Latest / Nonsense with cj little and Jeff Parker / The Great Displacement: Who’s Getting Fired Next?

Transcript

The Great Displacement: Who’s Getting Fired Next? · Nonsense with cj little and Jeff Parker

0:00
Welcome to Nonsense Season 4, Episode 23, our big story on our
0:02
second show of the week, I'm Jeff Parker.
0:04
I'm CJ Little. This time we're looking at
0:06
coding was only the beginning. The next 5 industries facing an
0:09
AI overhaul up next. First they came for the coders,
0:12
Jeff, and you said nothing. Because I was a coder, but
0:15
coders don't like to talk to people.
0:16
First of all, you, you weren't a coder.
0:17
You are a coder that doesn't just leave.
0:20
You're not like one day like I, I no longer develop software.
0:22
You're always developing software.
0:24
I do develop software but I do a lot of AI coding now, a lot of
0:27
vibe coding. OK, I enjoy it.
0:29
First of all, I have to re evaluate our relationship.
0:32
If you're a vibe coding now. Have you not tried having the AI
0:36
code for you? I have a little, it does make me
0:38
feel a little dirty. I'm more of like let it write me
0:40
like functions. I'm not like the full blown let
0:42
me build the thing. Oh, I started out doing
0:44
functions. That's exactly how I started
0:46
out, Yeah. And then I just got out of the
0:48
way. Yeah, the problem is I know
0:50
what's buried down there. And then when I go look at it,
0:53
I'm just, it just, it gives me the hives and I just, I just
0:56
can't. No, it's beautiful code.
0:57
Now you, you haven't done it in a while.
0:59
It's there's, there's nothing I can write now that's much
1:01
cleaner than what the AI writes. I did have this idea a while ago
1:04
where I was like, just don't look at the code, look at the
1:06
executable. Then you'll be fine.
1:07
You're like, oh, those ones in zeros.
1:09
No, no, no. Look at the code.
1:10
Now look at the code. OK, it's great.
1:13
What are you looking at? Python?
1:14
What do you? What are you having to write
1:15
stuff in? I'm reluctant to say Rust
1:17
because I know you're going to make fun of me.
1:18
No I'm not. I love Rust.
1:19
I'm pro rust, right? Esoteric weirdo.
1:22
I knew that was coming. I do.
1:24
I really. I'm not.
1:24
I'm not being a smart ass. I think it's a good choice.
1:27
I really do. OK.
1:28
It's like a solution looking for a problem that it'll never find.
1:31
But I think it's great. Fantastic.
1:32
OK, the industry's AI hits next. Let's go.
1:35
I'm just happy that podcasting is safe.
1:55
Coding was the Canary in the coal mine, but the mine just
1:57
collapsed. If you thought AI was only
1:59
coming for software engineers, think again.
2:01
The era of pure digital disruption is over.
2:03
The era of everything else has begun.
2:05
I love that in your mind there's only two parts to the world.
2:09
There's software engineering and then all the other stuff like
2:12
that's it. What's your point?
2:13
That's I mean, what's your point?
2:14
I'm just saying I appreciate that you think that way where
2:16
you're just like you're either making software, you're doing
2:18
one of the other things. From the operating room to the
2:20
courtroom, algorithms have graduated from writing scripts
2:23
to making life altering decisions.
2:25
We're watching 5 pillars of the global economy, healthcare,
2:28
finance, retail, manufacturing and law undergo a total
2:31
structural rewrite. It's no longer about a chat bot
2:34
answering emails. It's about autonomous agents
2:36
managing portfolios, robots designing their own parts and
2:40
diagnostics that see what humans miss.
2:42
The sandbox period is over. It's time to take a hard look at
2:45
the five industries currently being dismantled and rebuilt by
2:48
AI, and why the changes are moving faster than anyone
2:51
predicted. OK, so let me understand this
2:53
straight. Yes, yes.
2:53
You have a list of industries that AI is going to to decimate.
2:56
I do. And I have a list, yes.
2:57
So we're gonna we're gonna like take turns giving our
3:00
prognostication about which will interlace.
3:02
OK. And that's good.
3:03
Good. Use the word interlace.
3:04
Thank you. And we'll prognosticate about
3:05
which industries are likely to be the next that are disrupted.
3:07
Is that sort of? I mean, nobody really knows for
3:09
sure though, right? Oh no, I do I I my list is not
3:11
so much a list as is the answer. Sheet.
3:12
Oh, that's what we're gonna do now.
3:13
I see. Yeah, yeah, yeah, yeah, I see.
3:15
So I'll try to guess the right ones.
3:16
OK, the short answer is anything that involves structured logic
3:19
and high fidelity replication. Great, the only things in life I
3:22
find enjoying pretty much great. Awesome.
3:25
So chaos isn't being replaced yet?
3:26
That'll be awesome. Not so much, no.
3:28
But we say structured logic that tells me like the real estate
3:31
agents are going to be fine 'cause there's not a lot of
3:33
logic going on over there. Yeah, but the way houses are
3:35
shown, sure. The paperwork especially like in
3:37
California where the paperwork's somewhat complex, Yeah, that'll
3:40
all go away. That'll all change.
3:41
It's. Just not the same when you take
3:42
the AI and you put the low cut top on it, no and shove it into
3:45
a, you know, a Mercedes S class. It doesn't.
3:47
It's not but. You can see the tops of its
3:48
transistors. You can't see the good ones.
3:51
The AI's that sell the homes are the ones show off a couple
3:53
transistors at the showing. They're just like, hey, look
3:55
over here, look at all look at my dims and if you're really
3:59
lucky I'll show you one of its dim slots.
4:02
My AI agent likes to have two dims put into its slots at the
4:05
same time. That's what it's I'm sure it
4:07
does. It prefers to be double bang.
4:08
Please stop, I'm begging you. I'm more of a raid one sort of
4:11
girl. Oh, OK.
4:14
It's fun to have jokes. Knowing we'll get legal services
4:16
is my number one that I'm going to throw out there.
4:18
The legal industry is moving past.
4:20
Good job Kreskin. Amazing.
4:22
But I think it's going to get even worse, I think now I, I
4:25
think you still have law firms who are still, yes, clinging to
4:28
the old ways of doing. Things, well, you're always
4:29
going to have that, that, that that's not going to change.
4:31
I think you're always going to folks that just don't adapt to
4:33
the change. But I think those people are
4:35
going to be slowly finding themselves with less and less
4:38
work. I would disagree with you about
4:40
slowly. Yeah, I don't have to be
4:41
anything slow about that. I mean, there are so many kids
4:44
who just graduated law school and instead of hiring 20 kids
4:47
fresh out of law school, most big firms will pick up two or
4:49
three. And those two or three all think
4:51
they had oh, I made it. I made the No, you didn't.
4:53
Yeah, you made the cut for this week.
4:54
It begs the question though, with the power of these
4:57
efficiency tools, what is the number of of lawyers that we
5:00
need? Has that number gone up, down,
5:02
or stayed the same? It's.
5:03
Probably gone way down. Think how fast AI may generate a
5:06
contract and that contract, but look at kind of contemplate all
5:09
sorts of things that that individual lawyer might not have
5:12
contemplated if he's never written that particular contract
5:14
before. Great point, but I've got 2
5:16
counterpoints to this one. If you believe that AI is going
5:19
to make a bunch of industries more efficient, which I think we
5:21
can all kind of agree that that's going to happen, yeah,
5:23
then you can do more. And if you can do more, you're
5:26
going to need more, right? You're going to need more,
5:28
especially legalese. So I don't know what the balance
5:30
is, but we keep applying like, oh, well, the legal stuff that
5:33
we're currently doing, we need we're going to be able to do
5:36
much more efficiently. Well, yeah, maybe it goes down
5:38
by, let's say an efficiency factor of 80%.
5:40
A massive number. Yeah.
5:41
Well, if you get 5X the amount of stuff that comes out of it
5:43
that you need, then you're you're even Steve.
5:45
Think. About it in a really big case
5:47
where you've subpoenaed a ton of e-mail for sure.
5:50
I mean, there's millions. Discovery analysis and yeah.
5:52
Yes, exactly. For sure.
5:53
And all of a sudden the AI is going to do that in like a
5:56
matter of seconds. Of course, but like how many
5:58
attorneys out there are going, oh man, you know what I love
6:01
doing? I love reading Discovery.
6:03
That's my favorite. Thing, but it's a huge part of
6:05
the job. No, I get now we don't need
6:06
those hours. Those billable hours are gone.
6:08
I get that. But we have the opportunity to
6:11
do more things with what was previously those billable hours.
6:13
I don't really know. I'm not trying to be
6:15
provocative. I really think there is this
6:17
opportunity where we can do more because we have more tooling
6:20
behind it. So if we can do more, you may
6:23
need more effort from for more attorneys.
6:24
I don't know if it's going to go up by 10% or down or down by
6:27
80%. I just don't know.
6:28
But I suspect those things are going to balance each other out.
6:30
Yeah. If you're, if you're right now
6:31
spending, let's make up a number and say 10 grand to draft a
6:34
contract from your legal firm and you know they're going to
6:36
bill you for whatever 12 hours to understand what you're trying
6:39
to do, the deal you're trying to do and craft a contract.
6:41
If it all gets replaced with $0.10 of compute, what is the
6:44
value now that your your attorney is providing?
6:46
They're steering the AI, but they just marking up their time
6:48
more. Yeah.
6:49
At first it's steering the AI and it's making sure the AI.
6:52
Bias and hallucinity and not so much anymore, but.
6:55
Yeah, all that goes away, all that diminishes overtime.
6:57
What? Does the role of the lawyer
6:59
become, is it, are they just the one taking liability?
7:02
Are they just putting the liability shield saying, well,
7:03
you're using my LLM's, so now we're the liability shield on
7:06
your contracts? I don't know.
7:07
I'm I'm making stuff. Up yeah, I don't know, I don't
7:09
know what the value add. Is but right now you're like,
7:11
oh, hey, attorney, you've been doing this for 20 years.
7:14
What's your thoughts of like, am I gonna win in this case?
7:16
Right. If you're litigating, I think
7:17
when you get that into data analytics and you get that into
7:19
the world of compute, that just becomes a number that it's been
7:22
doubt that you of course aren't going to trust.
7:24
You're going to be like, Oh yeah, 97.6% chance of winning
7:26
this case. I don't believe you, but that's
7:27
going to get really interesting. That's a little bit like looking
7:30
at these stats of the basketball teams and trying to predict
7:33
who's going to win. You do actually have to play the
7:35
game. Sure.
7:35
You do have to actually get in front of the jury.
7:36
You do have to actually get in front of the judge, present your
7:39
case and be the guy who's got the best looking suit on and all
7:42
the things that help you in persuading people that you are
7:44
someone who is of merit that they should listen to.
7:47
But you are presuming that the judge stays the way the judge is
7:50
when the judge just becomes like a Ticketmaster.
7:52
Event window that opens and you just must submit your case by
7:55
10:00 PM on Thursday the 12th. And then you just jam it in and
7:59
it just spits out an answer. Like what then I like that idea
8:02
because I think you can probably get a less prejudice judge if an
8:07
AI is the judge. If you get a thing more you, I
8:09
would hope I would. I mean, certainly that's what's
8:11
happening in baseball in the strike zone.
8:13
A person who was normally judging, is this a ball, is this
8:16
a strike is now very, very accurately.
8:19
So we can all see, yeah, applying the law is the same for
8:22
everyone. If you you assume you're already
8:23
gonna have a big offset in the lawyering, right?
8:25
The lawyers are going to be. Yeah, utilizing AI, you're gonna
8:28
have less lawyers. Should the same happen on the
8:30
judge and clerking side? Should it be?
8:32
Yeah, sort of equal? It kind of makes sense, I guess,
8:34
right? One should necessarily lead the
8:35
other. We've solved the problem.
8:37
I like it. Yeah, no good work.
8:38
Your turn, Sir. OK, I like legal that I think
8:40
that is kind of an obvious one. I'm gonna let you do the obvious
8:42
ones 'cause you want your list to be perfect or whatever.
8:44
I'm gonna I wanna take some some swings.
8:46
Yes, fashion, fashion I'm. Gobsmacked.
8:49
Tell me more fashion and apparel.
8:50
Yeah, here's. The thought I have you've got so
8:52
much material so like you have to kind of guess that people are
8:56
gonna like your stuff and then you have to guess the sizes.
8:59
Right. And team who already did this
9:01
sort of just in time manufacturing where they try to
9:03
analyze trends. Yeah.
9:04
Show you a product that doesn't actually exist.
9:07
Sure. And then you go Oh yeah, I'd buy
9:08
those and then they they make them right.
9:09
It's the sort of this on demand. But they've already sold you the
9:12
product before it even existed Yeah.
9:14
As a physical good. Once you sort of see that I
9:16
don't understand how we keep doing what we've been doing
9:18
right. The whole like, make a product
9:20
six months ahead of time, ship it out and then let the tailors
9:23
deal with clearance stock and all that.
9:25
I think they call it dead stock. All that stuff that nobody
9:27
wants. Yeah.
9:27
So what if you just if you rethink that through the the
9:29
eyes of AI and manufacturing effectively, like sewing
9:32
machines become the 3D printers of clothes.
9:35
And now you've got an AI looking at trends and just going like,
9:38
hey, my, my cost to make this stuff is nominal.
9:40
Yeah. And I see this current trend on
9:41
TikTok that is people running around in marshmallow suits and
9:44
you're just like, I can make you a marshmallow suit custom
9:47
tailored to your size. Yeah.
9:48
Now you don't take the viral part You don't.
9:50
I mean, I assume you would always take the sizing part, but
9:52
now clothes just become this thing that is completely custom
9:55
for. You well even the sizing could
9:56
be yes, that's. What I mean the sizing, yeah.
9:58
Can look at you over and say I can make that perfectly for you,
10:01
yes. You've done a little body scan
10:02
with your phone. Everyone's listening to the
10:04
show's phone has LIDAR on it by now, right?
10:06
So you do like a little scan of yourself and then you pump it
10:09
into some equation. Some algorithm looks at it and
10:11
goes, OK, this is how big you are, this is how you cut the
10:13
stuff. And now you want to get that
10:14
jacket that you saw online. It just cuts the jacket for you
10:17
and sends it to you. I, I don't see a world where
10:19
something like that doesn't happen.
10:20
It just seems so obvious. To me.
10:22
But I like to go and try on clothes and see how they look on
10:25
me. Oh yeah, I mean, I guess.
10:26
You're right, we could never solve that problem with
10:28
software. There's no way how would we ever
10:30
put I can literally render Benjamin Franklin oil wrestling
10:34
with Thomas Jefferson and you're like, oh, how would it put
10:36
clothes on me? I.
10:37
Have no idea how would I look in that sport jacket.
10:39
Right? It's not possible.
10:40
Anyway, I don't know, I see something like this.
10:42
I can't tell you exactly when it's gonna happen, but I do feel
10:44
like a big chunk of that industry is gonna get replaced.
10:47
Especially you think about all the the management around the,
10:50
the sort of logistics of the the dead stock and just moving stock
10:52
around. Huh.
10:53
I think that's going to get wiped out.
10:54
I'm not entirely sure how the manufacturing process is going
10:57
to get changed so that you can have an autonomous system build
11:00
clothes soup to nuts. Yeah, to a given size.
11:02
But I do feel like that's a relatively easy to easy medium
11:05
problem to solve. And when that's solved, one, we
11:07
might actually get manufacturing back closer to here because now
11:11
what have you eliminated your marginal human cost?
11:13
Right. So now you can just deploy the
11:14
stuff in, you know, a warehouse in Arizona and it just cranks
11:17
out clothes. If you're a manufacturing depot
11:19
that can make stuff. All of a sudden, brands become
11:22
just marketing. And now, like, do you have to
11:23
have your own factory that makes jackets as long as they can make
11:26
it to your same specs? You have a set of factories that
11:28
makes jackets all over the country.
11:30
And then you put your brand on be like, well, mine's brown and
11:32
looks like this and like I stamped my brand on it.
11:35
Every piece of clothing becomes custom.
11:36
The only reason we couldn't do it before is because we had no
11:39
way to manufacture stuff at that scale.
11:41
Yeah. You basically put everybody into
11:42
smaller boxes. And you say we've got 6 sizes.
11:44
You can have any size you want as long as it's one of these
11:46
six. And once you go, well, no.
11:48
If I, if I can make the stuff and ship it cheaply, Amazon's
11:50
figured how to ship stuff cheaply.
11:52
Yeah. So now like it can all be one
11:53
off and that is. Super cool.
11:55
To me, that is kind of cool. That is kind of cool.
11:57
The creative side of it still baffles me a bit.
11:59
I I still think you need someone who dreams up Yes what these
12:03
things will look like at some point.
12:04
And I think though at some point that's going to get steamrolled
12:07
out, but at least for let's say the first part of this process,
12:10
yes, you want someone that's picking the trends and saying
12:12
this is what looks good and what doesn't.
12:13
But at some point if it if you're the clothing zeitgeist,
12:16
the AI clothing zeitgeist just kind of figures it out.
12:18
How do you drive the direction? I think it just becomes like a
12:21
like, almost like an influencer thing.
12:22
Yeah. Whoever gets the most views
12:24
becomes the one that's driving that.
12:25
Direction, I guess there are actually still designers like I
12:28
think back a few years I have things that are from Calvin
12:31
Klein that I really like of course that were because Calvin
12:34
Klein in his mind saw things a certain way.
12:36
Yeah. They're not vastly different
12:38
from everybody else's clothing. Sure, they're different enough
12:40
that I go, oh, I like. That, but I suspect there will
12:42
be more James purses. I don't know, man.
12:44
That's a one-of-a-kind brain. I don't think there's a ton of
12:46
those. When Ralph Lauren sold his
12:47
company, yeah, I have a hunch that all the assets of the
12:51
business walked out the door between his ears.
12:53
There's a pretty big asset in the brand.
12:54
I think in terms of growing the brand, most of that was probably
12:57
between his ears. Sure.
12:58
I'm looking at this differently. Are there brains out there that
13:01
can design things like a Loren or a Pierce that currently
13:05
aren't because they haven't come across the Capitol?
13:07
Or maybe. So I think this unlocks those.
13:10
Yeah, You get more of those. I'm not saying everybody should.
13:12
I'm not saying you and I should be designing pants, although I
13:14
think we should it be really funny.
13:15
No, But the point is there's a larger subset of people that can
13:19
design clothing well than the ones that are currently, and I
13:21
think this unlocks that. Yeah.
13:22
Anyway, legal. Yeah.
13:23
Amazing that you came up with that.
13:25
Such a such a Yeah. Prophetic.
13:26
Thing fashion is pretty good. I don't think that's happening
13:28
anytime soon, but I think that's very cool and I like where
13:30
you're going with. It's already happening though.
13:32
You're already seeing this on T move.
13:33
Let me come back to one now that's that's happening right
13:36
now. Another layup.
13:37
OK, it's another. Layup.
13:38
I'll give you another layup, accounting services.
13:40
Wow, man. Wow.
13:42
Accounting isn't just being disrupted, it's being
13:43
structurally rewritten. The industry has moved past
13:45
simple automation into a genetic accounting where AI doesn't just
13:48
suggest edits, it executes the workflow.
13:50
Yeah, data entries long gone. I mean, that's that's long gone.
13:53
You snap a picture of a bunch of receipts and they all get put in
13:56
perfect. Sure.
13:57
The only reason AI came for software development before
14:01
accounting was because AI was built by software developers and
14:04
we wanted to help ourselves. Yeah, they wanted to make their
14:06
own lives easy, but I would. Argue.
14:07
The only thing that is more logically straightforward than
14:10
software development is accounting.
14:12
Like, yeah, I'm pretty sure let's say I'm around like 15
14:15
years ago, you could just feed everything into Clippy and it
14:18
could have done all this. Microsoft just wanted to sell
14:20
more Great Plains licenses and they didn't do it.
14:22
Like, I don't feel like this is a far stretch to be able to just
14:25
feed everything and be like, here's your answer and just.
14:27
Spits it up. I think currently we're probably
14:28
at a point where bookkeeping is maybe 80% touchless.
14:31
That's gonna be 100% in the blink of an eye.
14:33
Yeah. Is it is 80% based on what
14:35
baseline? What?
14:36
Like what are you measuring back?
14:37
Because you've got, you still have to categorize transactions.
14:40
You've got to flag anomalies. No, I understand, but you're
14:42
saying it's all? That's gonna get fixed.
14:44
It's 80% touchless. Yeah, of all like the total
14:46
volume of work. OK, I see what you're saying.
14:48
I see what? Yeah.
14:48
I mean, most of that now is can be automated.
14:50
Most of that the AI can figure out.
14:52
There's still just a handful of things where you just going to
14:54
know what category to put this. Under 80% might be a low
14:57
watermark currently. I don't know.
14:58
I feel like it's higher. We have to talk to some, some
15:00
big enterprise accountants to see like, I'm sure they all want
15:02
to preserve their job and their partner status, But I feel like
15:04
a lot of those things now are doable.
15:06
Like, what are you really still touching?
15:08
What's the thing you're still doing?
15:09
I feel like, and I mean, I've been saying this now for at
15:11
least two, if not 3 decades. Tax preparation.
15:14
Well, that's the big one. Why there's a human in that
15:16
chain makes no sense to me. Yeah.
15:17
I'm 100% with you on this one. OK.
15:19
Amazing foresight. Incredible.
15:20
Thank you. Has anybody else come up with
15:21
this? No one, just me.
15:23
The amount of thought and effort you must have put into figuring
15:24
out that accounting is gonna be replaced by AI.
15:26
I'm like Kreskin. This one I I would put under the
15:28
bucket of layup, but I don't it doesn't get talked about a lot.
15:31
Oh, you're doing a layup now? It's layup.
15:32
Esque yeah, I'm gonna, I'm gonna, I want to give you a bone
15:35
here on this one. Education, I think education and
15:37
specifically like what we would now call tutoring like a custom
15:40
education development. Yes, that to me.
15:43
I'm gonna interrupt you and just say this is also my #3 so we can
15:46
combine and just discuss it because this is this is the
15:48
exact #3 that I. Have, oh, jumping on the
15:50
bandwagon. OK, Yeah, All right.
15:52
Well, I told you is an easy one. When you come in to do the
15:54
layups, you know I don't do, That's what I.
15:55
Do I think that becomes the alley oop?
15:57
I don't know anything about basketball.
15:59
No, clearly. It does seem like again, go like
16:01
go back to the example I used with yes, clothes.
16:04
We made clothes in different sizes because you could only
16:06
make a finite number of sizes. The number had to be relatively
16:08
small. You just get the fit in one.
16:10
They call that the assortment. Yes, I think the same thing for
16:12
education. We teach.
16:14
We have taught in a way because we did one-size-fits-all a
16:16
better way. Yeah.
16:18
And we had found ways like, OK, hey, we're going to do a
16:20
lecture, and then we're going to do some homework.
16:21
And then I'm thinking about, like, secondary school here.
16:23
Yeah. And then you're going to have
16:24
labs Like those are all different ways to teach the same
16:27
material. And if you missed a lesson or
16:29
you didn't comprehend a concept in, say, mathematics, yeah, and
16:33
you didn't figure that out, you were one of those students that
16:35
just fell behind and fell behind and fell behind.
16:37
And getting worse, especially for things like in math where
16:39
things stack up and if you don't get one of those concepts, you
16:42
are absolutely hosed. The AI can see you're not
16:44
getting this concept. They can keep teaching you that
16:46
and showing it to you in different ways until eventually
16:49
the light goes on and you go, oh, that's.
16:51
It also the other version of that, which is the, oh, you
16:54
already know these things. We don't want you to get bored.
16:56
You don't spend time on this. Let's rocket you into the things
16:59
that you don't know. Be able to dial in the education
17:01
to what is is the best for your human.
17:03
I think, right. Some people refer to this like
17:04
it's like the infinite Socratic method because now everyone can
17:07
have their own sort of Socratic method.
17:09
Once I start thinking this way, I can't like, I don't even know
17:12
why we're still teaching the way we're teaching right now.
17:14
Yeah, right. I mean, I know logistically it's
17:15
mainly because you've got to get the teaching staff up to speed
17:17
on this and that it's not going to be a a short or easy process
17:20
that's going to take some time. Well, that's why the tutoring
17:21
will go first. What do you mean?
17:23
Because you don't have to. Be credentialed tutor.
17:24
You can be a tutor. You can be a fabulous tutor and
17:27
have no credentials. Now you can go and you can go
17:29
online and you can talk to the AI and it can tutor you.
17:32
That will happen before the classroom education system in
17:35
schools begins to. Be a sure yeah, that's what
17:36
you're saying. Just 'cause it it can get there
17:38
faster. Yeah, there's less resistance.
17:40
For all sorts. That's a good point.
17:41
Yeah, that's a good point. And I think the the tutoring
17:42
should ultimately go away. To me, the tutoring should be
17:44
the learning, like what we now call tutoring should become the
17:47
learning. As you know, I have taught many
17:48
people to do code, and I enjoy the process of watching the
17:50
lights go on. Sure.
17:51
As I communicate things in many different ways until I
17:54
eventually get them. Oh, yeah.
17:56
Oh, yeah. OK, great.
17:57
And they now I just. That all goes away.
17:59
The last guy that I taught to code.
18:01
Yeah. AI was in existence.
18:03
Yeah. And I said here's what you're
18:05
gonna do. Yeah.
18:05
You're gonna download this compiler.
18:07
And I actually started him on Rust.
18:09
I said, you're gonna download this Rust compiler.
18:11
You're gonna install it on your machine.
18:12
I don't know how to do that. Yeah.
18:13
Yeah, that's right. You're gonna go to Gemini and
18:15
you're gonna ask it, how do I install a Rust compiler?
18:17
Taught him how to fix. Yeah.
18:18
Yeah. And I kept walking him through
18:19
the little bits. Yeah.
18:20
And I taught him Gemini. Yeah.
18:21
And he went, oh, that's how you do it.
18:22
Now write me a program that does this.
18:24
And Gemini helps him, gives him all the code and everything.
18:26
And he starts. Yeah.
18:27
And little by little by little, the lights were coming on so
18:29
fast and I did almost nothing. Literally in an evening, he
18:33
learned to the big the basics of coding.
18:34
Toward the end of the evening, I said, now go to Gemini and ask
18:37
what's another project I should have?
18:38
What's a little another little coding project I should do that
18:40
can teach me something slightly new, you know, just a little bit
18:43
of new stuff. Yeah.
18:44
And he started doing that and work great.
18:45
And as I walked out the door, the very last thing I said to
18:47
him is now I want you to go to the AI and I want you to ask it
18:50
what is GitHub? OK.
18:52
And by the next day he was committing his code.
18:55
He knew what he was doing. It was like, this is the easy.
18:57
I've never had such an easy time teaching someone how to do
19:00
something. This is wonderful.
19:02
This is customized learning, and I think that's what's so
19:04
amazing. More importantly, you taught
19:05
that individual how to fish. Yeah, that's to me is the most
19:07
amazing. Part, but mostly Gemini did the
19:09
work of. Course, that's the fishing part.
19:11
Yes, yes. And he'd never used an AI
19:13
before. He literally never used an AI.
19:14
He was blown away. But I need some sort of what
19:16
they call a command line to be able to get this to load.
19:19
Well, ask Gemini. It'll tell you how to do it as
19:21
you. Go around the Rubicon and I've
19:22
started to find, cause I've done this with people where they were
19:25
like, you know, call me for things and I'm like, well, let
19:27
me just show you like it'd be 20 years ago.
19:28
You show people how to Google stuff.
19:30
Yeah. And then that was sort of like
19:31
the answer. And then you still get some
19:32
calls will trickle through. Now you sort of show them how to
19:34
Gemini stuff. Or whatever other AI I'm not
19:36
pushing Gemini, of course, of course, of course.
19:38
Yeah, yeah. Just what any any other thing
19:40
should. I put you talk about a search
19:41
now you sort of teach them how to how to query an AI.
19:43
No, sure. They don't need us.
19:44
Nobody. Calls me anymore.
19:45
They don't need anything. Taught everybody how to like Oh
19:48
this is what it feels like to be the buggy whip salesman guy.
19:51
Should we move on to my next one layup?
19:52
Absolutely. Architecture and construction
19:54
design. Oh, I.
19:55
Wouldn't call it a layup. It has become a layup in my mind
19:58
because I've just used it remodeling my son's bathroom and
20:01
basically photographing the space and inputting the
20:05
photographs and saying now how can I more usefully use this
20:08
space? OK.
20:08
And it created several potential layouts, one of which was the
20:12
one OK. I went with was this a custom
20:13
construction AI or just like like regular old Gemini?
20:16
No, this is regular old Gemini. This is bland, boring as it gets
20:21
really, and it laid out out the room.
20:23
Now I weeded through some some choices, but it laid out the
20:26
room. Beautiful.
20:26
This one's got a giant dolphin on the wall.
20:28
That's out what? What is this?
20:30
A fireman's pole? That's out.
20:32
Like what? Did you give?
20:33
It a did you give it a floor plan or just pictures just.
20:36
Pictures, Pictures. Wow, that's impressive.
20:39
Yeah. Did you give it any measurements
20:40
or did it just assume them or infer them?
20:42
I guess infer them. It inferred them exactly.
20:43
It inferred measure. Stud spacing or something like
20:45
that. I don't know exactly how it did
20:47
it. Whatever, somehow with the right
20:48
measurements. Oh, I always.
20:49
Should have been at your house. I would have totally screwed
20:51
with it. I would have bought like an
20:52
oversized size beach or like a like a basketball, it was three
20:55
times the size and put it in there so that it thinks your
20:57
entire bedroom is miniature. That's really cool.
21:00
Yeah. I did not expect you to say that
21:02
that. You would and all sorts of
21:03
things that are code now, yeah. I mean I took pictures of inside
21:07
the walls and everything when the walls were.
21:08
Open. Wait a minute.
21:09
What's Wait a minute? What is California code for
21:10
where I live? How is this supposed to be?
21:12
Yes, and it knows. So you're basically saying that
21:16
good old vanilla beige Gemini, Yeah, is is smarter than 98% of
21:21
the contractors out there? I can't speak for the
21:23
contractors, but I can tell you Jim and I had everything pretty
21:26
much exactly right. Wow, that, that is, yeah, pretty
21:29
impressive. That is cool.
21:30
Local Los Angeles building codes, there's all sorts of
21:33
requirements and things that you have to do, and you've got to
21:35
take your stuff into the plan check.
21:37
Yeah. But if you want the teeter
21:38
totter, if you want things to continue to go faster, you got
21:40
to fix the other side too, which is the plan check.
21:42
Plan check should be using AI for sure.
21:45
Yeah because humans are looking over that.
21:46
It's a ridiculous process. The reason that architecture is
21:49
so similar to accounting and coding is the rulebook.
21:52
Coding has syntax, accounting has gap and tax law.
21:55
Yeah, architecture has physics and building codes.
21:58
That's it's a very simple thing that you can see why we iterate
22:01
from one to the other. They're all very similar.
22:03
Sure. Your turn, Sir.
22:05
Let me do my next one here. This one is is I think probably
22:07
the most controversial that I'm willing to bring up in this
22:09
particular setting. Yes, this is not going to make a
22:11
lot of friends, but I think it's a really good idea.
22:13
Yeah, insurance and and specifically the the actuarial
22:17
science behind it. This is, I think something that
22:20
makes people, a lot of people queasy.
22:22
But this this could be something where you're paying truly
22:24
tailored insurance. Think if you could dial in the
22:28
exact amount of premium for the exact amount of risk that you
22:31
are creating. That's kind of fascinating.
22:33
That's really the way that it should be.
22:35
Now all the people that are high risk, of course are gonna hate
22:37
this idea because. They're gonna pay bar in
22:39
premium. Sure.
22:40
Yeah, Right now they're getting a free ride.
22:42
Or not a free ride, but a partially free ride.
22:44
Yeah. And all the people that are low
22:45
risk, I assume would love this idea because they're low risk.
22:47
But I love the idea of it being real time or near real time.
22:51
I think this is controversial because there's nothing about
22:53
this I don't, I don't like the idea of because there are areas
22:56
where I think I might be considered a higher risk bucket.
22:58
Sure. But like for example, we talked
22:59
on the show some time ago about the the folks that were buying
23:02
cars and part of their contract was offloading their driving
23:06
data to their insurer. Yeah, right.
23:08
That was built in as part of their contract.
23:09
This is the problem with having AI looking at insurance data.
23:13
Well, but now you've got you certainly have some let's set it
23:14
aside the privacy constraints for a second, right?
23:16
You have some real privacy risks, but the idea of being
23:18
able to dial in insurance, like if you are consistently driving
23:22
your radically and changing lanes without signaling and, you
23:25
know, speeding at 2X the speed limit, you are without a doubt a
23:29
higher risk. Sure, right.
23:30
Sure. And if you have some basic guy
23:31
that can detect if you're driving when you're intoxicated,
23:34
you are a higher risk. Yeah, you should be paying a
23:36
higher premium. That should be the case.
23:37
Or maybe not even driving. And if you're not doing those
23:40
things, you should pay a less premium.
23:41
To me, driving is the easy. One driving is the first one to
23:43
go away because autonomous driving, yes, exactly.
23:45
So nipping at its heels that it won't really matter.
23:47
But you know it's not going to go away.
23:49
My new GE monogram fridge that when I open the fridge and it
23:52
goes yes, hey chubby, we see you grab the cake at 2:00 AM.
23:56
Maybe put that back or your healthcare premiums are going up
23:59
by 8 bucks a month. Like you want that and you don't
24:02
want. So many privacy issues here.
24:04
I I right, I am grimacing at the.
24:06
Privacy. But let's let's just assume for
24:08
a moment we can solve the privacy issue.
24:10
Let's assume for a moment that you can make this all run
24:12
locally, which is feasible. Huge jump.
24:14
I don't think it's that huge of a jump.
24:15
You could run everything locally.
24:17
You could have everything so that it's not putting back any
24:19
PII. Like let's just assume for a
24:21
second there's a path. Is this a world we want to live
24:23
in? And I'll tell you honestly, I
24:25
don't know, 'cause part of me is like, yes, absolutely.
24:27
And part of me is like, this is a terrible idea.
24:30
Terrible. And that's what be setting the
24:31
the privacy constraints aside. But the current version is, you
24:35
know, one size fits most and that's not a better solution,
24:38
so. Much of the hey fatty, you
24:40
probably shouldn't have that piece of cake at 3:00 AM in the
24:42
morning. And and you can say, look, my
24:44
genetics can handle it. No, it would know that your
24:46
genetics can't. No, no, I'm saying specifically
24:48
the one that would talk to me would say, hey fatty, you can't
24:50
have the. Cake, I know your.
24:51
Genetics can handle. They would just like, Sir, would
24:53
you like more cake? That's exactly what it should
24:54
do, right? It should know these things.
24:56
That's the whole point of the customization.
24:57
That's a bigger problem. If it starts assessing premiums
25:00
based on what it knows about your genetics, that's even even
25:02
bigger issue. Isn't that the whole freaking
25:04
point again, set the privacy. It's a concern because of
25:07
privacy. Again, I'm really torn on this.
25:09
I feel like I'm I'm running down the path of the Mobius strip and
25:12
I keep going from the one side of this is a good idea.
25:15
It's the other side of this is a terrible idea, but I'm on the
25:17
same side the entire time like it's not.
25:19
I really don't know the answer to it because it really should
25:22
be dialed in. The amount of risk that you're
25:23
creating. That is the right answer.
25:25
But here's the there's one thing I know for sure.
25:27
There's not much I'm confident of one thing.
25:29
I'm confident of this. This this is not going to go
25:30
away. Yeah.
25:31
You have this tool. We now have to figure out how to
25:33
wrestle with it. Yeah.
25:34
So we've got to solve this and find some middle ground where I
25:37
don't think you're going to find happiness.
25:38
I think you're going to find a, a tighter pendulum swing.
25:40
You're just going to have like, you're just going to be less bad
25:43
and less good. Just kind of like, OK, this
25:45
feels like this is probably the middle.
25:46
What else you got? Let's so my last one is going to
25:50
offend everyone I know in my neighborhood and that is I think
25:53
acting is basically going away. Oh, yeah, yeah, yeah, it's a
25:55
good one. OK, well, you say the best ones
25:57
for last. Yeah, the.
25:58
The ones, the actors who are brand names, the ones who bring
26:01
the sell tickets, sure, right. Those people are still good
26:03
because that's a, that's a viable brand.
26:05
You're actually selling tickets. You're opening a movie or
26:06
whatever you're doing. Everybody else, your part can be
26:09
created by an AI. And yeah, it's not going to ask
26:12
for a break to go to the bathroom where it's not going to
26:14
arrive late on the set because it hit traffic and sure, all
26:17
those things are going to just evaporate.
26:19
Sorry, there was there was a lot of data going between LAX 01 and
26:22
SO 2 on the fiber, so I'm a little late to show up for the
26:26
casting. So I'm a now a second late.
26:27
I OK, so first of all, this is awesome.
26:30
The here's the really good news from from a acting perspective.
26:34
Yes, this gives real opportunity for acting to all of the ugly
26:39
people, because now all of the ugly people can make their own.
26:42
Get a job too. Virtual.
26:44
No. Well, that's just I just
26:46
neutered that one now, didn't it?
26:48
Because now they can make you don't.
26:50
You're not going to have a finite set of actors.
26:51
I would argue if anything with AI, you're going to get more
26:54
actors, more virtual actors. And now those, those can be
26:57
creative. Yeah.
26:57
The fascinating part to me is you think about like, what is
27:00
acting? And now you can create that.
27:02
And that, to me, is kind of amazing.
27:04
Yeah, it's not that films are going away.
27:05
It's not that entertainment is going away.
27:07
It's just how we make it is completely changed.
27:09
Yes, there's a big sag after a negotiation 2026 that.
27:12
Yeah, that was concerned about likeness and rights issues.
27:15
Yeah, that's not gonna matter for those people who aren't
27:17
brand names, those people whose faces haven't been out there who
27:20
can open a movie. I would argue even for the the
27:23
people that have brands and our brand names, the people that are
27:26
brand names, they're gonna have short term value in that.
27:29
That's not a long term thing. I agree with you.
27:31
It's different for them. Well, sure, but it's.
27:32
Still, that is a short term, like long tail.
27:34
These are your options as a as a human actor.
27:37
You can either profit from your likeness or you cannot.
27:40
That's it. There's no need to.
27:41
But that's that's my. Why?
27:43
Why do I spend 1 cent on you? If we're not a known value, who
27:46
sells tickets? I can create a brand new human
27:49
from the AI, sure. I don't need to have your like
27:51
because I can just have some some other person created out of
27:54
whole. Class, well, because maybe we
27:56
need to shift to some of the value to the actors being in
27:59
real life and what they do in real life.
28:01
Because the one thing you don't have from a virtual AI actor is
28:04
an in real life version of it. So your likeness does.
28:07
It's harder to put them on The Tonight Show to promote your
28:09
movie for sure. Now in The Tonight Show's all AI
28:12
like that, I disagree with you there, but like for them to be
28:14
out and actually meet a human. But then you get into this whole
28:17
virtual influencer, which I am fascinated by, like these
28:20
influencers that are completely 100 percent virtual.
28:23
They're AI created. And then they go out, quote, UN
28:25
quote, into the real world and take pictures with like, here I
28:28
am with Mike Tyson at a Super Bowl party.
28:30
And you're like, wait a minute, one of those is real, and one of
28:32
those isn't. Yeah.
28:33
And I believe it to be a real image.
28:35
They've just, you know, superimposed the AI character.
28:38
I'm not even sure it's a real image.
28:39
Anymore, some of those I've seen I believe all.
28:40
Can you tell I see tons of them now that are completely
28:43
convincing. Yeah, I mean, it's it's, it's a
28:45
very interesting world we're going into.
28:47
My only point is I still think there is some value in the human
28:50
AI actor over a completely pathetic one.
28:53
But we're not, we're not making that valuable.
28:55
We have to see what value that has.
28:57
And it might not have as much as it we think it has previously or
28:59
as much as it has now, but that's it.
29:01
Your choice is you either get some value of your like this or
29:03
you don't. Yeah.
29:03
And now there's a whole new class of people that get to make
29:06
AI actors that that did not exist a couple years ago.
29:09
Pakistan and India both are creating feature films from AI
29:13
fully generated feature films. Yeah.
29:16
Now the first ones might be terrible, the 1st 20 might be
29:19
terrible, but eventually somebody's going to figure out
29:21
how I lean into the AI and do things that I, yeah, could never
29:24
afford to do. Totally.
29:25
If I receive and I had a tentpole budget, I couldn't do
29:28
this. Couldn't do.
29:28
This is more. Possible things that I can
29:31
create in some way, yeah, that I can create super, super cheaply
29:34
by just requesting this in the AI from prompting the AI to give
29:37
me this shot. And here's the best part of
29:39
that. They can make that entire
29:40
production right tip to tail, and they don't have to close the
29:43
streets in your neighborhood. Once.
29:44
I love that. Some number of people in this
29:46
industry here are going to embrace this, in addition to all
29:48
the people that aren't in this industry, right?
29:49
But some of the people here are going to embrace it, and those
29:51
are going to be the ones that are going to win.
29:52
I think you might have an advantage if you aren't in this
29:55
industry. I think you might have an
29:56
advantage of her coming. In some ways, because you don't.
29:58
You're not held back by the legacy thought.
30:00
In some ways, I think you're right.
30:01
Yeah, exactly. You're you're thinking from a
30:03
whole like the position of what it could be, not what it has
30:06
been. Yeah.
30:06
Yeah, I agree with that. I like this one.
30:08
Acting. Acting is a good one.
30:09
All right, that's my non layup. What's your last one My?
30:11
Last one, maybe as controversial to me as insurance, maybe not,
30:16
but I'll just I'll just throw this in here and this is 1 where
30:18
I don't know what I want with the insurance side to happen.
30:20
This one, don't you want to say it?
30:22
But I'm going to say it. Cooking.
30:24
Yeah. Food science.
30:25
Yeah. I still don't know exactly where
30:27
AI is going to find the most value in the cooking chain, but
30:31
I already read somewhere I read some some article about an AI
30:35
discovering that there are certain molecules in blue cheese
30:38
and chocolate that were identical and they led to this
30:41
this fusion dish that sounded absolutely disgusting.
30:45
But then you taste. It's probably delicious, so.
30:46
Amazing. So that kind of stuff.
30:48
I don't see a world where that's not gonna happen, right?
30:50
That's 100% gonna. Happen.
30:52
I'm always amazed when I go to these molecular gastronomy Yep,
30:55
tastings. And there's just things that
30:57
sound absolutely disgusting and many of them are absolutely
31:00
disgusting, but there's always a few that it's like, who in the
31:03
world ever thought that up? That combination of flavors is
31:05
fantastic. Absolutely.
31:07
Anyway, I'm excited to see what happens there.
31:09
But but I well, I I very much like you're you're acting
31:12
entertainment. I think cooking is into the
31:13
other one. That's going to be super
31:15
interesting. I'm excited about.
31:16
That I'm excited about the whole thing.
31:17
Every aspect of this we've discussed.
31:19
It will cause enormous disruption.
31:21
It will be very hard. A lot of people, we are going to
31:23
go through a seismic change in how people work.
31:27
And you know what it's still going to be.
31:28
It's really cool. Which one are you most excited
31:30
about have of of change coming? Which one are you the least
31:33
excited about change? I am the most excited about
31:35
seeing movies that are made entirely on AI from someone who
31:39
spends a couple of things of dollars with you and makes a
31:41
tadpole movie. I am the least excited about
31:43
seeing movies that are made by AI, and most of which are going
31:47
to be just dreadful. Really.
31:49
Oh yeah, there's just gonna be some awful, awful acting.
31:52
Oh, but don't watch. Those they I will tell you not
31:53
to watch those. Right, you guys, you're trying
31:55
to watch, by the way, movies are #6 don't do that, right?
31:58
People make really bad Movies Now.
32:00
And yeah, you, you have to sort through and find the good ones
32:03
too. But I think that's probably my,
32:04
the one that I think is gonna be the most interesting.
32:06
Oh, that's cute. It's gonna have both sides.
32:08
I think I'm the most excited about the the creative explosion
32:11
that I, I'm up the Mystic we're going to see with AI tools
32:14
around what would call acting or entertainment.
32:16
I'm actually very excited about that.
32:17
Yeah. Which one am I least excited
32:18
about? That's the question, man.
32:19
I don't know. It's probably some somewhere
32:21
between insurance and cooking and I don't, I have no good
32:24
reason to tell you why. I guess I'm that's kind of a
32:26
lie. I'm excited about all of these.
32:28
I think these are all actually very excited.
32:29
We picked ones that were exciting.
32:30
That's true. All right.
32:31
We have to get out here quickly before we do.
32:33
Have you seen or read anything good this last week?
32:35
Well, as you well know and I think most listeners do, I am an
32:38
avid listener of Ben and David at the acquired podcast.
32:40
Of course, like most the content that I enjoy, I try really hard
32:44
not to binge it. I I like to spread my content
32:47
without. I don't want to RIP through
32:48
everything and then just hit the end.
32:49
Yeah, that's not the way my brain works.
32:50
So I usually reserve their episodes for long road trips.
32:54
And last week, on the drive to Yosemite, I listened to their
32:56
Formula One episode. Oh, yeah.
32:58
It is a master class in business strategy.
33:01
It's a fantastic story. Sure.
33:03
It tracks F1's wild evolution from a deadly money losing
33:05
European racing series into a $30 billion global powerhouse.
33:10
Yeah. And whether you're a fan of the
33:11
racing or just interested in how Liberty Media used Netflix to
33:14
crack the US market and turn every team into a $1 billion
33:17
plus asset, yeah, the deep dive is essential.
33:19
Listening now, I'm a car fan. I'm a racing fan.
33:21
I went to Formula One races when I was a kid.
33:23
I've always loved it. You do not have to be a fan of
33:25
cars or racing to enjoy this episode.
33:27
It's really part like engineering World Cup, which the
33:30
engineering in F1 is amazing, like insane.
33:32
Second only maybe to NASA and the the ESA.
33:36
It's part what they say, Real Housewives of the Garage, which
33:39
I think is. Yeah, yeah, sure.
33:40
Super cute and the whole thing is just entirely fast.
33:42
So I think folks listen to our show would love this.
33:44
It's acquired Formula One episode.
33:45
This is from March of this year 2026 and I suggest you guys all
33:49
goes into it wherever you get your podcasts.
33:50
Awesome, awesome. How about you?
33:51
Death by Lightning, a four part high stakes historical thriller
33:54
detailing the assassination of President James A Garfield.
33:58
Sounds boring? Not even close.
33:59
Starring Michael Shannon is the intellectually brilliant
34:01
Garfield and Matthew McFadden as his deluded killer.
34:04
The series exposes a nation at a crossroads.
34:07
Garfield's unwavering integrity stood against the tide of
34:10
widespread graft and corruption personified by political Titans
34:13
like Roscoe Conkling. While Garfield sought to reform
34:16
a broken system, he was ultimately betrayed by medical
34:19
arrogance. Doctors who ignored germ theory
34:22
and turned a survivable wound fatal.
34:24
Oh boy. Featuring Nick Offerman and
34:25
Bradley Whitford, the story proves that America moves
34:28
fastest under leaders of deep intelligence.
34:30
When such light is extinguished by chaos and greed, an entire
34:33
nation is left in the dark. Amazing to me that this really
34:37
wasn't that long ago and you had a president of the United States
34:41
die of. Effectively infection.
34:44
Yeah, the bullet would not have killed him.
34:45
You could have left the bullet there and he would have been
34:47
fine. Doctors and I use that term
34:49
extremely loosely, who were just like, I'll put more leeches on
34:52
him, he'll be fine. It wasn't quite that, but, I
34:54
mean, the doctors even says, you know, I don't believe in ghosts.
34:56
There's no such thing as germs that are infecting him and all
34:59
this. That's ridiculous.
35:00
Yeah. Well, that's when germs run.
35:01
The ghost bucket hit the space bar.
35:02
It turns out I don't believe in ghosts.
35:03
But I do believe in germs. I have been to a daycare and let
35:06
me tell you, I believe in germs. Germs are real.
35:09
Death by Lightning on Netflix. Now I want to check this out.
35:12
I think this this sounds way more interesting than it's so
35:15
good. I think a lot of people give a
35:16
credit. That is the episode.
35:18
Thank you for joining us for all of this nonsense.
35:19
It's really terrible podcast from our awful company.
35:22
This is on the web at nonsense dot productions.
35:24
I'm I'm CJ Little. I'm Jeff.
35:25
Parker. If you like this program, please
35:26
follow, download, subscribe and like at Apple Podcast, YouTube
35:29
Music, Amazon Music, Antenna Pod, iHeartRadio, Spotify, My
35:33
Favorite Overcast, wherever you may get your podcasts from.
35:35
Nonsense uses the Internet Archive, A nonprofit digital
35:38
library offering free, universal access to digitized collections
35:41
of books, movies, software, music, websites, and more.
35:44
Through initiatives like the Wayback Machine, users can
35:46
explore snapshots of over 866 billion web pages, providing
35:50
invaluable insight into the evolution of the Internet and
35:53
human knowledge. The Internet Archive.
35:55
If you haven't already, check it out.
35:56
We'll be here every. Tuesday and Thursday mornings
35:58
For more nonsense, please join us.
36:12
Right now, if I brought an autonomous robot over to your
36:14
house, what would you want it to do?
36:16
But I don't know, I struggle with this.
36:17
What would I have the robot do? I I wouldn't, I'm sure I
36:19
wouldn't have it cook and wanted to do other things that I could
36:21
cook. I would say maybe 3 or 4 times
36:23
out of seven during the week I would have it cooked for me.
36:26
Like I enjoy it, but it's the difference between it having to
36:28
be utility and fun. And I'd say maybe 3% of the time
36:31
I can still have good fun and have the utility the other
36:33
times. When you don't want to cook on
36:35
those nights, you know we have these things called restaurants.
36:37
Go on. Tell me more.
36:43
Tell me more what? What future do you live in?