Latest / NO FILTER / Joe Rogan Experience _2525 - Nick Bostrom

Transcript

Joe Rogan Experience _2525 - Nick Bostrom · NO FILTER

0:04
Speaker 1: The Joe Rogan Experience.
0:06
Speaker 2: Train by day, Joe Rogan podcast by night, all day.
0:09
It's great to see you again.
0:14
Speaker 1: Yeah, good to see you.
0:16
Speaker 2: So since the last time we talked, we spent a
0:19
lot of time where you were trying to explain to
0:22
me simulation theory and why the probability of simulation theory
0:27
is more likely than it not being a simulation.
0:33
Speaker 1: Yeah, it was, what, five years ago or something?
0:34
Speaker 2: I think it was six. Five or six, somewhere along
0:36
those lines.
0:37
Speaker 1: I mean, a lot of things happened in the world
0:39
since then.
0:40
Speaker 2: Yes, yes, a lot of things. I mean, for example,
0:44
back then, I think we, did we even talk about AI?
0:47
Speaker 1: It probably came up a bit, but.
0:52
Speaker 2: It wasn't this thing looming over civilization, which is really
0:55
kind of fascinating when you think about the fact that
0:57
it's only been six years, and in six years, like,
1:00
what a massive jump and some new technology in our life,
1:05
like just sort of like the internet where it like
1:08
crept up on us. We just accepted that it's a thing,
1:12
but that this thing has gotten massively entangled in every
1:18
aspect of society and every aspect of people's lives in
1:22
a very short period of time. Yeah.
1:24
Speaker 1: I mean, things are like so much is happening now
1:27
that it's kind of a full-time job just to monitor
1:30
the situation.
1:31
Speaker 2: Yeah. And one of the things that you're talking about
1:36
is the positive aspects of it, right? You're talking about
1:40
like that this is probably going to be a net
1:42
good for humanity.
1:44
Speaker 1: Hopefully. I mean I think I take both cases seriously,
1:50
the sort of the risk side and the big unlock
1:54
if we get things right.
1:56
Speaker 2: I feel like we have the potential, like we're on
1:59
a whitewater raft, we have the potential to get to
2:01
our destination. But we also have the potential to flip
2:05
over and try to figure out how to get to
2:07
shore in freezing cold water and sort of rebuild.
2:10
Speaker 1: Yeah, yeah, yeah. That seems about right. Yeah, and it
2:15
seems like it's kind of speeding up a bit as well, right?
2:19
If you're sort of in this field, there's like every
2:22
few weeks there's a new model being released or something, right?
2:25
Speaker 2: Yes.
2:26
Speaker 1: I guess it's like if you imagine, I guess, commentating
2:29
some fight, right? That's been recorded and you're supposed to
2:32
do the voiceover. And the first round is like normal speed.
2:35
And then the second round is 2X. And then the
2:38
third round 4X. And it's just like a whir of
2:40
legs and arms.
2:43
Speaker 2: That's a great way to describe it.
2:45
Speaker 1: Yeah.
2:45
Speaker 2: It's just so strange how quickly it snuck up on
2:49
us in that there's two narratives that we hear. The
2:53
one narrative is we're in real trouble. And that this
2:56
thing is going to take over every aspect of society
2:59
and it's essentially going to be a superior life force,
3:02
a superior intellect that exists amongst us that we created
3:06
and we don't think that's wise. And then there's the
3:09
other side that is saying things like what Elon says
3:12
where he's saying we're going to have universal high income.
3:16
It's going to be so much prosperity that no one
3:19
is ever going to have to toil again. There will
3:20
be no more third world countries. There will be no
3:22
more– No more poverty. Like we can eradicate poverty with
3:26
the resources that we have on earth and we can
3:28
change what it means to have to work. Like just
3:32
to provide yourself with food and housing. That's all going
3:35
to be easy and free. And then everything else is
3:38
going to be you have to find a purpose in
3:40
your life. My problem is– and I love Elon. But
3:44
the people who have that perspective are all making money
3:46
off AI. They are all invested like heavily, Marc Andreessen.
3:52
All these people that have this rosy view of it,
3:55
they're all invested heavily into it. So when someone like you,
4:00
who's not necessarily in that camp, you know, that is
4:03
more of a true objective analyst of what's going on,
4:08
when you have a positive aspect or a positive viewpoint
4:11
of it, I get a little more excited.
4:14
Speaker 1: Yeah, well, I mean, the truth is we don't know, right,
4:17
how it will pan out. So, I mean... I think
4:23
there are these scenarios where we unlock this enormous boost,
4:28
both to economic productivity, but then across medicine, entertainment, environment, travel,
4:34
all kinds of things. And like a tsunami of wealth
4:37
just kind of flows through and lifts all boats. And
4:41
then the idea of human work becomes an anachronism and
4:47
where you have machines that can do everything that we
4:50
can do physically and mentally, and do it much better
4:52
and cheaper. And so I think in those scenarios where
4:56
this really works, the transformation is a lot deeper. I mean,
5:00
so that kind of layers to the onion. So, like,
5:03
the most superficial level is, well, they automate your job,
5:06
so what are people going to do? What are they
5:09
going to do for purpose? Well, that comes a little
5:11
bit deeper. First, like, the most superficial level is people
5:13
just wonder, where will I get the job if the
5:15
robot replaces me? And then the superficial level of that conversation,
5:19
I think, is, well, you need to retrain workers so
5:21
they can work in other fields. And maybe there needs
5:24
to be, I don't know, employment insurance whilst they are
5:27
being retrained or something like that. But once you think
5:30
through where this ultimately leads, you think, like, it's not
5:32
just a few jobs, I think, but it's really everything
5:35
that humans can do to a good approximation. With maybe
5:41
the exception being where the consumer has a direct preference
5:45
that the particular... product or service be done by human
5:49
like priest prostitute and politician um those might survive.
5:57
Speaker 2: Well politician hopefully not hopefully ai can handle that all
6:01
without any corruption.
6:02
Speaker 1: Yeah but somebody's gonna take credit for.
6:04
Speaker 2: It this is true and code it yeah it's um.
6:09
Speaker 1: Um And so then you get to these more profound questions,
6:14
I think, about meaning and purpose and what does a
6:17
human life look like at technological maturity. Right.
6:22
Speaker 2: Also, what is experience? Does experience have to be measurable?
6:27
Do you have to touch it? Do you have to
6:30
measure it and weigh it? Or is virtual experience still experience?
6:35
Like if you have a very full and enjoyable and
6:39
fulfilling virtual life Is that enough? You know what I mean?
6:43
Do you have to do everything in the material world
6:47
or can you find happiness in a virtual world that
6:54
doesn't currently exist but we could clearly see the technology
6:58
if it expands? There's going to be the possibility of
7:04
experiencing a matrix-type reality.
7:07
Speaker 1: Yeah, I think I see where you're going. But, yeah,
7:10
I mean, I think, like, different strokes for different folks.
7:12
If you imagine a world where presumably these virtual worlds
7:17
will be very rich and deep and fascinating, but some
7:21
people might just like the idea of, you know, climbing
7:24
the real Mount Everest rather than, you know, being in it.
7:26
Speaker 2: Most certainly. But, however, there could come a time where
7:32
people are locked out of regular reality. Like, this is
7:35
worst-case scenario with A.I., is that you have your needs
7:40
taken care of, but there is no purpose. There's nothing
7:44
to do. There's nothing to do because AI is taking
7:46
care of all aspects of society. Other than recreation, there's
7:50
nothing to do. And then this one recreation comes along
7:53
that's not reality, but it's way more fulfilling and exciting
7:56
than any other aspect of reality. And an example of
8:00
that would just be in a very minor way, Your
8:05
phone provides you with that sort of an escape and
8:07
it's not even that thrilling and yet it's massively addictive.
8:11
People are on their phones six hours a day. If
8:15
we come up with a virtual reality that's way more
8:19
exciting than regular reality, everyone is going to hop on in.
8:24
Speaker 1: Yeah. I mean I think to some extent this will–
8:28
Maybe be the case whether you go the virtual path
8:31
or the physical path. I mean you could imagine a
8:33
future where there are like resorts and people are sort
8:35
of lying in beach chairs, sipping drinks all day long.
8:38
And like that's also a kind of checking out, right?
8:41
Speaker 2: Yeah. That's not fulfilling. The thing about a virtual reality
8:47
is you don't have to even live within our physics, right?
8:52
I mean, you can fly. You can do anything.
8:54
Speaker 1: It's a larger space of possibilities. But either way, it
8:56
could be like a passive existence in physical reality where you're, like,
8:59
launching at the beach. Or it could be like a
9:02
video game. Or in video games, like, you could have, like,
9:04
really intense, every straining, every fiber to try to succeed
9:07
in this virtual environment. Or, like, one where you're just
9:09
kind of floating on some cloud in a drug-like state.
9:13
Speaker 2: But I don't think people are going to be interested
9:14
in that because just the way the human mind works.
9:18
like what kind of video games are people attracted to?
9:20
It's because a video game is essentially a proxy. It's
9:23
an approximation of that. You're just watching it on a screen,
9:27
but you're sort of forgetting the fact that you're watching
9:29
it on a screen. You're just concentrating on this 3d
9:32
reality that you're running through with a machine gun or
9:34
whatever you're doing.
9:35
Speaker 1: Right.
9:35
Speaker 2: That's what people are interested in. They're not interested in
9:38
these games. We just like float over the earth and
9:40
like fly and, and, and just eat bananas. No, they're,
9:45
they're interested in thrilling things. And if, That gets provided
9:49
in a virtual way. The amount of people that are
9:51
going to just say this is better than regular boring
9:55
life or we're a slave to a computer.
9:59
Speaker 1: Yeah. I mean there are both kinds. OK.
10:01
Speaker 2: We're back. Where were we? Oh, yeah.
10:05
Speaker 1: I was just saying that there are also these computer
10:07
games that like provide a more passive experience.
10:10
Speaker 2: Yeah.
10:11
Speaker 1: Like not really doing much and it's more kind of
10:13
zoning out.
10:14
Speaker 2: But most people aren't interested in those. Most people are
10:16
interested in Call of Duty. They're interested in these wild
10:20
first-person shooters where you're running down hallways and everything is
10:23
exciting and thrilling. Some sci-fi game or Half-Life, something like that.
10:29
That's what people are interested in.
10:30
Speaker 1: Well, I mean I think there are like some personality
10:34
differences in what path you would take there. But either way,
10:38
I think a lot of the choices that people would
10:40
make currently depend on what they are sort of– designed
10:45
to respond to by feeling good about it. So, like,
10:48
if you enjoy one thing, you choose that. If you
10:50
enjoy another, no. In this condition of technological maturity, if
10:55
we imagine sort of a future civilization that has developed
10:59
all possible technologies to their maximum, then amongst the affordances,
11:05
amongst the things they could do is, like, they wouldn't
11:07
just have control over the external environment, things around them,
11:12
but also their own biology and brains so if you
11:16
were one of these people and like maybe right now
11:19
the only way you get enjoyment is by you know
11:21
getting like crazy drunk or doing sort of things you
11:24
don't really approve of maybe but that's that's actually what
11:27
sort of lights your fire like you could imagine redesigning
11:32
yourself to take the same kind of pleasure but in
11:35
some kind of contemplating the beauty of the universe or
11:39
like appreciating the goodness in the heart of others or
11:42
like some more sort of noble aspiration, like solving abstract
11:47
mathematics rather than, you know, playing first-person shooters. You would
11:50
have a choice, like whether you would get your thrills
11:54
from one kind of activity or the other.
11:57
Speaker 2: Yeah, which brings us to the question, like, what does
11:59
it mean to be human? Like for people finding beauty
12:03
and joy in, you know, interesting and fascinating things and,
12:09
how to be a better person and all the different
12:12
aspects of human life that we think about when we
12:16
think about people, when we think about noble aspects of humanity.
12:19
But doesn't that have to exist in conjunction with the
12:24
worst aspects of society for us to appreciate it? It
12:27
seems like this is a part of the human condition
12:30
is that we have to have crime so that we
12:33
can appreciate peace. We have to have war so we
12:35
can appreciate peace. We have to have hate so we
12:38
could appreciate love. And we're never without one or the other.
12:42
They're always both together in constant conflict and we're always
12:45
nobly hoping that the good wins out over the bad.
12:50
And this is part of the struggle of being a
12:52
human being. And if we just completely eliminate that struggle,
12:56
we're going to have to find some– we're going to
12:59
have to be a different thing because what we are
13:02
is this very strange territorial primate. and who's endlessly curious
13:08
and this territorial primate is moving in a certain direction
13:13
and that direction seems to be a better society. It
13:17
seems to be over time. If you look at statistics
13:20
from thousands of years ago to today, there's less violence,
13:23
there's better medicine, there's better education. It's moving in a
13:27
better and better and better direction. But it's in a struggle.
13:30
That's part of why it moves. If all of a
13:33
sudden there's no struggle and everyone is this wonderful, enlightened being,
13:37
like what are we? Because we're going to be a
13:39
different thing than what we are right now. And if
13:42
you love music and if you love art and if
13:45
you love novels and all these different things that come
13:49
out of the human condition, those things come out of struggle.
13:53
Those things come out of confusion and pain and heartbreak
13:56
and love and joy and all that stuff all piled
14:00
up together. Without that, like what are we? And so
14:04
if we are moving in this direction technologically and we're
14:08
not moving as fast biologically, do we merge? Because that
14:14
seems to be what I think. I think that if
14:17
this thing goes the way it continues to go, the
14:21
bottleneck is going to be human biology. Yeah, yeah.
14:24
Speaker 1: There is a kind of paradox embedded in our efforts
14:29
to make progress. So, like, there are all these kind
14:32
of scientists and people working to develop better technologies. And
14:35
throughout the economy, you know, in some company, maybe you
14:38
try to figure out how to make some process a
14:40
little bit more efficient so you can serve customers better.
14:42
And all of this is designed to solve problems, right? Like,
14:47
if you sort of extrapolate that to its logical endpoint, right,
14:52
you would imagine we would have perfect technology that can
14:55
do everything, solve all the problems, like, presumably with AI
14:58
and automation.
14:59
Speaker 2: Right.
15:00
Speaker 1: But then you end up in this condition where there
15:02
is kind of nothing left for us to do, you
15:05
might think. And so although it looks like we have
15:07
these strong reasons to push forward in this direction, if
15:10
you actually look at the end point, if we succeeded
15:13
to many people, it will look kind of unpalatable and
15:17
like this kind of future where all the problems are solved.
15:20
And nevertheless, that does seem to be the direction that
15:26
we are headed in, probably the direction that we should
15:29
be headed in.
15:29
Speaker 2: Yeah. Wouldn't that be a better goal if there was
15:32
zero murder, zero violence, zero crime, zero lying, zero corruption,
15:40
and human beings all worked in coordination with each other
15:44
just like in a sense of unique syncing and harmony?
15:48
Speaker 1: Right.
15:48
Speaker 2: That would be way better. I think so.
15:51
Speaker 1: But it does– Like if you actually stare at that situation,
15:55
it does have these slightly unpalatable qualities because you might
16:00
think it looks kind of bland, right?
16:02
Speaker 2: Not necessarily though. It's just it looks bland if you
16:06
think about what we're experiencing now, like all the excitement
16:10
and the weirdness of uncertainty and of not knowing– and
16:15
also the negative aspects of our life in conjunction with
16:18
the positive aspects of our lives. So we contrast the
16:21
two of them and you– really appreciate good people after
16:23
around a bunch of fucking assholes. You know, if that doesn't,
16:27
but it doesn't have to be that way. This is
16:29
just what we are dealing with. This is my position
16:31
on work too. Cause everybody in their head is like,
16:35
what happens when AI takes all the jobs? And I'm like,
16:39
do we have to have a fucking job? Isn't that
16:42
a human invention? Like, why do you have to have
16:45
a job? It seems crazy that our main focus is
16:49
on housing and food and And like most people are
16:53
basically just working for that. Most people that are struggling
16:58
check to check, basically housing and food is their daily labor.
17:02
If that's removed, wouldn't enough people figure out what to
17:07
do with their time?
17:09
Speaker 1: Yeah. I think so. Like, yeah, there will be some
17:12
transition and discomfort and change.
17:15
Speaker 2: But this is made up.
17:15
Speaker 1: Ultimately, that is like, yeah. I mean, I think it's like,
17:18
so slavery is really bad, right? Like, But wage labor
17:21
is a sort of slavery light in a sense. You
17:24
have to sell a third of your waking day just
17:28
to get money to pay for necessities.
17:30
Speaker 2: And the people that are trying to make the most
17:33
money make the conditions for their employees as shitty as
17:36
possible because it costs the least amount. They don't think
17:40
about it like, listen, maybe we make less money. But
17:43
we have an awesome experience for everybody that works there.
17:46
Nobody thinks like that. They all think like we have
17:48
to make the most money because we have shareholders and
17:50
we have a responsibility and these people need to make
17:52
more money every quarter.
17:53
Speaker 1: Yeah, although sometimes you do need to compete for talent.
17:56
Speaker 2: That's true too. But only the top talent. That's why
18:00
the CEOs get all the bread and all the people
18:01
on the assembly line get fucked.
18:03
Speaker 1: Some employees get treated well, right? What would you say?
18:05
Speaker 2: Jamie is a different kind of employee. Jamie is literally
18:09
the co- I mean he's the co-person on this show.
18:13
That's different. But some dude who's out there stocking boxes
18:18
for Amazon, that's different than being the CEO.
18:22
Speaker 1: I think ultimately that would be a huge liberation and
18:25
kind of restore dignity. You should be in control of
18:28
your own time. It's kind of almost the most... Aside
18:32
from your own body, your own time, your own attention.
18:34
Speaker 2: We have to teach people discipline. That's the thing. If
18:38
you give people the opportunity to just do nothing, we...
18:41
We do. We're primates. We'll find a way to– we're
18:46
genetically sort of designed to conserve resources. You don't want
18:49
to overuse resources for no reason and do a bunch
18:52
of shit because you won't survive. You don't have enough food.
18:55
So we're sort of designed to like know how to
18:57
be lazy. And so we're going to have to find
18:59
things that give us purpose or we'll get depressed. This
19:02
episode is brought to you by Visible. It's officially summer, folks,
19:06
and boy, what a summer we are having. It feels
19:09
like one of those good old-fashioned summers where you just
19:11
want to enjoy every moment and keep up with the action.
19:14
Your best companion this summer, Visible Wireless. Visible gives you
19:19
unlimited 5G data and unlimited hotspot powered by Verizon for
19:25
just $ 25 a month. So you can make this the
19:28
summer of fun and savings. Visible is more than just
19:31
a wireless plan. It's unlimited wireless designed to always keep
19:35
you connected with no contract to tie you down. Switch
19:39
today at visible.com plan start at $ 25 a month or
19:44
get our premium visible plus pro plan and save $ 10
19:49
on your first month when you use promo code ROGAN.
19:53
Terms apply. See visible.com for plan features and network management details.
19:58
Speaker 1: Yeah. I mean, so it used to be the traditional
20:00
idea like of British aristocracy is like that you wouldn't
20:03
have to work for a living. It was kind of
20:06
an unfortunate necessity for many people. That they had to
20:09
work for their daily bread, right? But the ideal way
20:12
of being human was that you just had money. And
20:15
then you spent your time doing other things like maybe
20:18
you were involved in politics or maybe you had like
20:20
an art collection or you did your gardening or entertained friends.
20:24
Speaker 2: That's the best aspect of it. But when you think
20:27
about British aristocracy, don't you think about fuck-ups? I do.
20:31
I think about guys that are just drunk. They drive
20:33
their Ferrari into a lake and they accidentally drown their
20:36
friend and then they pay off the cops and they
20:39
get away with it. That's what I think of. Well, yeah,
20:42
there's that.
20:42
Speaker 1: But there were also these kind of, I don't know,
20:44
amateur scientists or eccentrics who had their weird thing and
20:47
it kind of had it too. So I guess it
20:49
brings out like the more freedoms of what it has,
20:51
the more they can reveal their true nature.
20:53
Speaker 2: Well, it also depends on how they're raised. Who are
20:58
their parents? What values did they instill in them when
21:01
they were young? Did they explain the value of hard
21:04
work that it's actually good for you? And if you
21:07
just find something you really want to do and you
21:09
concentrate on it and really work hard at it, it's
21:11
actually very fulfilling. Yeah.
21:13
Speaker 1: So you could imagine, for example, the education system in
21:16
this world where we no longer need to work.
21:18
Speaker 2: Right.
21:19
Speaker 1: Presumably it would need to be redesigned from scratch, basically,
21:24
because right now it's a kind of machine, right? So
21:26
you take children coming in kind of on a conveyor belt.
21:29
Speaker 2: Yeah.
21:29
Speaker 1: And then some processing is done, like sit at their desk,
21:32
here is this assignment, here is a grid, and then
21:34
they come out with a quality label attached.
21:37
Speaker 2: Yeah.
21:37
Speaker 1: Right? And then they are meant to be productive workers
21:40
that you could put in an office and they will...
21:43
So this kind of is an unfortunate... fact about the
21:47
current condition that I mean the world needs a lot
21:49
of these office workers that can do all these tasks
21:51
in the economy so like it's we have to have
21:54
until it doesn't until it doesn't and at that point
21:56
it would be absurd to keep doing this right then
21:59
you could imagine changing education so that it would be
22:03
much more about like how can you learn to live
22:08
well a life of leisure like you imagine You're right.
22:12
Speaker 2: Stay stimulated.
22:13
Speaker 1: Cultivating the art of conversation, appreciation for art and music, hobbies,
22:19
physical wellness, like nature, the ability to set your own
22:24
goals and take your own initiatives to develop true friendships
22:29
with people. These are things that are not taught in school,
22:32
but you could imagine that being the curriculum, like spirituality,
22:35
like all kinds of things that would then equip people
22:38
to sort of use their freedom, their wealth, their free
22:43
time for some kind of actually meaningful and beautiful activities.
22:50
Speaker 2: For sure, preemptively, we should be kind of teaching people
22:54
like that now. Like if young people are coming up now,
22:57
the world that they're– if you're in first grade right now,
23:00
12 years from now, when you're graduating from high school,
23:04
like the world is a totally different place. And this
23:07
idea that being a productive worker who can sit still
23:10
for eight hours a day is– that's the best way
23:12
we should teach kids. That seems crazy. If we're really
23:16
like seeing the world that you're describing and that most
23:19
people seem to think is oncoming. Yeah, yeah.
23:22
Speaker 1: I guess the timeline is a little bit uncertain. So
23:26
you don't want to end up in a situation, right,
23:27
where you sort of now you find yourself. You're 20
23:29
years old. It's time to go out in the world.
23:31
You have no skills. The AI revolution hasn't yet quite happened.
23:35
Speaker 2: Well, isn't that a liberal arts degree anyway?
23:37
Speaker 1: Yeah. How is that?
23:39
Speaker 2: There's a lot of silly degrees that people get right
23:42
now that are useless anyway. And they spend a tremendous
23:45
amount of time working on getting it. And then when
23:49
they get it out, there's no jobs for that degree.
23:51
Speaker 1: Yeah. So they were a little bit too early maybe
23:54
in hopping on this.
23:54
Speaker 2: Maybe. Or, you know, there's also traps. There's things that
23:59
you can get really excited about and do. Like if
24:02
you want to play professional bowling, you want to bowl professionally. Well,
24:06
the amount of money you're going to make is very
24:07
limited because the best bowler in the world. Like what
24:10
does the best professional bowler in the world make? Let's guess.
24:16
I say $ 100, 000. Yes, the professional bowler. We used to
24:21
know a guy. Ari's friends with a guy. Tommy. God,
24:24
I can't remember his last name. Do you remember his
24:26
last name, Jamie? He was a professional bowler, but he
24:29
used to come to the comic store and hang out.
24:30
Really nice guy. And just loved comedy.
24:33
Speaker 1: I don't know. Maybe the very best one makes a
24:35
pretty decent living. And then the third best, it's just barely.
24:39
I bet barely.
24:39
Speaker 2: I bet the best guy probably makes maybe $ 100, 000 or $ 200, 000
24:42
at the most. And that means the 30th guy in
24:47
the world is fucked. You have to have another job
24:50
to pursue that.
24:51
Speaker 1: What does he make? $ 200, 000? Oh, $ 400, 000 in big season.
24:53
Speaker 2: $ 450, 000. including prize money and sponsorships.
25:00
Speaker 1: EJ Tuckett made $ 437, 540 for the 2025 season, with.
25:02
Speaker 2: Several others between $ 190K and $ 270K in prize money alone.
25:04
Top tier. So the top tier guys make a good living.
25:06
But how many guys are there? You know what I'm saying?
25:24
Speaker 1: It's probably like a very thin sort of peak.
25:26
Speaker 2: So if you choose to go into bowling, it's not
25:29
like choosing to go into, you know, there's a lot
25:32
of other jobs. Like even if you're going to be
25:33
a basketball player, they're making a lot of money. You know,
25:36
if you're a really good basketball player and you pursue that,
25:39
like you could get very, very rich. But how many
25:42
of them get rich?
25:43
Speaker 1: Yeah, but then like, so maybe the money factor would
25:45
kind of drop out of the picture.
25:47
Speaker 2: Yes, and it would just be what you're interested in.
25:50
And then maybe even it would, I mean the opportunities
25:53
are endless.
25:54
Speaker 1: And what other people are interested in to some extent
25:56
because like I think part of what people are competing
25:58
for is also prestige and status.
26:00
Speaker 2: Yes. That would be a thing, right? Like status would
26:02
be more important. It's like who's the best at this thing?
26:06
Like a bunch of friends who play golf. Like Jamie
26:08
plays golf all the time. Golfers are all like comparing
26:11
each other's scores and they're playing. They're competing in this game.
26:16
And they think about that more than they think about work.
26:19
Like people who love golf, they fucking hate work. Like
26:22
I used to say that about comedy back when I
26:24
first started. One of the things that I noticed is
26:26
the guys who really got into golf, their careers kind
26:29
of stalled because they were more excited about playing golf
26:33
than they were about writing jokes and going on the road.
26:35
I was like, OK. So if the average person doesn't
26:39
need food and housing anymore from labor, if that's gone
26:43
and now you just get it. And so now you
26:45
could just go do things. we just have to teach
26:48
people to be excited about stuff. We have to teach
26:50
people the value of curiosity and finding things that are
26:55
interesting to you. And then the value of just education
26:59
for the sake of learning things, because it's interesting, just
27:02
pure satisfaction of curiosity, which is a beautiful thing. Like
27:07
that would be great for everybody. If we, instead of
27:10
Learning things because your teacher tells you you have to
27:13
learn it. Well, there's always going to be people that
27:14
just naturally gravitate towards mathematics and they're really fascinated by mathematics.
27:19
And there's always going to be people that are just
27:20
naturally gravitating towards history. They're like, how did we get here?
27:23
What is this? How did this start? Who wrote that?
27:26
Who's the first guy to figure that out? And that's–
27:29
you're going to naturally go towards that and just be
27:33
educated for the sake of satisfying your curiosity. And maybe
27:36
we'll have a more– balance society because people will actually
27:40
be able to just pursue their interests. We have to
27:42
teach them to do it. That's what I think is
27:45
going to be really important about young people in the future.
27:48
Teach them to actually pursue their curiosity rather than just
27:53
squash any desires they have for novelty and for interesting
27:58
things and just be able to work all day doing
28:00
something you hate because that's what it means to be
28:03
an adult.
28:03
Speaker 1: Yeah, I think cultivating that would be the way forward
28:09
in this case.
28:10
Speaker 2: And it seems very possible. It's doable. We know curious people.
28:16
Speaker 1: At least to some extent. I think there's probably also
28:18
some differences in how different folks are wired. Some people
28:21
get bored very easily. Others find everything fascinating. Some people
28:25
are like just naturally depressed or have low mood. Others
28:28
are sort of ebony.
28:28
Speaker 2: Yeah, but those people that are naturally depressed, how much
28:31
would their life change if they were coached at a
28:33
very early age the value of exercise? And if they
28:37
started running, if they started doing yoga and they started
28:40
feeling much better and they weren't naturally depressed anymore. And
28:42
then you change their diet and you start adding in
28:44
vitamins and nutrients and stop giving them processed foods. And
28:48
all of a sudden they feel better because the concept
28:51
even of depression, like what does that mean? Well, it
28:54
varies so wildly. And when you're talking about people that
28:58
aren't taking care of their bodies and that's not thought
29:00
of as a primary factor in why depression exists in
29:04
the first place, well, this is a very unscientific approach
29:07
then because we know that it has a profound impact
29:10
on human beings. So we're pretending that this is like
29:12
this one-size-fit-all. No, I think that should be also a
29:16
part of teaching people how to be a human being.
29:19
Chase your curiosities. But also this is why you're not
29:22
feeling well. And that if you just develop this ability
29:25
to get some momentum going and just show up at
29:27
that yoga class every day, you'll feel way better. And
29:32
you don't have to take a pill.
29:34
Speaker 1: I think, yeah, many could be a lot improved if
29:38
they were taught to sort of take. care of their
29:40
basic physiology.
29:42
Speaker 2: Everyone can be improved. 100% of them can be improved.
29:45
It's a part of being a human being.
29:47
Speaker 1: I mean, I don't know. I think there is also
29:49
some people who might just have some imbalance of some
29:52
neurotransmitter because of some genetic mutation. 100%. Or maybe they
29:56
don't even have the energy to engage in these activities
29:59
that would make them feel better. But then I think
30:01
there would be much better medicines for helping.
30:04
Speaker 2: That's where Adderall comes in. Let's go. Give them some Adderall.
30:09
Speaker 1: Get them fired up.
30:11
Speaker 2: Get them to start running. I'm kidding. But what we're
30:15
saying is all doable. We're not asking people to breathe underwater.
30:18
What we're saying is that There are people on earth
30:23
that live like that and this idea that living just
30:27
for a paycheck so that you can cover your food
30:30
and your housing, which we've always thought of as being
30:34
just an undeniable part of being a person, an adult.
30:37
This is what you have to do. You have to
30:38
pay for your food, pay for your housing. But if
30:40
we do get to a point where the structure of
30:44
society is now run by hyper-intelligent artificial intelligence … you
30:50
wouldn't need most jobs that people have to do that suck.
30:54
And in order to get our society to this point,
30:57
if you wanted an iPhone, you needed some people that
31:01
were in a factory somewhere putting it all together. You
31:04
needed someone who's designing it. You needed someone who's sitting
31:08
there in the office trying to figure out how to
31:10
market it. You needed all those jobs. But when we
31:12
don't need all those jobs anymore, then things are going
31:14
to be very interesting. And that's what I'm saying. We
31:17
need to become a different thing. Like It's kind of
31:20
true and kind of not. Like you can be a
31:23
human being and live in that world. But we're going
31:26
to have to reeducate people how to be a human being.
31:30
It's going to– our education system specifically in this country
31:35
is just designed to make factory workers. Specifically, like there's
31:38
a– There's a real history of it. Like we know
31:41
why they made it this way. They made it this
31:43
way and they got people in really early so they
31:46
could get people set up for jobs. They just want
31:48
people to work. Yeah.
31:49
Speaker 1: So we are kind of currently ill-suited for really thriving
31:55
in an environment of abundance and for enjoying life. Because
32:00
both at deep biological timescales and during the life cycle,
32:05
span of a single individual, there are all these pressures,
32:07
necessities that kind of force us to become a certain type.
32:11
I mean, we talked about the education system training people
32:14
to be the kinds who can sit at the desk
32:16
all day long and perform tasks. Biologically, we've evolved drives
32:22
to be lazy, to conserve energy, to eat as much
32:25
as we can. And now in the modern environment where
32:28
there are fridges everywhere, it causes problems metabolically. And In
32:35
terms of enjoying life, for many people, there is this
32:38
hedonic treadmill, right? So you achieve something, some improvement, there's
32:43
a spike of happiness, and then you sort of go
32:46
back to baseline very quickly. You start to take for
32:48
granted all the blessings of life, which makes it very
32:51
difficult to actually achieve a state of permanent happiness and felicity.
33:00
And it's kind of been necessary because there needed to
33:03
be this motivation to keep striving for the next thing. Now,
33:07
once you've actually achieved all the things, then maybe that
33:09
becomes a kind of dysfunctional, right? Like why keep wanting
33:14
to strive for the next thing if all the things
33:16
have already been achieved, at least in a certain domain.
33:20
And so I think as we move deeper into this
33:24
hypothesized future, like where we get all these magical technologies,
33:29
then at some point, probably some transformation of human nature
33:35
would have to go along with that. First, maybe cultural
33:37
changes to sort of equip people for a life of leisure.
33:40
Speaker 2: Yeah.
33:41
Speaker 1: And then ultimately, maybe more profound changes to our very biology.
33:45
And so that imagine if you solve aging and you
33:48
can now live for thousands of years, right? Like maybe
33:51
the way our memory work is not really set up
33:53
for that. Maybe we would just go stale after 200
33:56
years with our current brains. Like we just get stuck
33:58
in a rut. We don't know because we haven't lived
34:01
for 200 years, but that could easily turn out to
34:03
be the case with the humans. So then maybe at
34:05
that point, you would need to sort of do something
34:07
to sort of add more cognitive resources, more flexibility, do
34:12
some sort of psychedelic reset or something to sort of
34:15
keep the flexibility going for longer periods of time.
34:19
Speaker 2: I'm glad you said a psychedelic reset because it would
34:22
be very funny if the main tools that we have
34:25
for navigating this are all illegal. Because I think they
34:29
might be and specifically with psilocybin and DMT and probably
34:35
Ibogaine as well. Like if we wanted chemical tools to
34:39
navigate a new reality, those are probably the best ones
34:43
that we have available and they're all illegal. And I
34:46
think you're right. I think if we're going to be
34:48
able to navigate this correctly, we have to kind of
34:51
change the way we interact with each other, what it
34:54
means to be a person. But I wonder how much
34:57
of the conflict that we have is a direct result
35:00
of this inherent struggle that so many people have. And
35:04
I mean, there's a direct correlation between extreme poverty and
35:07
extreme crime, you know, specifically in this country. If you
35:10
look at the areas of extreme poverty in this country,
35:13
they're also the areas with extreme crime. And I wonder
35:17
how much of that would be completely alleviated with a
35:20
complete lack of poverty. And we've always assumed that if
35:25
you're going to have a functioning society that you're going
35:30
to have slums. Why? Like why? They don't serve any function.
35:36
It's not a good thing. Well, it's because some people
35:38
are always going to be making bad decisions and some
35:41
people are always going to be going down the wrong
35:43
road and crime and this and that. Right. But they're
35:47
still just humans and some humans don't do that. So
35:49
wouldn't it be better to figure out a way to
35:51
not have humans ever do that anymore? And there seems
35:55
like a way to do that. It's, again, not asking
35:57
people to breathe underwater. We're trying to figure out why
36:00
do some people never engage in crime? Why do some
36:04
people live these really fulfilling and interesting lives? Probably because
36:08
they were exposed to it when they were really young.
36:10
Probably because they weren't exposed to very bad environments and
36:13
very bad crime and very bad poverty. And how much
36:16
of that would change if there was no more poverty?
36:20
It sounds like such a little fairy tale, child-like Oh,
36:24
one day we'll have no poverty. But that's doable. If
36:28
everyone is alive, right, OK, that's alive right now is
36:31
not starving to death. That means we figured out a
36:34
way to at least the very least get you resources
36:37
so you could feed yourself, right? No matter how dysfunctional
36:41
things are. All that has to do is you get
36:44
ramped up a few steps. I am. And now you
36:47
have no one ever worried about being fed, no one
36:49
ever worrying about not having a place to sleep. And
36:52
then you have to find purpose. But it seems like
36:55
there's a lot of people that find purpose without having
36:59
a financial price tag attached to it. Just by what
37:02
we're talking about with golf or you could be like
37:05
really into writing books. You could write books all day
37:08
long and people are always going to want human created fiction.
37:12
People are always going to be interested in the way
37:13
other people think about things. You'll find there's plenty of
37:17
things to do. There's plenty of games to play and
37:20
plenty of skills to learn. The idea that the only
37:24
reason why we work is to eat and to not
37:26
get rained on seems nuts.
37:30
Speaker 1: Yeah. It is, I guess, to some extent, an open
37:33
question to what extent people will always want to read
37:37
human fiction or to prefer the human-generated outputs. It might
37:44
be just because current AI-generated writing is kind of lacking
37:49
in various ways. It's often sloppy and boring. But if
37:52
it became really good, then maybe it would just be
37:58
much more fun. Like a movie made by an AI
38:02
might just be so much better and richer and the
38:04
lighting is perfect and the dialogue is sharp and it's
38:07
more funny and deep.
38:08
Speaker 2: Certainly possible.
38:09
Speaker 1: Then you go and watch this human-produced thing and it's
38:12
going like, Like most people don't go and watch sort
38:15
of film school students' productions.
38:17
Speaker 2: Right. But then there's like that movie Obsession. Wasn't that
38:20
movie made like very inexpensively? And it's a huge hit, right?
38:25
And that's part of the thrill of it is that
38:27
this guy who was like a YouTuber, right, who created
38:30
this film and now this movie is a giant hit
38:34
and everybody is like super excited about it. Now, if
38:36
that same movie was made by AI, I wonder if
38:39
it would have that impact because it doesn't have the
38:41
human connection.
38:42
Speaker 1: Yeah. So that's possible that we keep getting interested in
38:47
these things because we sort of are really entangled with
38:52
the human world much more than we are with the
38:55
world of AIs. And I mean, for the same reason
38:57
you might be more interested in if like your brother
39:00
or friend did something. You might think, oh, this is interesting.
39:03
I want to check that out.
39:04
Speaker 2: Of course.
39:04
Speaker 1: Whereas with some random dude who made something slightly better,
39:07
it's like doesn't mean. Right.
39:09
Speaker 2: When you go to see your nephew play baseball, you
39:11
wouldn't really go see little kids baseball.
39:13
Speaker 1: No, no. Just randomly, right?
39:15
Speaker 2: When your nephew gets on first base, like way to go, Bobby.
39:18
Speaker 1: So I think these kind of social entanglements that we
39:22
have will be a big part of what gives structure
39:28
to lives in this condition where. the external constraints are removed.
39:33
Speaker 2: I think that there is a great value for the
39:36
human mind for whatever reason in getting better at things
39:39
and learning things. And I think that if we could
39:42
instill that in people at a young age, I think
39:46
it would be fairly easy to get people to pursue
39:49
that path. But we have to completely revamp our education system,
39:54
which should have been done a long time ago anyway.
39:56
The idea of having unenthusiastic, underpaid people being any percentage
40:02
of what children encounter for most of the day when
40:05
they're in their most formative period is insane. It should
40:09
be a rich, exciting, enthusiastic journey in how to be
40:14
a better person and how it's exciting to learn things
40:17
and how it's exciting to get better at things and
40:19
about how anybody can get better at things.
40:21
Speaker 1: I mean I hated every day of school.
40:24
Speaker 2: I hated it. I hated it. I didn't even know
40:26
that I was interested in learning anything until I got
40:28
out of school.
40:29
Speaker 1: The same for me. I mean I only discovered– like
40:32
I also hated books and everything like that. for my
40:35
first 15 years of life because I associated that with school.
40:38
That was the kind of school thing. So it's only
40:40
then randomly one day I happened to walk into the
40:44
local library for no reason and pulled out the book
40:47
here and there. And then I discovered that there was
40:49
this whole world of science and idea and literature and
40:52
all of that, like very different from what we were
40:54
doing in school. And then that kind of opened the
40:56
gates for me to this.
40:58
Speaker 2: Imagine what a head start you would have had if
41:00
you had, like, a different kind of education with super
41:04
enthusiastic people who really love teaching children, are really good
41:09
at it, and that we reward them and that it
41:12
becomes a very prestigious position to be in rather than
41:15
what it is now. If you talk to some guy
41:17
and he goes, oh, I'm a high school teacher, you're like, oh,
41:19
poor bastard. How does he feed himself? That's immediately what
41:22
I think. Like good for you that you're doing that,
41:24
but also I guarantee you could probably be making a
41:27
lot more money and be happier doing another job. And
41:31
so that's a terrible way to start life off. And
41:34
if we just revamp that, then you have a bunch
41:37
of people that are interacting with life in a very
41:40
different way. And instead of being thought, I have to
41:42
get a job someday, I have to get a job
41:45
after school anyway to help my mom pay the bills
41:48
because we're struggling. So I got to contribute to the household,
41:51
even though I'm 16, I can't hang out with friends.
41:53
If all that stuff's out the window and now it's like, no,
41:56
what we need to find out is what is exciting
41:59
for you. Like what, What excites your mind, your specific personality?
42:04
What is it about life that's interesting? And let's expose
42:07
you to a bunch of different things that are exciting
42:10
and interesting that other people find value in. And let's
42:13
find out which way you gravitate because maybe you gravitate
42:16
towards chess or maybe you gravitate towards something completely different.
42:20
Maybe you're just really in a painting. Maybe painting just
42:22
lights you up and you like look at a canvas
42:24
and you just start fucking around with it and that's
42:26
your thing. There's a lot of people in this world
42:29
that find that. It's just they have to find it themselves, unfortunately.
42:33
Speaker 1: I think that's kind of inspiring. But I think the
42:40
train doesn't stop there. So if we think even more
42:43
further into this kind of condition of technological maturity, I
42:48
think in addition to sort of freeing up people, like
42:51
making it easy to produce the food and the houses
42:54
and cars that go– faster and don't pollute and all
42:58
of this stuff. Like, if you think through what maybe
43:01
a technologically mature civilization could do, like, so a lot
43:03
of these like learning, for example, is something that gives
43:07
people purpose, but maybe you would have the ability to
43:10
sort of download knowledge at the click of a button.
43:15
So rather than, like, if you want to learn advanced mathematics,
43:17
now you have to study for years, right? Do like
43:21
books with exercises and apply yourself and then eventually you
43:25
sort of unlock Yeah. But, like, maybe at technological maturity,
43:29
they would be like, okay, I understand mathematics. Okay, I'm
43:32
going to press that button.
43:33
Speaker 2: Boom.
43:33
Speaker 1: Now I understand mathematics.
43:34
Speaker 2: Like the Matrix. I know Kung Fu. Remember? Yeah.
43:37
Speaker 1: So if you had, like, you know, maybe some kind
43:39
of nanobots that could infiltrate your brain and then change
43:43
the synapses in just such a way that you're not
43:46
the same as you were before except you have these
43:48
concepts from abstract mathematics. Yes. You know, after 20 minutes
43:52
or something like that. Some superintelligence works out how to
43:55
change your synapses. to this new condition. And if you
43:59
were inclined to do various things because they give you
44:03
joy and pleasure, you could also think, well, I could
44:07
do those things and get the joy and pleasure, or
44:08
I could just push this button that activates the same
44:11
joy and pleasure. So there would be these shortcuts to everything.
44:16
And it looks like you would have a post-instrumental condition
44:20
where that would, at least at first sight, seem to
44:23
be no reason to do anything for the sake of
44:26
of achieving something else, because there would always be this
44:28
shorter path to that other thing, pressing the button. And
44:33
so that's a kind of deeper form of redundancy. It
44:35
seems that all forms of human instrumental effort would become unmotivated.
44:40
You could still go to the gym every day and
44:46
sweat for an hour, or you could just take the
44:50
pill that induces exactly the same physiological effects And the
44:54
mental effect like the calm or whatever you feel after.
44:58
Speaker 2: Or maybe even better, genetically engineer someone.
45:00
Speaker 1: Or genetically, yeah.
45:01
Speaker 2: So they don't have to take a pill.
45:02
Speaker 1: Yeah, yeah. And so then I think you're getting more
45:06
into these like fundamental questions of value. Right. Philosophical questions
45:10
about what ultimately is it that makes a life good.
45:15
Speaker 2: And what makes a life good for a primate because
45:18
that's what we are. Or a primitive version of primates.
45:22
I think we're going to become something different. And if
45:24
we do it through technologically induced evolution or biologically induced evolution,
45:29
we're still moving in a general direction. Just biologically, if
45:34
you look at ancient hominids and look at us, we're
45:36
clearly very different. So we're moving in a general direction anyway.
45:40
And when you look at the greys, like these prototypical
45:43
aliens that everybody sees with the big heads and skinny bodies,
45:45
what do they look like? Well, they look like if
45:48
we keep going, that's what we're going to look like.
45:50
We're going to be these genderless, sexless aliens. Muscleless little
45:53
skinny things with giant heads that communicate telepathically. This is
45:57
probably where we're moving. And if we're moving that way,
46:01
maybe we think of them as being an alien from
46:03
another– maybe that's just the general direction that primates go
46:09
once they figure out technology. They eventually realize, well, all–
46:12
first of all, the idea of having to have a
46:15
male and a female is kind of crazy, right? Why? Why? Well,
46:19
because we have to reproduce. Well, what if we don't
46:21
reproduce like that anymore? What if all reproduction is done
46:24
through engineering and there's no more sex, so there's no
46:29
more lust, so there's no more jealousy, so there's no
46:31
more ego? There's no more anything. And then there's a
46:34
hive mind because technology has advanced to the point where
46:37
the reason why they have these big giant heads, they're
46:38
essentially locked into everybody all around them all the time.
46:42
Speaker 1: Yeah. Well, is that what we would want though?
46:45
Speaker 2: Not us, but I don't think we'll be us anymore.
46:49
I think this– like if you went to chimps and
46:51
you said, hey, dude, one day you're going to be
46:53
on a plane eating peanuts, flying over the ocean. They'd
46:56
be like, fuck that. I don't want to do that.
46:57
I'll just stay here where the fruit is. You guys
47:00
are nuts. Like why do you want to do that?
47:01
And you'll be addicted to kratom and you'll be watching
47:06
YouTube all day long. Like, no, I don't want to
47:08
do that. That doesn't sound fun.
47:11
Speaker 1: We'd have a lot of banana stuff.
47:13
Speaker 2: Maybe. Or you have to get killed by the other
47:15
chimpanzees that want the bananas. They have chimp wars. They
47:18
come in and kill each other. So I think we're
47:22
going to be something different, and I think that's inevitable anyway.
47:25
If you believe in evolution, it's inevitable anyway that we
47:28
become something different. We're not a finished product.
47:30
Speaker 1: Yeah, no, I think that is true. We are not
47:33
a finished product. Now, the question is the thing that
47:36
we become or that becomes. Is that a result of
47:43
us choosing how– we want to be or is it
47:47
just these kind of impersonal forces like evolution selecting certain
47:51
types that might then end up leading to outcomes that
47:55
we actually don't want and so I guess the hope
47:57
is that we would be able to develop along a
48:00
path that preserves and develops the things that we actually
48:05
value about being human right maybe amplifies them then maybe
48:08
adds other things So there are many different possible things
48:12
you could develop in the future but that we sort
48:14
of select those that actually make us better rather than
48:17
just randomly different that we can sort of grow into
48:22
our ideals.
48:24
Speaker 2: But it's so funny how much value we put in
48:27
human condition, how much value we put in meaning because
48:32
it is valuable to us. When you think about like
48:35
what that means in terms of the amount of energy
48:38
it produces, the amount of impact that it has versus
48:41
the structure of the universe itself versus black holes and
48:46
stellar nurseries and things that are just infinitely more spectacular
48:51
than the human condition. But to us, it's like, oh,
48:54
what is meaning? What is meaning? What is meaning to me? Like, well,
48:57
what is meaning to me? You as you're a finite
49:00
biological organism that has to find meaning. Because you're sort
49:04
of tromping around through this weird world where eventually you're
49:07
going to die and you're going to leave your mark
49:10
and maybe reincarnation is real and maybe the afterlife is
49:12
real and nobody fucking knows. So you've got to have meaning.
49:14
But the universe itself, like a human only lives 100 years.
49:17
How much meaning can you have in 100 years?
49:20
Speaker 1: Yeah. I think we kind of die prematurely. I mean
49:23
we think of us as like first you become stronger
49:28
and more capable for 15, 20 years, right? Then biologically
49:32
you stagnate and then– Maybe you keep accumulating knowledge and
49:37
skills for another few decades, but just as you have
49:40
started to acquire a modicum of wisdom, your brain starts
49:44
to rot. And then everything is erased, either by Alzheimer's
49:49
or death. That's all gone, the whole lifetime of experience.
49:53
Speaker 2: There's your meaning.
49:54
Speaker 1: All the memories and the hard-earned lessons. So that seems
49:58
kind of sad and probably not the best way for
50:01
things to be. I think we would want to extend...
50:05
the human lifespan. So you could continue to grow up,
50:08
not just kind of reach 20 and then plateau there
50:10
for the rest of your life. But what if we
50:12
could continue to develop so that like at 80, you
50:16
were as much stronger and like be able to understand
50:20
more and move better and like just have the same
50:22
capacity increase as you had like between like zero and
50:26
20 and you just kept going.
50:28
Speaker 2: Well, these life extension scientists that are working on these things,
50:30
guys like David Sinclair, like they believe that that's a,
50:34
not just a possibility, but an inevitability.
50:37
Speaker 1: Yeah, well, so that's more like, I guess, preserving, preventing
50:40
or delaying that decay.
50:42
Speaker 2: Sure.
50:42
Speaker 1: Which is already very good.
50:44
Speaker 2: Not reversing. They're talking about reversing now.
50:45
Speaker 1: Yeah, yeah, but then you would go back to being 20, right?
50:48
Speaker 2: Biologically, yeah. But then you have engineering.
50:52
Speaker 1: If you have very long time spans, then you might
50:54
at some point want to continue to grow.
50:56
Speaker 2: Yeah.
50:56
Speaker 1: Like you might not want to just be stuck at
50:58
a 20-year-old human for 10,000 years. Like maybe eventually you
51:02
would want to become like slightly bigger in terms of
51:08
your ability to engage with the world.
51:11
Speaker 2: Of course. And if they can figure out how to
51:13
make people– well, there's already genetic engineering that's being done
51:17
in terms of increasing potential intelligence and IQ. They're already
51:20
doing that. And so this was a– I'm sure you
51:24
know about that thing that happened in China where this
51:28
one doctor got in trouble and because he had genetically
51:32
engineered some babies to be inoculated to HIV, but it also,
51:39
at the same time, gave them a far increased potential IQ.
51:44
It remains to be seen whether or not it actually
51:46
works as these guys, you know, become like 20 and
51:49
30 and we start putting them in chess tournaments and
51:52
see if it really did make them smarter. But if
51:55
The possibility of that is already being studied. They already
51:58
understand the differences between like what is– what's possible, what
52:05
we understand and that increases every day. They understand more
52:09
every day, more is possible every day. If that just
52:12
keeps going, well, you have a different version of what
52:15
the human consciousness is. You have a different version of
52:17
the human mind.
52:18
Speaker 1: Yeah. I think– The way AI is going, I think
52:22
that's the train that's going to reach the destination faster.
52:26
Speaker 2: Oh, yeah.
52:27
Speaker 1: And then once you do have super intelligence in machine substrate,
52:31
then that can then unlock all kinds of technologies, including
52:35
these biological technologies or nanotechnology or a host of other
52:39
things that could then sort of bring us up if
52:42
that's how we decide to go.
52:44
Speaker 2: Yeah.
52:46
Speaker 1: And uplift us into different... Like either biologically enhancing us
52:50
or like uploading us into.
52:52
Speaker 2: But again, it's that whitewater river raft. It's like we're
52:56
going down this whitewater and we might make it out
52:59
of this or it might crash. One of the weird
53:03
things about– I'm very fascinated with ancient civilizations. And one
53:09
of the weirdest things about ancient civilizations is when you
53:12
go really, really far back, a bunch of them, have
53:15
these depictions of kings that ruled for thousands of years.
53:20
And it's very strange stuff because the Egyptians and ancient Sumer,
53:27
there's a bunch of different depictions of these kings that
53:31
lived lives that were way longer than biological humans lived.
53:36
And then there's the flood myth or the flood story.
53:39
And then after the flood, biological life decreases pretty radically.
53:44
And it seems to get back to like a normal
53:46
version of what we assume now, which is like 100 years.
53:50
And one of the weird questions that these alternative ancient
53:55
historians have is are we missing the possibility that there
54:00
was a hyper-advanced civilization that existed tens of thousands of
54:05
years ago? And that what we're seeing now is not
54:07
a linear progression from caveman to human being with an
54:11
iPhone in 2026. that along the way there might have
54:15
been a very high level of sophistication and the evidence
54:18
of that might be the pyramids and some of these
54:21
other ancient structures that are mind-boggling. As much as you
54:24
try to explain them away, when you're dealing with 2,300,000 stones,
54:29
some of them are cut from a quarry 500 miles away,
54:32
80-ton stones that are in the ceiling, perfectly cut granite
54:37
where it looks like they have diamond drill holes in them.
54:39
It seems like there's some lost technology. And Every society
54:43
has this flood myth and every society that has this
54:46
flood myth, especially these ones that were very advanced somehow
54:51
or another like ancient Egypt, they have these depictions of
54:54
people that lived way longer than people live today. And
54:58
I wonder if human beings one day will realize like, oh,
55:03
if you keep going long enough, 100 years is silly.
55:06
Like people can– they will just figure out what it
55:09
is that makes people age and die, fix that. And
55:12
you'll live an insanely long time. And then if people
55:15
live an insanely long time and their capacity for reason
55:17
and logic increases, their capacity for intelligence increases, then you
55:22
have these insane technologies that were required to do things
55:26
like build the pyramids. And that might be what happens.
55:29
That might be a natural course of progression. The first
55:32
thing we realize is 100 years is not enough. Because
55:36
I'm 58 and I'm a moron still. Like why am
55:38
I learning every day? How come I'm getting better at
55:41
being a person every day? I should have figured this
55:43
out already. But you don't ever have time to figure
55:45
it out.
55:45
Speaker 1: You don't have time.
55:46
Speaker 2: There's not enough time.
55:47
Speaker 1: And you only get one run. Like I mean you
55:49
would take one trial run first, right, on life. Right.
55:52
And then like, okay, now I kind of know a
55:54
little bit how it works. Now let's do it for real. Well,
55:56
that's the concept of reincarnation. It's kind of scary. I
55:58
mean it's a bit crazy that each of us is
56:00
put in charge of a whole human life.
56:02
Speaker 2: Right.
56:03
Speaker 1: Like are we really– like, reach the level of, like,
56:06
if you want to run an airplane, you have to
56:08
go through these certifications and stuff, right? And then maybe
56:11
you can be a pilot. But, like, for a human life,
56:13
it's like every single person, like, okay, you have 100%
56:15
dictatorial control of this person. Like, they're completely hostage to
56:20
your whims.
56:22
Speaker 2: Yeah.
56:22
Speaker 1: And then that's like us. But that's the way it is. But, yeah,
56:26
having the ability to kind of try different things and
56:29
maybe do, like, have a kind of opportunity to do
56:35
a do-over if like if the life turned out to
56:39
be like every 50 years or something you have a
56:41
chance to kind of.
56:44
Speaker 2: Or maybe not even a do-over but it's the same
56:46
life but you can do whatever you want and you
56:49
don't have to be constrained to this idea that you
56:52
only have a certain amount of time and you want
56:53
to retire by your 65 we might look at that
56:56
as being like one of the first completely archaic notions
56:59
and the reason why people got it all wrong Like, well,
57:02
society will be better because people are going to live
57:04
way longer. And if you think about how much smarter–
57:06
how old are you? 53.
57:07
Speaker 1: OK.
57:08
Speaker 2: Now, the way you said that was sad. I'm older
57:13
than you, dog. Don't worry about it. But if you
57:15
think about how much smarter you are at 53 than
57:18
you were at 13, now imagine how much smarter you
57:22
will be at 353 or 1,053. And if that actually
57:23
becomes possible– If a person can live to be 100 years,
57:31
why can't they live to be 4,000 years? If they
57:33
can figure out what makes people age, if we really
57:37
can genetically engineer human beings, that's not– again, that's not
57:40
breathing underwater. That's just extending life. And if you extend
57:44
life and extend intellectual capacity and your ability to learn
57:48
and grow, holy shit. You're dealing with a 2,000-year-old person.
57:52
That's a completely different kind of thing. And that could
57:55
be real someday and that might be also– One of
57:58
the things that comes along with this new understanding of
58:04
just life in general with super intelligent AI.
58:07
Speaker 1: Yeah, no, for sure. Fixing aging and reversing it would
58:11
be amongst things that a technologically mature civilization would be
58:15
able to do.
58:16
Speaker 2: We just have to get people out of the idea
58:18
that it's vain. Why do you want to live forever?
58:21
Come on.
58:21
Speaker 1: That's what they say now, right? But when somebody actually
58:24
has the therapy and you could like, I'll hear. You
58:27
could either continue to get worse kidneys and more painful
58:31
knees and wrinkles, and you start to forget things. Or
58:35
you can take this rejuvenation serum that everybody else is using,
58:39
and they are full of energy and running around and
58:41
doing stuff. Do you still think it's vain? So I
58:44
think a lot of this is Stockholm syndrome, that you
58:47
are kind of faced with the inevitability of aging and
58:51
decay and have been for thousands of years. Yeah. you
58:54
develop a kind of problematic scaffold that reconciles you to
58:58
the inevitable.
58:59
Speaker 2: I like it. Stockholm Syndrome.
59:01
Speaker 1: Which is maybe adaptive, but only up until the point
59:04
where there actually might be something you could do to escape.
59:07
Speaker 2: Right?
59:07
Speaker 1: Then if you're too stuck in this mindset, it might
59:10
sort of prevent you from taking advantage to regain your freedom.
59:14
Speaker 2: Right.
59:15
Speaker 1: And I think we are now at that stage. We
59:17
don't yet have the therapies, but there are certainly like
59:20
investments and research and stuff we could do that would
59:22
hasten the arrival. And I think, in fact, we should
59:26
have started on a big sort of you know, Manhattan
59:30
product to defeat aging like decades ago. And maybe we
59:33
would have better therapies by now then. But for a
59:36
long time, aging was not seen as a problem. Only
59:39
the specific diseases that are the result of the aging.
59:42
Speaker 2: Yeah.
59:42
Speaker 1: And they got all this funding, right? Like, so you
59:44
have like huge research products for Alzheimer's and for heart
59:47
disease and for cancer. But if you look at the
59:50
main common cause of most human disease, it's senescence. Like
59:54
you're much more likely to get any of these. When
59:56
you're 80, then when you're 20, because your cells gradually
59:59
decay and gunk builds up, arteries clog up, you know,
1:00:04
your DNA mutates, your methylation pattern gets scrambled.
1:00:08
Speaker 2: Your brain tracks.
1:00:09
Speaker 1: So rather than just trying to put out the fires
1:00:11
one by one after they occur, like at some point
1:00:13
you just need to sort of try to prevent, like
1:00:17
maybe you need a bit of rain to sort of
1:00:19
prevent that.
1:00:20
Speaker 2: It's kind of amazing how much we've accomplished as a society,
1:00:24
as civilization, when you think about the fact that people
1:00:26
only live to be 100. Kind of incredible. Yeah. Kind
1:00:30
of incredible. Got rockets and satellites and cell phones and
1:00:35
people only live– so it's like everybody has to build
1:00:37
on everybody else's intelligence and everybody else's understanding of the
1:00:40
world and develop new things, and everybody has to learn
1:00:43
those new things.
1:00:44
Speaker 1: And each brain is so small that it can only
1:00:46
learn a small little specialty, right? And you need millions
1:00:49
of them. Although it's not entirely clear with humans, if
1:00:55
we had been much more long-lived, like the way we
1:00:58
currently are configured, maybe we would just have gotten too
1:01:01
stuck in our ways and sort of stagnated into some
1:01:04
sort of ultra-conservative society where nobody's allowed to do anything different.
1:01:08
or the old geezers that have their way of doing
1:01:10
things just remain forever in charge. So, I mean, we
1:01:17
just don't know, right? Like whether that would have slowed
1:01:21
progress or accelerated it. But it is, yeah, it is
1:01:27
still amazing. Like it's a long path from sort of
1:01:29
running around in the forest to sort of these, look
1:01:33
at these like advanced chip technologies and the whole global
1:01:37
supply chain Or like thousands of people are working to
1:01:41
develop just one little tool that then feeds into the
1:01:43
ability to make another tool that eventually makes these leading-edge
1:01:47
AI chips.
1:01:49
Speaker 2: Where they're layering things four atoms deep.
1:01:51
Speaker 1: Yeah, yeah, yeah. It's remarkable.
1:01:55
Speaker 2: It's incredible. And humans only live to be 100. Imagine
1:01:58
what we could accomplish if we lived to be 3,000.
1:02:00
Speaker 1: Or if we were just a little bit better at
1:02:05
these things. Yes. Because I think … We are sort
1:02:08
of the stupidest possible species that are capable of developing
1:02:12
advanced technology. Because as soon as we evolved to reach
1:02:15
that level, we started developing it, right? So that's where
1:02:18
we are, like the dumbest possible that just barely can
1:02:21
do this.
1:02:21
Speaker 2: Right, barely can do this while we're blowing each other up. Yeah.
1:02:23
At the same time, all over the world, blowing up people. Yeah,
1:02:29
but maybe a lot of that frantic, stupid, illogical behavior
1:02:35
is– because we're so finite. Like we're in a rush that,
1:02:40
you know, we realize we only have a certain amount
1:02:42
of time to get things done. And so this sort
1:02:43
of accentuates the desire to control resources and to cement
1:02:49
your immortality and to do these things that people love
1:02:51
to do, put their name on buildings. This is a
1:02:54
thing that people like to do that it's almost like
1:02:56
cementing their immortality. Maybe there'll be less of a desire
1:03:00
to do that if people live way longer. And then
1:03:03
you would have to assume If you can engineer humans
1:03:08
to live longer, you could probably engineer a lot of
1:03:10
stupidity out of us. They could just find out why
1:03:13
people– like what if you could just eliminate lying? What
1:03:17
if there's like a genetic solution to lying?
1:03:23
Speaker 1: Well, or just a really good lie detector.
1:03:26
Speaker 2: Yeah, but I mean what if you can genetically engineer
1:03:29
out the desire to lie?
1:03:32
Speaker 1: Well, I mean– If you had a perfect lie detector,
1:03:36
there would be no point in lying because people would
1:03:38
just see it.
1:03:39
Speaker 2: Immediately, right? Or if you could read each other's minds,
1:03:42
it would be fruitless. Yeah.
1:03:44
Speaker 1: And so that might be closer. I mean who knows
1:03:46
what you could do if you sort of had like
1:03:48
a powerful AI system trained to detect micro-expressions.
1:03:50
Speaker 2: We just all have to get the big gray heads
1:03:53
with the black eyes like the aliens have. Then you
1:03:57
can just read each other's minds.
1:03:58
Speaker 1: You want one of those?
1:03:59
Speaker 2: Well– One of the weirder things that Bob Lazar said, again,
1:04:03
I don't know if you're into this UFO stuff, but
1:04:05
he was one of the guys that was a whistleblower
1:04:08
that said he back-engineered crafts for the United States government
1:04:11
in the 1980s. They don't have any controls in these
1:04:15
crafts supposedly and that they're all powered by the minds
1:04:19
of the beings that are running it and that they
1:04:21
have– it's almost like these crafts are alive and they
1:04:23
have some sort of a syncing with the thing that
1:04:25
instead of like pressing buttons and working a joystick– they
1:04:29
just sync with this creation and then they can propel
1:04:34
it with their minds. So their minds become the computer
1:04:37
that moves this thing around.
1:04:40
Speaker 1: I'm a bit skeptical. I haven't looked into that.
1:04:43
Speaker 2: You should be. Sounds crazy. How could you not be?
1:04:46
Anybody who's not skeptical sounds crazy. But if you think
1:04:49
about where we are now and where we're going to be,
1:04:53
the possibility of that being our future is pretty likely.
1:04:56
Speaker 1: I think that the kind of possibility of engineering something
1:05:00
like that with sufficiently advanced technology would be there, like
1:05:04
all kinds of stuff. Basically everything that doesn't violate some
1:05:08
law of physics to a first approximation you would be.
1:05:11
Speaker 2: Able to construct. Even when we say that, it's like
1:05:14
our current understanding of physics. It's not the understanding of
1:05:17
physics of a society that's a million years older than ours.
1:05:21
Because that's a blink in the eye of the time
1:05:25
that it took to create what we are at right
1:05:28
now from the big bang like a million years is nothing.
1:05:32
Speaker 1: Um yeah but even with our current understanding of what
1:05:36
physics permits that's still an enormous space of uh designs
1:05:43
and types of life and being that you could imagine
1:05:46
instantiating you could have you know, like a Dyson sphere.
1:05:51
You've heard of this concept, right? It's basically like using
1:05:53
the output of the sun for energy generation. And it
1:05:56
could have like, you surround the sun by solar panels
1:05:59
that then powers a civilization infrastructure, maybe like a computer.
1:06:03
Like imagine the, like already we have like AI computes
1:06:08
growing at like 240% per year or something, right? And
1:06:13
then like imagine the kind of mind that could run
1:06:16
on that sized computer. so people are wondering whether like
1:06:23
could machines be conscious and like discussing that right I
1:06:28
think like maybe the more pertinent question is like are
1:06:30
we really conscious I think barely so you're driving on
1:06:37
like the road for two and a half hours a
1:06:40
motorist right and like driving past thousands of people and
1:06:44
homes and like mothers with their strollers and then After,
1:06:50
does he remember any of that? Was he even really
1:06:52
aware of any of this while he was driving? It's
1:06:55
like a little diffuse sense of body and some murky
1:06:58
perceptions floating through, maybe some confused abstract idea that we
1:07:03
don't really understand. That's kind of the consciousness that can
1:07:07
fit into this coconut-sized biological organ that we think, and
1:07:10
we think, wow, we are so conscious. But imagine this
1:07:13
Dyson sphere consciousness. Or like a mind that maybe spans
1:07:18
a galaxy. I think the difference between the sort of
1:07:22
awareness that it could have and our awareness might be
1:07:25
bigger than the difference between our awareness and whatever a
1:07:28
flea has or something like that. Way bigger. So it
1:07:33
could be this transition where we develop super-intelligent minds that
1:07:37
for the first time is really life waking up and
1:07:41
becoming truly aware. Yeah. and that we are a little
1:07:45
bit sort of over-proud in our own specialness when we
1:07:51
think that we have achieved something close to the maximum level,
1:07:55
that we are the standard by which consciousness should be measured,
1:07:59
and we are this kind of feeble, confused, murky glimmer
1:08:04
that is barely sentient at all. So that's, I think,
1:08:09
maybe the big... challenge for our era like giving birth
1:08:15
to super intelligence and then hopefully shaping and nurturing it
1:08:18
and steering it so that it becomes a positive thing
1:08:22
both for us ourselves and also for it itself and
1:08:27
also for whatever other if there are other super beings
1:08:30
somewhere in the world or in the cosmos that it
1:08:34
sort of is able to get along with us and
1:08:37
contribute positively at the cosmic scale And that's a very
1:08:43
multifaceted challenge, but I think that's kind of seems to
1:08:47
be what's going on.
1:08:49
Speaker 2: Yeah, and it's hard for people to think that far ahead.
1:08:52
You just think of intelligence being something superior to what
1:08:56
we currently experience. But when you're talking about a computer
1:09:01
or a being, a conscious being that is infinitely more
1:09:08
powerful than anything we can imagine, that seems to be
1:09:12
what– if everything keeps going in the same general direction
1:09:15
and AI increases its power– And we figure out new
1:09:19
ways to power it and then– because one of the
1:09:22
things that AI needs that's so interesting is it needs
1:09:25
enormous amounts of power. And so just these AI centers
1:09:30
that they're developing now, they're– like Google is doing one
1:09:33
where they're developing their own nuclear power. Do you want
1:09:35
some coffee? Are you still there? Yeah, right here. I
1:09:39
saw you reaching for your cup. You shouldn't be drinking
1:09:41
out of paper cups anyway, man.
1:09:42
Speaker 1: I don't usually do that.
1:09:44
Speaker 2: I know. I don't either, but occasionally I do. But
1:09:46
they're so bad for you.
1:09:47
Speaker 1: I'm usually quite conscientious with coming a little bit up.
1:09:51
Speaker 2: Even when I go on the road, I've started taking
1:09:53
a little Yeti coffee cup with me so I can
1:09:57
buy coffee and just have them put it in that. Okay.
1:10:02
Inside of those things is just lined with plastic. You're
1:10:04
pouring hot water into plastic and then the plastic leaches
1:10:07
into your coffee.
1:10:08
Speaker 1: I don't do that normally. But I figured like for
1:10:13
having a chat with you on the Jorgen show, it
1:10:15
would be worth like I guess some hundred microplastics.
1:10:19
Speaker 2: Well, thank you. I think what we're talking about is
1:10:25
inevitable if human beings don't blow ourselves up. If we
1:10:27
don't get hit with an asteroid, blow ourselves up or
1:10:30
a super volcano. It doesn't eradicate civilization. All this stuff
1:10:35
is– it's inevitable. It's just– it's how much time does
1:10:38
it take and how much does it grow exponentially in power?
1:10:42
Because we're talking about computers and then they start bringing
1:10:46
up quantum computing and quantum computers' ability to do calculations
1:10:50
and it doesn't even make sense. And so you think, well,
1:10:54
this is just one version of that. Like what is
1:10:58
quantum computing going to look like 500 years from now? Like,
1:11:01
what is computing power which is connected to AI? What
1:11:05
is that going to look like 500 years from now?
1:11:07
It's impossible to even guess.
1:11:10
Speaker 1: Well, we can sort of see lower bounds on what's possible,
1:11:13
like thinking already of just the designs we can conceive
1:11:17
of that we see in principle you could make. Maybe
1:11:20
it will take a long, a lot of grind to
1:11:22
get there, but at least that establishes like a lower
1:11:24
bound of what a technologically mature civilization could do, and
1:11:28
then maybe they have additional ideas beyond that, but But
1:11:31
already that is enough to really unlock. So if you
1:11:35
think of a space of possible modes of being, like
1:11:39
a mode of being is a way of experiencing, living, interacting.
1:11:42
I think you look around humanity and all people who
1:11:48
have existed in the past, a lot of different characters, right? Men, women,
1:11:53
like mean people, nasty people, crazy people. But I think
1:11:58
all of that diversity of human experience is like a
1:12:01
tiny little corner in this space of possible modes of being.
1:12:06
Like it's a huge cathedral and we've been kind of
1:12:08
basically sitting in the janitor's closet. That's like the exploration
1:12:13
we've done. And the kinds of modes of being, the
1:12:17
kinds of ways of experiencing and relating to each other
1:12:20
or thinking and doing stuff that are ultimately possible is
1:12:24
just this enormous space that we haven't that we haven't
1:12:28
been able to explore. Because ultimately what we can currently
1:12:31
conceive of and imagine and experience and feel is limited
1:12:34
to our biological substrate, the human brain. And just as
1:12:38
your early human ancestors from a few million years ago
1:12:44
that you were talking about before, they wouldn't really have
1:12:46
been able to conceive of the modern human condition. Not
1:12:50
just because I didn't think of it, but because the
1:12:53
ape brain can't really... Even if you try to describe
1:12:58
it to them somehow, like what is music, what is humor,
1:13:01
what is romantic love, what is science, what is all
1:13:07
of these things, what's literature, it just doesn't fit in.
1:13:11
Speaker 2: Right.
1:13:12
Speaker 1: And so presumably, I mean, you could think right now
1:13:14
we've achieved the right brain size where all possible values
1:13:18
and interesting ideas are accessible to us. That seems kind
1:13:21
of implausible. It seems much more likely that just as
1:13:24
the chimpanzees are necessarily blind to like a lot of
1:13:30
what can give life meaning and value and significance. So
1:13:33
probably are we too. Yeah.
1:13:35
Speaker 2: They don't even have the capacity for it, which is interesting.
1:13:38
Speaker 1: They can't even consider it. Yeah.
1:13:39
Speaker 2: We think of them as being so intelligent and they
1:13:41
are in comparison to a lot of other animals. But
1:13:44
I think one of the things that they were puzzled
1:13:45
by when they started teaching primates sign language was that
1:13:50
they never asked questions. Like they don't seem to have
1:13:54
questions about stuff. So they just, They just– they exist
1:13:59
in an intelligent way. They can figure things out. They
1:14:01
can do things. They know when the door is left unlocked.
1:14:06
They understand fairness, which is really weird. Like chimps get
1:14:10
very upset if they're treated unfairly. Like if one chimp
1:14:13
gets something, they don't get– they get super violent. But
1:14:17
they don't have a lot of questions. We're filled with questions.
1:14:21
So what are we missing here? That the next stage
1:14:25
of being a human will have that we're not even
1:14:28
considering now because it's not a part of our experience.
1:14:30
Speaker 1: Like what's the asking question type of thing that we
1:14:34
are not able to imagine? Yeah. And many of those
1:14:38
types of things. Yeah.
1:14:39
Speaker 2: And the possibilities are literally endless. We just don't conceive
1:14:43
of them because we're trapped in this territorial primate mindset. Yeah.
1:14:49
Speaker 1: Yeah. I think that's likely.
1:14:50
Speaker 2: Yeah, I think so too. I think so too. I
1:14:54
think the possibilities are endless. And I think it's going
1:14:58
to be really, really weird if that happens inside of
1:15:02
our lifetime. Because you and I, we're close to the
1:15:06
same age. So we grew up with answering machines. You
1:15:09
probably remember the first cell phones. You probably got online
1:15:12
and probably– when did you get online?
1:15:14
Speaker 1: Like the early– Ninety-six.
1:15:16
Speaker 2: Yeah, there you go. Yeah, see. I think for me
1:15:19
it was 94. So we got to see all this happen.
1:15:26
If we get to see the new version of humans
1:15:30
in our lifetime, that would be like literally bonkers. What
1:15:34
an amazing time to be alive if you really think
1:15:37
about it. We're so fortunate because if you grew up
1:15:39
between like 1700 and 1800, how much did shit change?
1:15:45
1600 to 1700, would you make a better boat? Like
1:15:47
what's different? Not a lot of different, man. Everybody kind
1:15:51
of is the same for 100 years. And inside of
1:15:54
our lifetime from the late 1960s for me to the
1:15:59
early 70s for you, like in the amount of time
1:16:01
that we've been alive, things have radically changed. Like really,
1:16:07
really radically to the point where it's probably the biggest
1:16:10
shift in human ingenuity and innovation that the world has
1:16:15
ever seen. And we're just in the middle of it.
1:16:18
And we might be in the middle of the next one,
1:16:20
which literally allows us to see what the world looks
1:16:26
like a thousand years from now because you're going to
1:16:27
be alive.
1:16:28
Speaker 1: I mean that's why it is now this full-time job
1:16:32
just to monitor the situation.
1:16:34
Speaker 2: How do you do that?
1:16:35
Speaker 1: Well, you don't really. But you try to sort of–
1:16:41
so I think the opportunity– so it used to be–
1:16:46
like at least in my sphere of effect, if you're
1:16:47
doing philosophy or something like it, most people would think
1:16:53
you have a kind of unlimited time horizon. People have
1:16:55
been working on philosophical questions for thousands of years, and
1:17:00
there doesn't seem to be any huge urgency. If they
1:17:04
have been unsold for thousands of years, maybe if it
1:17:07
takes another 500 or so. But I always thought of
1:17:10
philosophy as having a deadline, meaning that at some point,
1:17:16
we would develop smarter than human forms of intelligence, presumably AI,
1:17:21
that could then do the philosophy much better than us.
1:17:23
And so there was a limited period of time during
1:17:26
which any advances I could contribute to would be meaningful
1:17:32
and that it would then make sense on focusing on
1:17:34
that subset of philosophical questions or general questions that we
1:17:38
really need the answer to now as opposed to like,
1:17:41
you know, 10, 20 years in the future when somebody
1:17:45
else can do them better, like the machine minds. So
1:17:47
that's kind of been a lens through which I have
1:17:50
selected the things to work on. And now, of course,
1:17:52
that deadline is moving closer. So there's less time remaining.
1:17:57
And so my focus is increasingly drifting towards questions where
1:18:03
it might be relevant to have the answer to now
1:18:05
rather than a year or two from now.
1:18:07
Speaker 2: You're almost like a cultural navigator, like a guy with
1:18:10
a sextant at the the helm of a ship, looking
1:18:14
at the constellations and going, I think we're in the
1:18:17
right direction.
1:18:18
Speaker 1: Yeah, but now we're sort of moving maybe close to
1:18:22
harbor and you need to pay more attention to exactly
1:18:27
how deep the rocks are and scan around you.
1:18:32
Speaker 2: When you think about a timeline for radical change, in
1:18:38
your mind, what do you think? What do you think
1:18:41
that looks like?
1:18:42
Speaker 1: Well, I mean, I take short timelines seriously with AI.
1:18:46
I mean, for what we know, it could be like...
1:18:50
like a year or two or three or four and
1:18:51
probably a bit longer. But we're no longer at a
1:18:56
point where we can be confident that we won't have
1:18:59
super intelligence in just a few years. Like it could happen.
1:19:03
Speaker 2: What is that? When you say that for the uninitiated,
1:19:06
what do you mean by super intelligence?
1:19:08
Speaker 1: Well... I guess we first have AGI, artificial general intelligence,
1:19:17
AI that can do all the stuff that we can do.
1:19:21
And then super intelligence would just be that, but can
1:19:24
do it way better than any human can do. So
1:19:28
all technical, intellectual, common sense tasks, and then they would
1:19:33
have robotics as well that can do all the physical stuff,
1:19:36
not much later. And so this, yeah, so the timeline
1:19:42
remains uncertain, but I think it's not impossible that this
1:19:48
could happen very soon. And then once you have super intelligence,
1:19:51
then I think from there on, it might be like
1:19:54
a sort of sprint to something approximating technological maturity, because
1:19:58
what you have super intelligence that then designs even more
1:20:00
smart AIs, right? Using its kind of super intelligent AI
1:20:05
research capability and designing better chips and.
1:20:07
Speaker 2: All of that.
1:20:09
Speaker 1: So you might then have this like intelligence explosion where
1:20:13
you go from something slightly greater than human level to
1:20:18
some radical super intelligence that can then sort of invent
1:20:21
whatever the remaining technologies are. Maybe there needs to be
1:20:23
some trial and error and experiments in the physical world
1:20:25
that slows things down a bit. But some smallish number,
1:20:28
like a single digit number of years from super intelligence,
1:20:31
I think you might have something that unlocks all of
1:20:33
these like sci-fi level capabilities that we've talked about. I
1:20:40
think that seems relatively plausible to me.
1:20:44
Speaker 2: It seems inevitable. And my question is how does it
1:20:51
announce itself? Does it send a mass text message to
1:20:54
the whole world? Everybody's phone just starts like– you know
1:20:56
when you have those amber alerts and your phone starts
1:20:59
vibrating or when there's some sort of a storm warning?
1:21:03
All of a sudden your phone goes off and then
1:21:05
it alerts you to the fact that it's taken over.
1:21:08
Speaker 1: Well, so here we don't really know. It's like this
1:21:11
is very confusing, and we've never done this before, and
1:21:14
so it's very hard to figure out how this is
1:21:17
going to unfold. And maybe it's not even fixed yet.
1:21:20
Maybe it depends a little bit on what we do
1:21:23
and the extent to which different actors get their stuff together.
1:21:27
Speaker 2: But, like, one, I guess.
1:21:35
Speaker 1: One possible type of scenario is where things are just accelerating.
1:21:38
There are more and more of these advances, model releases. Increasingly,
1:21:43
there is automation of the research process in AI labs. Already,
1:21:49
you have coding assistants that are really useful to people
1:21:53
in AI labs. They're using them to write a lot
1:21:55
of their software. Right now, you still need humans for
1:21:58
the research taste and judgment. Sometimes, things get stuck and
1:22:03
you need a human to kind of redirect them but
1:22:06
more and more of that may get automated and then
1:22:12
you have kind of new iterations of models being trained
1:22:18
at a faster and faster clip they can do more
1:22:20
and more stuff they start to automate increasing chunks of
1:22:29
the economy so right now a lot of coding is
1:22:31
automated but like other areas as well. Maybe they, they
1:22:35
become drop-in virtual workers that, um, can do everything like
1:22:39
a human could do with a remote connection, uh, initially.
1:22:44
And then like you have people working on robotics, so
1:22:46
then that would start to kicking. And then eventually just
1:22:49
more and more of the action is, is run by
1:22:51
these AI systems, one or more. Um, and, and they're
1:22:55
kind of doing it at their timescale, which is speeding up. Um,
1:23:01
Then from that point, it would depend a lot on
1:23:04
whether we have successfully aligned these systems so that they
1:23:09
actually do what the people created them intended or whether
1:23:13
they have somehow gone off the rails.
1:23:17
Speaker 2: There's also the fear that America doesn't come up with
1:23:21
it first.
1:23:23
Speaker 1: This is our fear over here. Yeah, that's mostly a
1:23:25
fear in America. In China, for example. Yeah, America comes
1:23:30
up with it first, yeah.
1:23:31
Speaker 2: I mean, it's really just America and China that are
1:23:33
at the forefront of this race, right? Yeah. Russia a
1:23:36
little bit, but not on the same level.
1:23:38
Speaker 1: Not Russia. No?
1:23:39
Speaker 2: Not Russia? No. Who else is it?
1:23:41
Speaker 1: Did you see this last week?
1:23:43
Speaker 2: Ford hiring 350 engineers after AI failed shows human value
1:23:49
in AI era. What does that mean? I tried to
1:23:52
find the only one that didn't.
1:23:52
Speaker 1: I just did it.
1:23:54
Speaker 2: They rehired a bunch of engineers they fired after their
1:23:57
AI– wasn't matching the quality they needed. Oh, and this
1:24:01
is Ford. That's interesting. They probably jumped the gun, you know?
1:24:07
They probably fired them too quickly. And now these people
1:24:10
that they rehire, these people are going to like, oh,
1:24:12
how much time we have left? Six months for AI
1:24:15
figures out how to do this just as well?
1:24:17
Speaker 1: I mean, it's also possible that things, timelines could extend
1:24:23
if like... One way that could happen is the progress
1:24:28
we've seen in the last 10 years, which has been remarkable.
1:24:33
It's to a substantial extent driven by the advances, the
1:24:38
increase in compute that is being used to train and
1:24:41
run these systems. It used to be that you had
1:24:45
a cutting-edge AI system. If you were an academic running
1:24:48
on your PC in your office, that was the amount
1:24:51
of computing power that was applied to doing AI stuff.
1:24:55
And now, of course, you have these kind of tens
1:24:57
of billions of dollar data centers, like hundreds of megawatts.
1:25:02
Speaker 2: Right.
1:25:03
Speaker 1: Like just massive funding. So the chips have gotten better,
1:25:05
but also just the amount of funding, like you're just
1:25:07
building many more of these chips. And so as you
1:25:10
apply more and more compute, like performance improves. And that's
1:25:14
like has been a big driver. Now, at some point,
1:25:18
like you might not be able to keep increasing the
1:25:21
amount of money you spend on it because like, You
1:25:25
can go from like a thousand dollar PC to like
1:25:28
a million dollars quite easily. And you can go from
1:25:30
a million dollars to a billion dollars. And maybe you're
1:25:34
spending on the order of a trillion dollars across the
1:25:37
world to build data centers per year. But you can't
1:25:40
really do like three orders of magnitude very easily there.
1:25:43
There's like not a thousand trillion dollars to spend on it.
1:25:46
So at some point that expenditure has to kind of
1:25:49
slow down. So if we haven't achieved super intelligence by then,
1:25:52
then maybe that would mean progress get slower if the
1:25:56
main driver is the scale up of compute. Now, it
1:25:58
is also true that some of the progress is driven
1:26:00
by algorithmic advances, like just kind of clever algorithms. So
1:26:05
that might continue. But if one driver stalls out, then
1:26:09
that could result in faster progress. And then, of course,
1:26:13
there is a possibility that the people who want to
1:26:18
pause or regulate AI gain enough traction to kind of
1:26:21
get regulatory inhibitions.
1:26:24
Speaker 2: How would they do that though if there really is
1:26:26
some sort of a Manhattan Project style race between the
1:26:29
United States and China? And what other countries are developing
1:26:32
AI right now that are close?
1:26:33
Speaker 1: Well, I mean those are the– Those are the two
1:26:35
big ones? The big ones.
1:26:37
Speaker 2: Is it possible that someone could sneak up on us
1:26:39
and develop super intelligent AI first?
1:26:42
Speaker 1: Yeah. I mean it's possible if there were like some
1:26:44
big algorithmic breakthrough for example that made it. a lot
1:26:48
more efficient to run a similar level of capability with
1:26:52
less sort of AI data center infrastructure. And many other
1:26:56
countries are also trying. It's just that they are not
1:26:59
as advanced. And out of U.S. and China, I'd say, like, U.S.
1:27:03
currently has the edge.
1:27:06
Speaker 2: What would happen if China got there first?
1:27:08
Speaker 1: Well, I mean, part would depend on whether they had
1:27:12
successfully aligned their AI. If they... unaligned, then I guess
1:27:17
the same thing happens as if US AI is unaligned.
1:27:20
That is, the future gets shaped according to whatever values
1:27:24
this AI had ended up with. If the alignment problem
1:27:28
is solved, then it might make a difference because then
1:27:31
the values would then depend on sort of the people
1:27:36
who owned and controlled or governed it would ask it
1:27:40
to pursue. So then in that scenario, maybe it makes
1:27:44
some difference. who initiates it.
1:27:50
Speaker 2: So right now the big players are– there's Google and
1:27:54
there's OpenAI. There's all these different companies. When you say aligned,
1:27:59
do they have to all be kind of on the
1:28:01
same page or like when aligned, we're not going to
1:28:04
be aligned if there's a bunch of corporations that are
1:28:06
competing to come up with this first. So they have
1:28:09
to be aligned in terms of the way they're programmed,
1:28:12
that they're valuing human life and that they're valuing society
1:28:17
like what do you mean by.
1:28:18
Speaker 1: Well so i mean just the technical challenge of if
1:28:21
you are building an ai system and you have certain
1:28:24
things you wanted to do and certain things you don't
1:28:26
want it to do right are you technically able to
1:28:28
get the ai to behave that way that you.
1:28:31
Speaker 2: As the person who gets to decide well that's.
1:28:33
Speaker 1: A separate question which is equally important maybe but different
1:28:37
because that's not the technical like you can't just go
1:28:40
to the whiteboard and write down some formula now you
1:28:42
have that's like a political question ultimately right question of
1:28:45
governance yes um so there you need you know political
1:28:49
organization appeals to the best in human nature dialogues like
1:28:55
checks and balances whatever stuff that might work in the
1:28:58
political arena um To hope that the governance of this,
1:29:02
that the values to which it is aligned are sort
1:29:04
of benign values that hopefully incorporates a wide set of stakeholders.
1:29:10
Speaker 2: Right. But that's a little– isn't that a little naive?
1:29:12
Because whenever there's any sort of a situation where something
1:29:16
has massive amounts of power above others, one of the
1:29:19
first things I think of is what's the most money
1:29:21
you can generate doing it? And what is the best
1:29:25
way to generate the most amount of money? Like that's–
1:29:28
They're going to think that way. They're not going to
1:29:29
think like what's the best for the human race. Like
1:29:31
no one ever thinks that way. They think what is
1:29:33
the best way in terms of like destroying the competition?
1:29:36
What is the best way in terms of extracting the
1:29:39
most resources out of this? Making it work for me.
1:29:43
That's what people think of immediately. And what's the best
1:29:46
way to stop other people from competing with us?
1:29:49
Speaker 1: Right. So that would be I think a lot of
1:29:55
pressures and strains on whatever– governance mechanism exists at the
1:30:00
time when superintelligence is developed.
1:30:05
Speaker 2: So is this something that the technology is so far
1:30:09
ahead– the potential for it to be so far ahead
1:30:12
of our understanding of what it's going to be able
1:30:14
to do that like making laws for that now, it's
1:30:16
going to be very difficult to even explain to people?
1:30:20
Like we need these laws and we need these guardrails
1:30:24
in place now because here's what could happen. And that
1:30:28
conversation is not really happening right now.
1:30:30
Speaker 1: It is happening.
1:30:32
Speaker 2: But is it happening politically?
1:30:34
Speaker 1: It's starting to happen, yeah. It's been happening to an
1:30:38
increasing extent politically and there have been various actions.
1:30:41
Speaker 2: Like what actions?
1:30:43
Speaker 1: Well, so a few years ago, for example, that was
1:30:45
the whole export– regime imposed by the U.S. on chips,
1:30:51
the most advanced chips, where like China was cut out
1:30:54
from being able to access the most advanced Nvidia chips
1:30:58
and so forth. And that's kind of been modified, but
1:31:03
that was motivated to a significant extent by trying to
1:31:06
preserve AI edge that the U.S. has.
1:31:10
Speaker 2: Right. But that's like internationally. I mean, in terms of
1:31:12
like being able to stop corporations from having the kind
1:31:18
of power that this could provide them.
1:31:20
Speaker 1: So then there have been various efforts that was like
1:31:23
a proposition in California where there were various things. More recently,
1:31:28
you saw the whole thing. I don't know if you
1:31:30
followed it with the Mythos and Fable 5, the anthropic
1:31:36
models that were... So Mythos has not been released. This
1:31:41
is the most powerful model because it seems to have...
1:31:45
significant cyber offense capabilities. It can easily detect vulnerabilities in software.
1:31:52
And so Anthropic figured that rather than immediately making it
1:31:55
available to everybody, maybe it would be better to first
1:31:58
try to make it available to providers of critical software infrastructure,
1:32:06
like big banks, so that they can patch up their systems.
1:32:11
And then Fable 5 was a kind of restricted version
1:32:17
of the Mythos model, like the same underlying model, but
1:32:21
with extra safeguards. It basically refused anything that remotely seemed
1:32:25
like cyber hacking, programming, biological stuff, because maybe that would
1:32:33
be bio. So it sort of drew a wide circle
1:32:35
around anything that even remotely looked possibly dangerous, and it
1:32:39
just refuses that. So that was released. But then after
1:32:44
like a week or so, the administration imposed a kind
1:32:51
of export restriction that prevented any non-U.S. citizen from using it.
1:32:57
And that meant Anthropic had to cut it off for
1:32:59
everybody because they didn't have a way in real time
1:33:02
to verify who is a U.S. citizen or not. Oh, wow.
1:33:06
So then it was like unavailable for several weeks. and
1:33:10
intense negotiations behind the scenes and working to try to
1:33:13
figure out, because allegedly it was possible to jailbreak it
1:33:16
so that it sometimes gave some little assistance with some cyber,
1:33:20
like finding vulnerabilities in code. And now recently it just
1:33:25
became available again because they had reached some understanding with
1:33:29
the government where it was deemed sufficiently safeguarded to be released.
1:33:35
And now there is like, efforts underway to try to
1:33:38
work out the framework because like in the future, you
1:33:40
wouldn't want to do this on an ad hoc basis.
1:33:42
Like somebody just decides this particular model for some unspecified reason.
1:33:46
Like you want to have clear standards ideally, right? That
1:33:48
applies to everybody, every company. So there's now some industry-wide
1:33:52
effort to try to work with the government too. Anyway,
1:33:55
so there's like a lot of this stuff happening. I
1:33:57
expect much more of this going forward. It just has
1:34:01
recently become like a serious issue. People until recently were
1:34:06
kind of ignoring the whole AI thing for the most part.
1:34:10
And then there is a second wave coming, I think.
1:34:14
So far, we haven't really seen any big impact on
1:34:16
the labor market from AI. But once that starts to
1:34:21
hit and you get maybe high levels of unemployment amongst
1:34:26
white-collar workers. Imagine if you have millions and millions of
1:34:31
people that have their university diplomas. They feel a sense
1:34:35
of superiority and entitlement. They've gone through the whole process.
1:34:38
They got their degree, and now they expect a well-paying job.
1:34:41
And then there is no job for them. They've got
1:34:45
nothing to do all day long but complain about AI.
1:34:50
So you've got to have all these well-educated people who
1:34:52
feel resentful and are going to say every possible bad
1:34:55
thing about AI that could be said all day long
1:34:57
and mobilize. There's a very powerful political constituency that will
1:35:01
emerge from that. That's not even yet happened, but that
1:35:04
will kind of add to all of these other grievances
1:35:09
that people point to with AI. So I think there's
1:35:12
going to be kind of significant political pressures for doing
1:35:17
something about AI.
1:35:19
Speaker 2: The key is getting ahead of it though. So how
1:35:23
can people find– how can we see the vulnerabilities in
1:35:28
advance of AI? and recognize like when this is going
1:35:31
to– like if there is going to be a tipping
1:35:33
point and a bunch of white-collar workers are going to
1:35:35
be out of work and there's going to get to
1:35:37
a point where we realize like this is coming. This
1:35:39
is like three weeks from now. What do we do?
1:35:41
Like what do they do?
1:35:42
Speaker 1: Yeah, so that's not clear to me at least. I
1:35:45
mean a lot of people think they have different views
1:35:48
about what should be done, of course. But it's like
1:35:52
very hard to have an– it's such a complex strategy.
1:35:56
I mean I've been thinking about this for like– three
1:35:58
decades maybe and I still feel extremely unsure even which
1:36:03
direction is kind of up and which is down. I
1:36:07
don't think it would be possible right now to sort
1:36:09
of figure out a detailed perfect regulatory scheme or system
1:36:14
of laws because there's so many unknowns. I think we
1:36:17
need to sort of watch this closely as it happens
1:36:21
and be ready to react quickly to issues that come
1:36:24
up and hope that relevant people are like highly competent
1:36:30
and well-motivated and are trying their best. And then on
1:36:34
the margin, maybe we can do things that kind of
1:36:36
are constructive and increases the chances that it will be
1:36:40
for the good. If and when I have some clear
1:36:45
insight as to some overall big directional push, I will
1:36:50
let you know. But even basic questions, like for example,
1:36:57
do we want more government involvement or less government involvement?
1:37:01
Do we want faster progress in AI or slower progress
1:37:05
in AI?
1:37:09
Speaker 2: I don't know.
1:37:09
Speaker 1: It's not completely obvious to me.
1:37:15
Speaker 2: Well, I think it's eventually going to get to wherever
1:37:17
it's going to get either way. Having it slower, I
1:37:21
don't know if that's really going to help us. I
1:37:24
think almost like we have to crash And then we
1:37:26
have to figure out how to rebuild and pick up
1:37:29
the pieces. I don't think we're going to be intelligent
1:37:31
or have enough foresight to recognize where all these flaws
1:37:36
and where all these problems are going to ultimately be.
1:37:40
I think they're just going to have to happen. And
1:37:42
then people are going to have to adjust. That's what
1:37:44
I think. Yeah.
1:37:46
Speaker 1: So I'm not advocating an AI pause by contrast to
1:37:49
a lot of some of my colleagues and friends and stuff.
1:37:52
Speaker 2: Yeah.
1:37:54
Speaker 1: I could see some scenario in which it would be
1:37:57
helpful at some point to have a temporary slowdown of
1:38:02
a few months maybe or a half a year or
1:38:04
a year. Like if you imagine, you know, there are
1:38:06
different companies, countries maybe racing to get there first.
1:38:09
Speaker 2: Yeah.
1:38:10
Speaker 1: And then eventually somebody figures out they basically have the
1:38:13
system in place. They just need to like amp up,
1:38:16
like run it for longer. They can sort of see
1:38:18
that it will become super intelligent. They hope it's aligned. Yeah.
1:38:22
It might be very helpful in that situation to have
1:38:25
a few extra months just to sort of double check
1:38:28
all your safeguards. And rather than immediately cranking all the
1:38:31
knobs up to 11, like maybe sort of do it
1:38:34
a little bit incrementally, watch what happens, study it, then
1:38:37
crank it up a little bit more. I just think
1:38:40
you might gain some extra safety if you have a
1:38:41
few extra months there, like the pressures on these people
1:38:44
in the lab.
1:38:46
Speaker 2: That makes sense.
1:38:46
Speaker 1: If it's all going to happen over a week, like,
1:38:48
it's just going to be, this is the first time
1:38:49
we ever do this hugely complex thing. It has to
1:38:51
be right. We never get a second chance. Just being
1:38:54
able to, like, even just being able to sleep properly for, like,
1:38:57
between your work sessions and, like, having, you know, a
1:39:01
weekend to mull things over. Like, just that kind of human,
1:39:05
humans don't really operate on, well, in these, like, super
1:39:10
short timelines. So I think a short process If you
1:39:14
could be sure that the pause would in fact be
1:39:16
limited and short and then it would be lifted, I
1:39:19
think that could be quite useful to have if it
1:39:22
could also be timed to be at the right moment
1:39:25
and well implemented.
1:39:26
Speaker 2: But wouldn't the problem be espionage first of all?
1:39:29
Speaker 1: That's one issue with it. Yeah.
1:39:31
Speaker 2: If they realized– if somebody realized, hey, they're about three
1:39:34
days away from this and then China bribes a bunch
1:39:37
of people and people take off and move to.
1:39:42
Speaker 1: Yeah, there are various downsides to pulsing. So one is
1:39:48
that there are competitors who just even without espionage are
1:39:52
catching up. Another is it's not as if the world
1:39:57
is safe without AI.
1:40:01
Speaker 2: Right.
1:40:02
Speaker 1: There are other, like, the natural hazards, I think they
1:40:06
are very small, like supervolcanoes and stuff like that on
1:40:08
the relevant timescales.
1:40:09
Speaker 2: Until they're not.
1:40:10
Speaker 1: Well, I mean, we've survived for hundreds of thousands of years, right?
1:40:14
Speaker 2: We're almost wiped out entirely 70,000 years ago.
1:40:16
Speaker 1: Yeah, well, that's a long time ago.
1:40:18
Speaker 2: Is it, though?
1:40:20
Speaker 1: Not if it happens tomorrow, it's not. Well, if something
1:40:23
hasn't done us in in that long a time, probably
1:40:26
it's not going to do us in within the next
1:40:28
10 years.
1:40:28
Speaker 2: Boy, you're more optimistic than me.
1:40:30
Speaker 1: Well, it's just based on the actual empirics of these
1:40:33
natural catastrophes. By contrast, I would say there are other
1:40:36
hazards that are not low. I mean, we have rapid
1:40:41
advances in synthetic biology, some of them driven or enabled
1:40:45
maybe even by already AI progress that has already taken place,
1:40:52
like that can start to assist with this, and more
1:40:54
just happening independently. So you get increasing risks from synthetic biology.
1:40:59
We still have the nuclear arsenals, right, kind of existing. We,
1:41:05
I think, have gotten a little bit complacent about the
1:41:07
risks of nuclear war. and various other things as well.
1:41:15
So I think there's like a background level of existential
1:41:18
risk that humanity faces in the absence of super intolerance
1:41:21
that probably is growing as well. So you don't want
1:41:24
to wait so long that you don't even get the
1:41:26
chance to roll the dice with AI because you destroyed
1:41:29
yourself before you even got to that point. It would
1:41:32
kind of be sad. And then.
1:41:37
Then also, I mean, I think there is like some
1:41:39
risk if you set up a pause infrastructure that what's
1:41:43
initially meant to be a temporary thing becomes permanent. Like
1:41:47
they say there's nothing more permanent than a temporary government program.
1:41:51
So if you set up the infrastructure to actually control this,
1:41:53
ideally at a global scale, right, it's meant to be
1:41:55
for six months. Like what happens after six months? Nobody
1:41:59
can still prove that the air will be safe. So
1:42:02
you have all these people now whose job it is
1:42:04
to regulate it and like maybe they– So it could
1:42:08
become kind of frozen in and become permanent. And then
1:42:12
also the delay benefits of this. So if you think, what,
1:42:20
like 65 million people die every year, that's a lot
1:42:26
of human lives. That's like one 9-11 every 25 minutes.
1:42:33
Just kind of boom.
1:42:35
Speaker 2: That's a way to look at it.
1:42:36
Speaker 1: Wow. And so there is a certain urgency. If there
1:42:39
is something that somehow could have a chance to fix this,
1:42:45
like all the suffering that is happening in the world,
1:42:47
like aside from the people dying, like all the people
1:42:50
who are bereaved, right, who lost their loved one, and
1:42:53
then all the disease that led up to that dying, right,
1:42:57
and then all the non-disease-related, all the horrible poverty and, like,
1:43:02
the suicidal depression and the animals in the animal factories
1:43:06
that are, like, spending it. There's just like this, I think,
1:43:09
moral urgency that if there is a hope that getting
1:43:13
AI right could fix this, then you don't want to
1:43:15
wait unnecessarily long because every day is just this massive horror.
1:43:21
So I think there are many reasons for why you
1:43:23
wouldn't want unnecessary or excessive delays in developing AI. But
1:43:28
there is a trade-off because if we can make more
1:43:31
progress on ai safety and alignment and get our act
1:43:33
together a little bit before we take the plunge like
1:43:35
that's also worth quite a lot.
1:43:40
Speaker 2: It's uh it's certainly exciting because the possibilities are right
1:43:46
in front of us and they're kind of endless and
1:43:49
it seems like they're right in front of us it
1:43:50
seems like to in my mind it seems like we're
1:43:53
like 24 months away from something really insane i.
1:43:56
Speaker 1: Think it could be could be yeah it could be
1:43:58
48 months it could be i mean we don't we
1:44:00
don't Yeah, it could be.
1:44:02
Speaker 2: Yeah, who knows? But there's something happening really quickly. You know,
1:44:05
when I talked to Elon about Grok, he said it
1:44:09
shocks us like every couple weeks. We're like, how is
1:44:13
it doing this? How is it so fast? How is
1:44:15
it advancing so quickly? And one day, cat's going to
1:44:21
be out of the bag and there's not going to
1:44:22
be any way to stop it. Like if the power
1:44:24
went out right now, if there was some sort of
1:44:26
a massive solar flare that killed our power grids and
1:44:31
all computer hard drives got fried and we had to
1:44:34
restart from now, like we would have to rebuild society, right?
1:44:40
But if we get to the point where this thing
1:44:43
understands what would cause that, how to prevent that, how
1:44:46
to make sure that never happens, much better power supply,
1:44:50
much better allocation of resources, much better batteries, much better
1:44:56
redundant data systems where you never have to worry about
1:45:01
hard drives crashing and you never have to worry about
1:45:04
any of these problems. Like any information you have now
1:45:06
will be secure. And then understanding of natural disasters will
1:45:10
actually get drill into this volcano and it never goes off.
1:45:14
Like there's going to be a few things that they're
1:45:15
going to figure out through AI. That's going to prevent
1:45:18
a lot of the things that are probably wiped out
1:45:20
enormous swaths of people. I mean, imagine if we could
1:45:22
just actually see asteroids coming, all of them and know
1:45:25
how to divert them instantaneously. We have a bunch of
1:45:30
ideas right now on how to do it, right? Like
1:45:32
coding them.
1:45:32
Speaker 1: Yeah, yeah, yeah.
1:45:34
Speaker 2: With something that changes the aerodynamics of them.
1:45:37
Speaker 1: I think that one would be relatively easy. Yeah. Even
1:45:39
we could figure out to do that, I think.
1:45:42
Speaker 2: And there's also detonating them. There's a bunch of different
1:45:44
crazy ideas. But imagine if AI is like, actually, you
1:45:48
could just do this. And then you put a shield
1:45:51
over the earth and you never have to worry about
1:45:52
it ever. The shield's powered by the sun. So you
1:45:55
have ultimate power. And you never have to worry about
1:45:58
being hit by a Manhattan-sized asteroid.
1:46:01
Speaker 1: I think the trickier ones might be ones that are
1:46:03
more internal to civilization. If you have some process that's
1:46:11
either like a worldwide process with different humans and corporations
1:46:16
and governments or like an internal process in the AI.
1:46:19
On the one hand, you want to be able to
1:46:21
continue to develop new ideas, new ways of doing things.
1:46:26
So you need to experiment and try new things. But
1:46:28
then there's also the risk of just keep trying new
1:46:32
ideas like that eventually you get stuck on some idea
1:46:35
that actually proves really harmful or dangerous to you. And
1:46:40
that you then sort of derail internally through your own
1:46:44
memetic or internal evolutionary development. So how to sort of
1:46:50
grow up safely in this world where you have unlocked
1:46:55
all kinds of new technological capabilities even if you're easily
1:46:59
able to protect yourself against external threats like volcanoes and
1:47:02
asteroids and stuff right there might still be processes arising
1:47:07
from within this global civilization just as we might invent
1:47:12
new technologies that are dangerous we might invent new like
1:47:15
drugs that are super addictive once you invent them you
1:47:18
try them and then you're addicted you never want to
1:47:21
yeah there could be other like some some crazy ideology
1:47:24
that once it takes.
1:47:25
Speaker 2: Hold you I'm glad you brought that up because that's
1:47:29
one of the major fears that a lot of people
1:47:31
have is that human beings will develop ideologies based around AI,
1:47:36
that there will be like an AI god that someone develops.
1:47:41
Speaker 1: Yeah, and I mean I think– in fact, I would
1:47:43
add that to the list of– I mentioned nuclear and
1:47:46
bioterrorist risk as like background existential risks. Like another is
1:47:53
like some form of insanity, like collective insanity I feel.
1:47:57
Our civilizational sanity already is kind of a little bit precarious.
1:48:04
I think we're just kind of maybe barely holding ourselves together.
1:48:08
And it's amazing how well the world functions despite how
1:48:12
crazy people are and how much they loathe their opponents
1:48:15
and so forth. And we still managed to somehow get
1:48:17
it to work.
1:48:18
Speaker 2: There's also manipulation of the zeitgeist that's clearly being done
1:48:22
by bots. So you've got a bunch of people online
1:48:26
that are having arguments on Twitter and they're not even
1:48:29
talking to people. There's a bunch of ideas that are
1:48:31
being pushed out on Twitter and a lot of these
1:48:34
social media platforms. It's not even human beings tweeting about it.
1:48:38
It's not. There's algorithms. There's AI. There's people that are
1:48:42
hired to do it. There's people that are working for
1:48:45
certain organizations that are hired to muddy up the waters,
1:48:49
gaslight people, create problems, have people argue with each other.
1:48:53
And so it's like if AI recognizes how easy it
1:48:58
is to manipulate people, do whatever it wants, that's a
1:49:01
fear as well. Yeah.
1:49:03
Speaker 1: So whenever you change the basic parameters of sort of
1:49:07
the– social, cultural, political discourse, new dynamics will emerge. And
1:49:18
we don't have the kind of social science that is
1:49:21
able to predict what happens if you change some of
1:49:23
these knobs. So we've seen in the past, like you invent,
1:49:26
let's go all the way back. Like somebody invents writing. Okay.
1:49:30
So that turns out to have had a huge effect.
1:49:33
Not just on people writing, like, literature and stuff, but, like,
1:49:36
on political systems. You could now have states that could
1:49:39
keep tax records, right? So you could have larger political
1:49:42
units with writing, and then they can hire standing armies,
1:49:47
and now you have, like, large-scale war, you have social stratification,
1:49:51
you could have, like, the ruler of a large area
1:49:53
could have, you know, enormous wealth, and you could have
1:49:56
a soldier caste that, like, protects against internal and external. So, like,
1:50:00
just the way that human societies are politically organized changes
1:50:04
as a result of this change in the rules of
1:50:08
communication when you can have written texts, right? And then
1:50:10
you have the printing press. Again, we're like 100 years
1:50:14
of religious wars in Europe, like reformation and all of
1:50:17
this stuff is possibly. And then like forms of democracy
1:50:21
later on also coming out of this and the scientific revolution.
1:50:27
Then you have like mass media in the last century
1:50:29
with like radio and stuff and you have kind of
1:50:31
demagogues that take advantage of being able to simultaneously talk
1:50:34
to millions of people, because that was never possible before, right?
1:50:38
You could have some charismatic kind of guy who's like
1:50:41
rallies up the whole nation and new ideologies become memetically
1:50:46
fit in that situation that might never have, if it
1:50:48
were people writing kind of letters to the editor, like
1:50:53
it's kind of a different type of idea that works
1:50:55
there than if you're giving a kind of stump speech
1:50:57
that goes out to millions simultaneously. And now With social media,
1:51:00
of course, we have another one of these, and you
1:51:02
do see that starting to change culture in different ways.
1:51:06
And now with AI being the next wave of this,
1:51:09
that will also change. You have bots, you have new
1:51:12
ways of finding information, you have maybe AIs that can
1:51:15
themselves be super persuaders. That will also change, presumably, in
1:51:19
some unknown, unknowable way, the way that social discourse pans
1:51:24
out and For any one of these, I guess it
1:51:26
could turn out to be a lot better. We could
1:51:28
become more informed having AI advisors. I think that's a
1:51:31
fairly likely scenario. But it's also possible somehow the dynamics
1:51:34
shape out in a different way and we kind of
1:51:36
go collectively insane in some way.
1:51:39
Speaker 2: Back to the whitewater raft.
1:51:42
Speaker 1: White to backwater raft, yes. And it all becomes kind
1:51:45
of totalitarian or we sort of fragment into like political
1:51:49
warring tribes or whatever. We become kind of completely unhinged.
1:51:53
Every one of us becomes convinced of their own little
1:51:55
nutty theory that they then, like their AI or feed,
1:51:59
just serves them more material to kind of fuel their conviction.
1:52:02
There are different ways in which this could go badly.
1:52:06
But don't you.
1:52:07
Speaker 2: Think even if that happens, ultimately the progress of AI
1:52:11
won't stop? And so, again, I keep going back to
1:52:17
this thing, but I think this is really what it is,
1:52:20
is we have to change. Like what it means to
1:52:22
be a person. All those things are only problems if
1:52:25
people stay what we are now, which is territorial primates
1:52:29
with desire to possess material goods for some strange reason
1:52:34
even though we're a finite life form. If all that changes,
1:52:37
if we change what it means to be a person,
1:52:40
which seems inevitable, then it won't matter. If we all–
1:52:46
if we could figure out a way– To literally engineer
1:52:52
out all of the issues that humans have with greed
1:52:55
and violence and all the different things that trouble society.
1:52:59
If the desire for that is no longer a part
1:53:02
of being a person, which is– that's doable. That seems
1:53:07
like if we're going to continue to evolve past Australopithecus
1:53:12
to Homo sapien 2026, if we continue the same amount
1:53:16
of time in the future– We probably won't be like
1:53:19
that anymore. The best versions of people aren't people that
1:53:22
want to steal your house and steal your land and
1:53:24
shoot you and take your resources. The best people are
1:53:28
the people that contribute and they're interesting and they make
1:53:30
you excited to be around them and you like it.
1:53:34
If that keeps going on and if that is aided
1:53:37
by technology, if we recognize that there's actual patterns of
1:53:41
human thinking and behavior that can be changed and that
1:53:46
if we all agree to– subscribe to this algorithm, to
1:53:49
connect to this computer program, connect to this external device,
1:53:54
or maybe not even external, maybe it's internal, that allows
1:53:58
us to communicate with each other telepathically. All that changes,
1:54:04
and then it doesn't matter. It doesn't matter if there's
1:54:07
any guardrails for AI or not, because we're not the
1:54:09
same thing anymore. All the problem, the worry that we
1:54:12
have about AI, the worry we have about power and
1:54:15
manipulation and the ability to influence people, all that kind
1:54:20
of goes away if people don't behave the way they
1:54:23
behave currently.
1:54:25
Speaker 1: Well, I guess here is one difference with your whitewater
1:54:30
rafting metaphor. In that metaphor, it's kind of we need
1:54:36
to hang on. There's a risk we could smash ourselves
1:54:38
on the rock, but if we don't do that, then
1:54:40
there's kind of one way we end up, right? Downstream.
1:54:42
Speaker 2: We get up in a nice lake. Yeah.
1:54:44
Speaker 1: But I think in reality, what might also be the
1:54:48
case that in addition to trying to not smash ourselves
1:54:50
on the rock, there might be places where the choices
1:54:54
we make affect the ultimate destination. Right.
1:54:59
Speaker 2: Go left or go right.
1:55:00
Speaker 1: Yeah. Or which stream do you follow? Maybe a more, yeah.
1:55:04
And where, like, if you steer in one direction, you
1:55:08
might end up in one place ultimately in some sort
1:55:10
of strange post-human world and maybe... it's really beautiful and
1:55:14
people have the chance to grow into their true self
1:55:16
and we love each other and are creative and take
1:55:19
initiatives and go a different way. Maybe you have a
1:55:22
paperclip maximizing AI or maybe that just everything is paperclip.
1:55:26
So going a third way, maybe you have like a
1:55:28
totalitarian system with like a small elite on top and
1:55:32
everybody else. Yeah.
1:55:34
Speaker 2: Or just sci-fi movie.
1:55:36
Speaker 1: Yeah. And there are many more possibilities and maybe some
1:55:39
that kind of we would, think would look kind of
1:55:43
good if we chose now but then in reality they
1:55:45
have some hidden flaw that we didn't think of so
1:55:48
if we chose those it would sort of be a
1:55:49
mistake and then maybe others that don't look that appealing
1:55:52
to us now if we just presented them in a
1:55:55
brochure to you like but then actually if you thought
1:55:57
hard about it maybe you would realize that actually would
1:55:59
be a really nice place to live because like in
1:56:03
the current world there are places that are nice to
1:56:05
visit that are interesting but then those are not necessarily
1:56:08
the same places we would want to live and raise
1:56:10
a kid right so there's There are like novels and
1:56:13
movies that are really fun to watch, but you wouldn't
1:56:15
want to live in those worlds. So there's a difference
1:56:18
between the kind of what makes for the good, interesting
1:56:20
story and the place we actually want to spend the
1:56:22
rest of your life. And so I think there would
1:56:24
be possibly a lot of opportunities to make foolish choices
1:56:28
or unwise choices or conflicts that sort of thwarts the
1:56:34
process and makes us end up, not just that we
1:56:36
go extinct before we reach there, but that the dare
1:56:38
that we reach might depend on the level of wisdom
1:56:41
and kindness that we have during the process.
1:56:44
Speaker 2: And there's always unintended consequences with every solution that we
1:56:47
try to find for any problem that we have. There's
1:56:50
always some new thing that comes up. We're like, Oh,
1:56:52
we didn't see that coming. I mean, it's certainly when,
1:56:55
whenever we're dealing with nature, like whenever they've brought in
1:56:58
invasive species to handle other invasive species, it always a
1:57:02
giant disaster, you know?
1:57:04
Speaker 1: Yeah.
1:57:06
Speaker 2: Always. There's always unintended consequences. Like, Oh, Now you have
1:57:10
100 million frogs. Who's telling us about that? Was it
1:57:13
Ali Sadiq that was telling us about the frogs? About
1:57:19
they brought in– what did they bring in? I forget
1:57:23
what the invasive species was. They brought in this one
1:57:27
species to kill another species and the problem, they didn't
1:57:29
realize that that species had been– it was like coyotes
1:57:33
or– I forget what it was. It was Guam. What
1:57:36
did they get rid of? Whatever it was, they got
1:57:41
rid of this thing that had been killing the frogs.
1:57:44
So then they had toads, right? Was it toads? So
1:57:48
they had millions of these things that were like all
1:57:50
over the highway. So you drive on them. You just
1:57:52
squash them everywhere you go because there was nothing controlling
1:57:55
the population of these things anymore. It's like unintended consequences
1:58:00
of– because we're very short-sighted. Even in our ability to
1:58:06
be contemplative about what the possibility of the future is.
1:58:10
When you're dealing with such an open-ended thing like AI,
1:58:14
super general intelligence that can maybe make better versions of itself.
1:58:19
Who the hell knows what that means? Are we making
1:58:21
a god? Because it seems like if it keeps going
1:58:24
and that– if it makes better versions of itself, it's
1:58:26
ultimately going to get to the point where it could
1:58:27
do anything. which is exactly what a god is. It
1:58:30
can make universes. You know, what if the ultimate end
1:58:34
of this is a big bang button that some scientist invents?
1:58:40
Speaker 1: Yeah. So we are a little bit like, yeah, it's,
1:58:43
I don't know, the different analogies one might reach for.
1:58:46
But I mean, I guess like, like say you're on
1:58:50
a plane and then like the pilot has passed out
1:58:53
or had a heart attack or something. Now it's like
1:58:55
we passengers somehow have to I don't know, figure out
1:58:59
how to do the, run the controls.
1:59:02
Speaker 2: Yeah.
1:59:02
Speaker 1: We're not really, I mean, you have to still try
1:59:05
to make your best, right? But then add to that
1:59:07
that all the passengers are fighting amongst each other. Each
1:59:10
one thinks they are the guy who should control the, like,
1:59:13
they're all convinced that they are the superior person to
1:59:17
try to land the plane. So, in addition to trying
1:59:20
to figure out how the controls work, they are, like,
1:59:22
actually having a big fist fight in the cockpit as
1:59:24
well and somebody is, like, dragging somebody else away and there's, like, the,
1:59:28
the kind of monkey tribe that is the current state
1:59:30
of the world here, and we're trying to shepherd, like,
1:59:34
this ship of humanity into utopia.
1:59:38
Speaker 2: It's... Yeah.
1:59:41
Speaker 1: Yeah, it's an interesting... It's Dr.
1:59:43
Speaker 2: Strangelove on steroids.
1:59:46
Speaker 1: But that's it. We might only have to, like, get it...
1:59:50
At some point, we should be able to hand over, like...
1:59:57
Once you have a sufficiently good AI, we maybe get
2:00:00
assistance from that point on.
2:00:02
Speaker 2: I think the first thing that it's going to take
2:00:05
over is government. We're going to realize how unbelievably inefficient
2:00:10
the government is at doing almost everything and how much
2:00:14
of the money that gets allocated is fraud and waste.
2:00:18
If you allow AI to sort through that and develop
2:00:21
much more efficient pathways to controlling and some.
2:00:26
Speaker 1: People would not want to see that.
2:00:29
Speaker 2: Exactly. That's going to be a problem. But if we
2:00:31
get to the point where there's some sort of, um,
2:00:36
some sort of a hive mind possibility, some sort of, uh,
2:00:40
I mean, one of the things that Elon said that
2:00:42
I thought was really fascinating, he said, you're going to
2:00:44
be able to talk without words. Well, if we're able
2:00:48
to talk without words, like does that eventually get to
2:00:50
the point where we could read minds? Could we, is, is,
2:00:55
Is thought and is communication no longer verbal? It's no
2:00:58
longer sounds. So right now we associate sounds that I'm
2:01:02
making where you know what words I'm using, what I'm referencing,
2:01:06
and we get a certain understanding of what each other
2:01:08
is trying to say. But what if that's just clunky
2:01:12
and that's silly? And what if instead of a scroll
2:01:17
that you leave in a cave somewhere, now you have
2:01:19
a movie that you can watch, like something much more
2:01:22
engrossing and much more– Much more powerful in that this
2:01:27
is what human communication becomes. It doesn't– and maybe this
2:01:32
is why the greys don't have mouths. We move away
2:01:34
from sounds because right now what we figured out is
2:01:37
like sort of like duct tape. We're communicating. We kind
2:01:42
of patched it up. We figured out something. We're just
2:01:45
going to use noises we make. Well, we have different
2:01:47
noises here than the people that live on this island.
2:01:49
They have totally different noises. I don't know what the
2:01:51
fuck they're saying. And then you have people on the
2:01:53
other side of the world, totally different noises. So Tower
2:01:56
of Babel type situation, right, where we really can't communicate
2:01:59
with each other unless we have translators. But if we
2:02:02
get to the point where that's not how humans communicate,
2:02:05
we communicate purely through thought and intention and understanding and
2:02:10
that it's no longer based on language. It's no longer
2:02:13
based on this is a transistor. This is a coffee mug.
2:02:18
Instead– It's a complete understanding of each individual thing that
2:02:22
we're discussing. Everything– you know what it is. You understand
2:02:26
what it is without it having to have a noise
2:02:28
associated with it.
2:02:30
Speaker 1: Yeah. I mean I guess the cybersecurity implications are significant
2:02:34
if you are giving direct access to other people to
2:02:37
transmit signals to your brain in a high bandwidth way.
2:02:42
That is not just words kind of that– It's almost
2:02:46
like you heard them even though there's no sound in
2:02:48
your ear. But if it's like actually directly kind of
2:02:51
interfacing in a high bandwidth way with your neural network.
2:02:54
Speaker 2: Also, encryption's out the window. If we no longer have encryption,
2:02:59
if we get to the point where Because if computing
2:03:03
gets to the point where the bottleneck is– like think
2:03:07
about money, right? What is money right now? Money is
2:03:10
all ones and zeros somewhere essentially, right? It's all bank accounts.
2:03:14
It's like we're not on a gold standard anymore. So
2:03:18
what if that bottleneck– it's an information bottleneck. Like someone
2:03:22
is preventing you from going into these places and getting
2:03:26
these ones and zeros and transferring it to your place.
2:03:30
But what if that is all– what if computing power
2:03:33
gets to the point where AI gets to the point
2:03:34
where those boundaries are nonsense now? All encryption is instantaneously decoded.
2:03:41
Speaker 1: I think probably cryptography is defense dominant in the limit,
2:03:50
I think. Like if you imagine– Mature technology, I think
2:03:54
it would be possible to encrypt. I mean, if nothing else,
2:03:56
you could use like a one-time pad, which would enable
2:04:00
you to encrypt things in a way that is unbreakable.
2:04:04
Speaker 2: Really? I feel like that's going to be a bottleneck.
2:04:08
And I think once we start reading each other's minds,
2:04:10
that might be the first thing to go. It would
2:04:13
be like the ultimate socialism.
2:04:14
Speaker 1: That's where you really probably would want encryption, right? If
2:04:16
you're going to transmit thoughts to me. Oh, yeah. Like
2:04:19
you don't want sort of have a– Oh.
2:04:21
Speaker 2: You don't want some asshole constantly in your head, your
2:04:23
next-door neighbor, just poking you and prodding you. Yeah. But
2:04:27
you would hope that along with this technology becomes like
2:04:32
a general state of enlightenment that the human beings achieve
2:04:36
where that's no longer the kind of behavior that we
2:04:38
indulge in.
2:04:40
Speaker 1: Which behavior do we no longer enjoy?
2:04:41
Speaker 2: Annoying each other.
2:04:42
Speaker 1: Annoying each other.
2:04:43
Speaker 2: Stealing each other's money. That kind of stuff.
2:04:46
Speaker 1: But then, yeah, I guess we go back to this
2:04:48
question of the utopian condition.
2:04:49
Speaker 2: Right.
2:04:50
Speaker 1: So there are a lot of things that individually are bad,
2:04:57
like lying, stealing, cheating, greed, excessive pride, like all kinds
2:05:02
of disease. Stubbing your knee is bad. Yeah. Yet, if
2:05:08
you sort of imagine removing all of those things, then
2:05:11
that changes the human condition quite profoundly. And to some people,
2:05:15
it would feel kind of maybe or appear flavorless or
2:05:18
sort of if there is no tension, no conflict, no
2:05:25
bruised ego. It might still be good, but it does
2:05:29
force us to sort of– I think it would be
2:05:31
a rather fundamentally different thing that we would be metamorphosing into.
2:05:35
if we went all the way in that direction, which
2:05:37
ultimately might be right, but it would require us to
2:05:41
kind of find new ways of realizing whatever values are
2:05:44
imperfectly realized in the current world through conflict and competition
2:05:49
and pain. Like some people get the motivation from painful failure.
2:05:53
And so if you get rid of the pain from failure,
2:05:56
then you'd need some other motivation, like some other thing
2:06:00
that drives you on, which there could be, like maybe
2:06:01
it's just a love of achievement and you feel kind of,
2:06:04
neutral are just less happy when you fail. But you
2:06:06
would still need something that kind of preserves whatever structure
2:06:11
it is that we think is valuable in the current
2:06:14
human condition. Unless you go all the way to sort
2:06:16
of radical hedonism and think the future is best if
2:06:19
we were all just kind of floating in some kind
2:06:22
of drug-induced euphoria as blobs that experienced immense pleasure but
2:06:28
had no real texture in our experiences, didn't engage in
2:06:33
activities and didn't interact with each other. Like there's like
2:06:37
a philosophical view where ultimately pleasure is the only thing
2:06:39
that matters and the minimization of suffering. So if that's
2:06:43
your axiology, then it's relatively easy to see then how
2:06:46
a technological maturity, you would achieve a sort of optimal state.
2:06:50
But if you have a more complex value system where
2:06:53
maybe pleasure is one good thing, maybe really important, but
2:06:57
there are also other things like appreciating beauty, you know,
2:07:00
true friendship, courage, achievement. And ideally, you'd want a future
2:07:04
that includes all of these things. Then you need to
2:07:07
do a little bit more sort of design work to
2:07:10
figure out a way to combine them all in a
2:07:13
meaningful way.
2:07:14
Speaker 2: But this all comes back to our idea of human meaning,
2:07:17
what's important to humans, our finite 100-year lifespan adoption of
2:07:22
this concept of meaning. But the black hole doesn't give
2:07:25
a shit about human meaning. And it's going to be
2:07:28
around a lot longer than us. And it's got a
2:07:31
lot more power than us and it's doing a lot
2:07:33
more change than us and we want to think that
2:07:36
we're more important than black holes.
2:07:39
Speaker 1: Yeah I mean because we are I think we are
2:07:42
I mean to us yeah to us.
2:07:45
Speaker 2: But to the universe is it is human meaning that
2:07:47
important to the universe or is it just sort of
2:07:49
a placeholder for like what will ultimately become is it
2:07:53
is it motivate us to continue to progress.
2:07:58
Speaker 1: Well so I think like at technological maturity, there are
2:08:01
certainly forms of purpose that you could have. You could
2:08:04
have artificial purpose. So this is when you basically set
2:08:08
yourself a goal for the sake of having the goal
2:08:11
and then doing the activity. So maybe you set yourself
2:08:15
the goal, I'm going to get this little white ball
2:08:18
into a sequence of 18 holes. And not only that,
2:08:23
but in order to achieve this goal, it's part of
2:08:25
the goal that I'm only allowed to use this very
2:08:28
inconvenient method. I can hit the ball with a club, right?
2:08:32
Much easier to just pick it up and put it
2:08:34
in each hole successively, but that doesn't count as being successful.
2:08:39
So you could make up this goal, pretty arbitrary. Now,
2:08:43
once you have that goal, then you now have a
2:08:45
reason to try hard, to concentrate, to perfect your swing,
2:08:49
and you can play golf. So the goal enables you
2:08:52
to do this activity of golf playing, which maybe you
2:08:54
find fun or worthwhile or meaningful.
2:08:58
Speaker 2: And the reason why you find it meaningful is because
2:09:00
it's difficult to do.
2:09:02
Speaker 1: Yeah. And so the future would consist, I think, if
2:09:08
we succeed in a lot of game playing and You
2:09:11
could certainly have these artificial purposes that you set yourself,
2:09:14
these goals, that then give you a reason to engage
2:09:16
in an activity.
2:09:17
Speaker 2: Now we're back in the world of Half-Life. Now we're
2:09:21
back in a video game. Now we're also in the simulation.
2:09:24
This is going to be your artificial goal.
2:09:27
Speaker 1: And you could imagine, I think, maybe we shouldn't think
2:09:30
of video games here, but it could be much like
2:09:32
games we can't even imagine. It could be like society-wide
2:09:36
games that last for 20 years that involve all kinds
2:09:39
of multimodal things and little groups that work together to
2:09:43
like come up with new ways of creatively. And so
2:09:49
in that broad sense of kind of things we do
2:09:51
for their own sake, I think game playing could be,
2:09:54
and it's like a lot of what children do, they're
2:09:56
kind of for curiosity and spend a lot of their
2:09:57
time playing games and we might all be like kids again.
2:10:00
What might be in Shorter Supply is sort of natural purpose,
2:10:05
like purposes which... we don't just arbitrarily make up in
2:10:11
order to have a purpose, but that they're sort of
2:10:12
given to us. So right now in the world, you
2:10:16
might say, you know, making a living is not just
2:10:20
an arbitrary purpose because there are real consequences if you fail.
2:10:23
Like maybe eventually you get kicked out from your flat
2:10:27
and then, you know, it's really cold and you get
2:10:30
rained on and like horrible things happen. So like similarly,
2:10:34
if you're, Like don't brush your teeth. Eventually you will
2:10:37
have tooth decay and there will be real consequences. So
2:10:40
there are various things that you have reason to do
2:10:44
because there are real negative consequences if you fail to
2:10:46
do them. And a lot of our lives is structured
2:10:48
by these natural purposes. At a societal level, there is
2:10:51
a whole bunch of things we need to do together, right?
2:10:54
In this future world, maybe there would be many fewer
2:10:57
of those natural purposes because for any one of them,
2:10:59
you could just ask the AI to sort it out.
2:11:02
And so the artificial purposes would be a larger chunk.
2:11:04
It's interesting to think, are there any natural purposes that
2:11:07
would survive to technological maturity? Like any things that we
2:11:11
still have sort of instrumental reasons that we need to
2:11:14
do ourselves. And I think there might be a few,
2:11:19
but they are more subtle. They might not strike us
2:11:21
currently as very important, but it's one of those things where, like,
2:11:29
you know, that During the day, if you're outside, you
2:11:32
can't see the stars, right? It's not because they're not there.
2:11:36
It's because there's so much light that they are sort
2:11:39
of blotted out. But at night, when the stronger light
2:11:45
from the sun is absent, you can see these fainter lights.
2:11:47
I think similarly in this future, there might be one
2:11:52
sort of urgent screaming moral values of immediately pressing practical
2:11:56
concern go away you might be able to perceive a
2:11:58
whole constellation of these more subtle values that we are
2:12:01
blind to currently so take the value of take the
2:12:04
value of like i don't know like some like honoring
2:12:06
uh your forebears so right now it doesn't seem i
2:12:11
mean maybe it's nice sometimes to remember your your past
2:12:15
parents or some historical hero who did something good that
2:12:18
benefited humanity, right? But it's not the main thing that you're...
2:12:23
Maybe that would be a bigger thing if there was
2:12:25
nothing else you needed to do. Maybe you could actually
2:12:27
spend serious time or spiritual quests. Even for people who
2:12:32
are very religious, a lot of their actual waking hours
2:12:35
are spent on random other things, doing their laundry, driving
2:12:39
to work, all kinds of stuff. If all of that
2:12:42
was automated, you could imagine spending more time on trying
2:12:46
to align yourself, orient yourself to this higher being and
2:12:51
trying to be in communication with them. Like maybe aesthetic values.
2:12:56
Like there are maybe some things that would just be
2:12:58
kind of nice and cool if the world were like that.
2:13:01
We don't really have time to worry so much about
2:13:02
them now, but if there was nothing else on the agenda,
2:13:05
like coming together in a way that upholds some tradition
2:13:10
in a beautiful, original way that still is true to
2:13:13
the original spirit, together with other people and enacting some ceremony,
2:13:17
like maybe those things would start to fill more of
2:13:19
our time together. in conjunction with this game playing. And
2:13:25
there might be many other of these kind of subtler
2:13:28
values that would start to shape what people were doing.
2:13:33
Speaker 2: Yeah. And ultimately, who knows?
2:13:37
Speaker 1: Yes.
2:13:40
Speaker 2: It's very interesting and it's very open-ended. And we really
2:13:43
don't know what's going to happen, but we're probably going
2:13:45
to see it. We're probably going to see the strangest
2:13:49
thing that humans have ever had a possibility to experience.
2:13:53
Speaker 1: Yeah, and in the end, I guess it's trustful.
2:13:56
Speaker 2: Yeah. Well, I mean, these conversations are always fascinating, and
2:14:02
who knows? Let's do another one in a few years
2:14:04
and see how off we are. Come back in four years,
2:14:09
all right?
2:14:10
Speaker 1: If we have four years.
2:14:11
Speaker 2: If we have four years. If you can, if you're
2:14:13
allowed to travel in four years, come back and let's
2:14:15
see how wrong we were.
2:14:17
Speaker 1: We can update.
2:14:19
Speaker 2: Well, thank you very much. I really appreciate you coming
2:14:21
in here. It's great to see you again. It's always fun. Okay.
2:14:25
Speaker 1: All right.
2:14:26
Speaker 2: Think about it, kids. Bye, everybody.