Latest / Bad Decisions Podcast / Trump Releases UFO Files 😮
Transcript
- 0:00If you believe that aliens are real, today may just be your
- 0:03lucky day because the US government just declassified
- 0:07official UFO sightings. Two robots can now fully clean a
- 0:14bedroom autonomously. That means you don't have to
- 0:17worry about cleaning your room anymore.
- 0:20And we also end today's show with a conversation with the
- 0:23founder of Polycam about the future of AI and its impact on
- 0:28our lives. Farhad, are we ready to do this?
- 0:32Welcome back, everybody. It's Friday and we are back with
- 0:35another show. We have very exciting topics,
- 0:39some of them just released a few hours before this show.
- 0:41So you're very hot. It is very hot.
- 0:44As always, we're bringing you guys the latest and the greatest
- 0:47in the world of technology Tuesdays and Fridays.
- 0:51Don't miss it. Now for everybody had who has
- 0:55been watching and listening, we just wanted to say thank you for
- 0:58supporting the show up until today.
- 1:00You guys are amazing. We read your comments, we see
- 1:03your likes and we see the love and we send it right back.
- 1:06Now before we begin, if you enjoyed the show in the previous
- 1:10episodes, don't forget to subscribe.
- 1:12That means a lot to us. Let's get it.
- 1:15Let's go. My hat.
- 1:17I'm sorry you did it. Why did you take it away?
- 1:23Never trust. That's the sign that no, no,
- 1:26that's that's a sign that we should never trust you.
- 1:30You brought up a fist bump. I never like.
- 1:33What was it then? Was it a fist bump?
- 1:35We can put a screen of that. People saw it.
- 1:38You put up a fist bump and then you took it away.
- 1:40I shouldn't have actually fist bumped you.
- 1:42That would have been funny. It's OK, you can always do it to
- 1:44yourself, but we got some important stuff we got.
- 1:47Go ahead. No fist bump for you anymore.
- 1:50All right, High 5. Let's do this Today. the US
- 1:55government did something that it has never done before.
- 1:59It opened a public website, war.gov/UFO, and uploaded 162
- 2:09files. Photographs from the surface of
- 2:13the moon showing strange objects.
- 2:16Audio of astronauts on Apollo 17 describing very bright particles
- 2:24drifting past the spacecraft. An infrared video from a Navy
- 2:29sensor over the agency in Greece showing an object making 90°
- 2:35turns above the water at 80 miles an hour.
- 2:39I mean, look at how the object is moving.
- 2:42That is impossible for any aircraft that we know of today
- 2:47at that speed and those turns. It has not been seen before.
- 2:53These all of these documents that I just mentioned are never
- 2:57seen before. Documents now declassified and
- 3:01shared publicly as part of the pursue directive from the
- 3:06President. Photos, scanned documents,
- 3:10videos spanning across 80 years of UFO sightings from as early
- 3:18as 1945 during World War 2 to incidents as late as April 2025.
- 3:27Now, this is not the end. There are more batches of these
- 3:32documents set to be released every few weeks.
- 3:37Farhad, this is perhaps a day that many alien activists.
- 3:47Alien. UFOs enthusiasts have been
- 3:51waiting for because we've always had conversations around aliens
- 3:56and UFOs. Actually, Joe Rogan took his
- 3:58promise during the podcast right that Trump has to release this
- 4:01document before his election presidency.
- 4:04You mean before he was elected as president?
- 4:07Joe Rogan asked him for this during the podcast.
- 4:11That's why I did mention that this is part of the Pursue
- 4:14directive. This was an initiative to allow
- 4:19for transparency between the government and the people,
- 4:23mostly for American public, but of course, everybody has access
- 4:26to the website. The government here is not just
- 4:30the president. We're talking about the state
- 4:33government. We're talking about the
- 4:34president. The FBI, even NASA is involved.
- 4:39They've shared documents, official documents, but this is
- 4:42a an insanely large initiative that has just been released.
- 4:49By the way, this is as early as today as we're recording this
- 4:53and we're uploading this show today.
- 4:54So this is very hot right from the oven.
- 4:58We're giving you this. It's an insanely long article
- 5:01here on X. You can check it out straight
- 5:03from Department of War. They have the link to the
- 5:06website. When you go into the website, as
- 5:09Pratt will help us view right now, you'll see that the entire
- 5:13description of what Trump's initiative is all about is
- 5:16there, the directive. And then below, if you go all
- 5:20the way down, you'll get access to all of these files.
- 5:23There are 17 pages right here. All of them.
- 5:27Everything. PDFs, scanned documents, videos,
- 5:30photos, audio recordings, and they all have dates and
- 5:36locations. You are not just exclusively
- 5:40looking at files from the United States.
- 5:42We have files from the moon, We have files from the Middle East,
- 5:45from South America, from Japan, from Greece.
- 5:49Different people have reported these over the past 80 years.
- 5:55So you may well be asking yourself, how do I trust these
- 5:59files? Well, one of the first questions
- 6:01we asked was the same. And so we started looking into
- 6:04the files with Claude. We asked Claude to go and look
- 6:07through the documents. We didn't have enough time to
- 6:09look through all of this. Some of the PDF files, can you
- 6:12just quickly go inside? Some of these PDF files are 100
- 6:16plus pages. Oh shit, these are reports right
- 6:19from, I don't know, like Army and people in the field
- 6:22probably. That's a great question.
- 6:25We asked Claude that because I wouldn't trust it if it's just
- 6:29straight from the government only.
- 6:31But it involves NASA astronauts, U.S.
- 6:35Navy and Air Force pilots, FBI special agents, scientists,
- 6:41foreign government, intelligence services and civilians.
- 6:45So yes, many of them can be associated with the government,
- 6:53the roles and the responsibilities.
- 6:55So people who would be incredibly skeptical can say,
- 6:59oh, this could all be paid, You know, all of these people could
- 7:03have been. Paid.
- 7:03Why would you pay someone to give this reports?
- 7:07I've no idea. Yeah, no.
- 7:09But you could to make up stories.
- 7:11Oh, yeah, of course you can. It's been done before many
- 7:14times. But the the important thing is
- 7:18to understand that there's different demographics of people
- 7:22making these reportings. It's not just FBI agents or just
- 7:28Air Force pilots. We've got different demographics
- 7:30of people making these reports. And it really feels like I'm
- 7:36looking at artifacts in a Call of Duty World War 2 game, you
- 7:40know it. It feels almost unreal that as
- 7:44civilians we get access to these classified documents that I
- 7:48don't think in my lifetime I would have even seen.
- 7:51These are, again, you can see the stamps on all of them.
- 7:54They're all classified secret documents released incredibly
- 7:59publicly. The entire world is going to see
- 8:00this. Joe Rogan's going to be the
- 8:02first person covering this. It's to me do we have photos and
- 8:07videos as well? Everything.
- 8:09Yeah, we have everything. Can you pay attention to this
- 8:12moment? I found out you didn't realize
- 8:13this. Can you play this?
- 8:14This is at 80 mph. Look at the way that bright
- 8:18light starts moving. It's all over the sea.
- 8:21Look at that. Did you see how it just turned?
- 8:25It's fully 180° turn. Up look, we don't have anything
- 8:32that looks like that. So these are all objects or they
- 8:35have seen species as well? OK, great question.
- 8:40Are these aliens? You probably are asking that
- 8:45question. I did lie to you.
- 8:47I have to be honest. I did start by saying UFOs and
- 8:51the only reason I said that is because most of us are familiar
- 8:54with the pop culture trend and name of UFOs.
- 8:58And what does UFO stand for? Unidentified flying objects.
- 9:03Problem is most people associate that with flying objects because
- 9:07that's what it's called. The saucers that we saw in in
- 9:10films or little green alien men creatures, yes.
- 9:16What scientists and the US government have changed with
- 9:20regards to these reports over the years and specifically this
- 9:25release, is that they're no longer calling it UFOs.
- 9:29They've switched a name out with UAP, which stands for
- 9:33unidentified anomalous phenomena.
- 9:36Anomalous means deviating from the normal.
- 9:40Oh, so it can be an object or it can be a species as well?
- 9:43It is. It is anything that we cannot
- 9:47identify. So imagine it could be a light
- 9:49source or an object that is not human.
- 9:53It's just something we cannot identify.
- 9:55It's not normal. It's abnormal.
- 9:57We've never seen anything like it before.
- 10:00And this is the interesting part. the US government official
- 10:04position here is not that these are aliens.
- 10:07They're just saying we don't know what they are.
- 10:10These are unresolved cases specifically because even their
- 10:14own personal can't tell what they are.
- 10:17But there hasn't been any harm to any human in all of these
- 10:22documents. Or has there been?
- 10:24That's a great question. Do we have, can you ask?
- 10:27Can you, Pratt is talking to Claude, Can you pull up Claude?
- 10:30Is there Pratt, can you do me a favor?
- 10:32Do we have anything sensitive to show there or we can pull that
- 10:35out? I would love for people to see
- 10:36how we're using Claude Co work. No, no, no.
- 10:39We can pull it up. Let me pull it up.
- 10:41As long as we don't have anything sensitive there, I just
- 10:44have to make sure. So for anyone, can you describe
- 10:47what cloud cohort is for everyone?
- 10:49So in the cloud desktop app you have 3 features, Chat, cohort
- 10:56and code. Chat is your LLM where you talk
- 10:59to cloud different models. Code is where you code.
- 11:02Cohort is the agentic interface where you can ask it to control
- 11:09your computer, to go to your browser, do tasks for you.
- 11:12It's like open claw. Exactly.
- 11:15And the reason why we use it is because right before the show we
- 11:22became aware of this news piece. As I mentioned earlier, there's
- 11:26thousands of pages of documents and I wanted to make sure we're
- 11:29giving you the right information.
- 11:31So we wrote out this long prompt and asked it to go ahead and
- 11:36analyze the news, and it did so very well.
- 11:39It even came back to us with the report as if I spoke with a
- 11:42journalist to do the job for me. So right now, we're going to ask
- 11:46you one more question. Farah, what was the question?
- 11:48Has there been any damage to humans?
- 11:51Wait, wait. No, no.
- 11:52Say that again. Oh, wow.
- 11:54OK, sorry. Can we do it again?
- 11:55Sorry, I got so excited. OK, it's I thought, I thought
- 12:00price is going to type. OK, 321, has there been any
- 12:04damage to humans or any other assets by UFOs or Uaps?
- 12:13Can I add a little bit more? No, no, no.
- 12:15That's my question. Send it, send it, send it OK.
- 12:18We're going to wait for this to finish.
- 12:21In the meantime, what do you think about this whole thing?
- 12:26The fact that they have released all this document, I think is I,
- 12:29I told you, is part of the promise that Trump made in the
- 12:33beginning of the campaign. But do you really think they
- 12:35have released everything? Of course not, no.
- 12:38What do you mean they released everything?
- 12:40Do you know how much files they have?
- 12:43It's impossible to release everything.
- 12:46It's not possible to release everything or make that full
- 12:50screen or something. Just passed by, right?
- 12:53Play it again. That looks like a rocket.
- 13:00It looks like a rocket. Now, this is the interesting
- 13:03thing. I want to be very clear about
- 13:04this because I want to make sure we don't think of this like
- 13:10anybody who's just a UFO enthusiast and say, yes, aliens
- 13:13are real. You have to understand
- 13:15something. Even the US government has
- 13:18stated this, that any of these objects could very well be a
- 13:22drone, balloon, a natural phenomenon that they couldn't
- 13:26understand or properly visualize, a foreign tech,
- 13:32Chinese drone, or in rare cases, something truly unknown like an
- 13:38alien. So we don't know.
- 13:39Again, the whole point here is that these are unidentified
- 13:43objects. Let me rephrase my question.
- 13:48Do you think aliens are real? Yes.
- 13:51You. Yes, but they do not be in the
- 13:56form of green creatures or blue which which?
- 13:59Green creature you've seen as an alien.
- 14:01Grey, green, blue, Just aliens in ET.
- 14:04The first the first Alien movie I watched was ET actually.
- 14:08Why can't we have 4K8K footages with red cameras?
- 14:12Why are they all horrible quality?
- 14:15Right It's it's impossible. But some of these are infrared.
- 14:19Many of them are, I think also from fighter jets right when
- 14:22they are flying and doing patients.
- 14:24Install some RED cameras on these fighter jets.
- 14:27Did cloud cohort came back? No, it's still.
- 14:30Wow. Still.
- 14:31Doing the research, how much document is out there?
- 14:33There's there's tons. Like I said, 17 pages here.
- 14:38In each of these you have about 10.
- 14:42Rows right now, so that's 170. Rows.
- 14:46Each of them can have hundreds of pages.
- 14:49There's tons of documents in there.
- 14:51I recommend that you guys check them out yourself.
- 14:53I think it's pretty interesting that we get access to this.
- 14:57But I always like to ask the question why?
- 15:00And I think you should do the same, too.
- 15:02I do that with content. When I see people release
- 15:06content on social media, and I think the content has a meaning
- 15:09behind it. I always ask myself why this
- 15:11creator, this company has released such kind of content.
- 15:15Why has this been film been made?
- 15:18And I think this is actually one of those cases where you
- 15:20definitely have to ask the question, why would the US
- 15:22government suddenly release all of these documents?
- 15:26I'm not saying that. Oh, they have a horrible
- 15:31intention with it, but many presidents could have done the
- 15:35same. Alex Jones has been talking
- 15:37about it for ages. Yeah, I know, you said Trump
- 15:40promised. But he could have also not done
- 15:42it. Actually, why now?
- 15:45Out of all the time, no. I don't think, I don't think
- 15:48anything would have happened to the government if they didn't
- 15:50release these. That's true, actually.
- 15:52Why would they release them like that?
- 15:54I mean and by the way, the website is so well done if you
- 15:57actually. The new administration websites
- 16:00are generally all of them are good.
- 16:02I think they got a really good design team.
- 16:03OK, we got the response back. I cannot read it from here, but
- 16:07is it yes or no? The headline answer is yes,
- 16:11they're document cases, but the record is uneven and you should
- 16:16grade them by tier on camera. Oh, so there has been damages?
- 16:21But go to the tier one. Yeah, there's a Tier 1, Tier 2,
- 16:24tier 3. Just read just overly.
- 16:26What is it saying? Has there been damages or
- 16:29injuries or deaths? No, it doesn't mention that.
- 16:37So they're just flying around observing. 3 civilians,
- 16:42including a 7 year old, suffered severe burns.
- 16:46UAP encounter. There's a lot, there's a lot of
- 16:52it, by the way, that it was not caught on video nor are there
- 16:56photos of it. For instance, there's one I
- 16:58wrote about a super hot orb at AUS military base out running a
- 17:03helicopter described to the FBI by a senior U.S. intelligence
- 17:07officer. Oh, so just a single?
- 17:09Even there was another count the size of half of a football
- 17:12field. But these are all reported by a
- 17:16single individual. No, no, no.
- 17:17They're witnesses. Cross Cross.
- 17:20Reference with multiple people actually, could you?
- 17:23Is there anything you're trying to show us here?
- 17:25No, no, it's good. So no official documents go up
- 17:29here. There are the general photos.
- 17:31There's a render made on the reporting that was done.
- 17:35Look at the size of the orb, zoom in.
- 17:37Yes, go to the right, Go to the right.
- 17:39Right, right, right, right. More all All jet cameras.
- 17:44Yeah, look at this one. This is based on the reporting
- 17:47of multiple people. This is not a real Photo by the
- 17:50way. They did a render to try to
- 17:52imagine what multiple people described to be.
- 17:56At the farm they saw this. I don't think it's.
- 17:58A farm, I don't know where it is.
- 17:59Pride. You can probably read the text
- 18:01below it for us. No, it doesn't say where, but
- 18:07it's like a bright Epsiloy bronze metallic object metallic
- 18:11realizing out of a bright light in the sky, 130 to 195 feet in
- 18:16length and disappearing instantaneously.
- 18:20Wow. It's very interesting.
- 18:23It's very interesting. What do you guys think man?
- 18:27Yeah. I do, but also another one that
- 18:29blew my mind is Apollo 12. What happened with Apollo 12?
- 18:33These are the astronauts. That they saw something in moon.
- 18:37I mean not sure they are in moon so.
- 18:39If, for instance, there are astronomers or physicists, I
- 18:46don't know what area astrophysicists, thank you,
- 18:50astrophysicists who can come and describe exactly what these are
- 18:53and say, Oh no, these are specific planets or stars and
- 18:56they are things that we can identify, then I'm like, OK,
- 19:00cool. But if astrophysicists say what
- 19:04are these? We don't know these exist.
- 19:05But we are deep in the space. Of course we don't know what
- 19:08they are. I mean, the one on the moon,
- 19:09this is from the moon, right? This is from the moon.
- 19:12This is from the moon. Does it look like planet Earth
- 19:15to you? This is from the moon.
- 19:17The first question is, did we go?
- 19:18To the moon Did we go to the moon Now you are.
- 19:21Releasing a document from the moon.
- 19:23Did we go to the moon? Did we go to the moon man?
- 19:26We are asking guys some deep questions.
- 19:29So what do you guys think about aliens?
- 19:31Pratt, Aliens, real or Not? I don't know man, I will say yes
- 19:35for now. Let us know what you guys think
- 19:39in the comments. This could be an ongoing
- 19:41discussion. Considering that they'll be
- 19:43releasing more batches, you bet your ass we're going to be
- 19:46covering it right here on this show that is about technology
- 19:49and also aliens. Oh.
- 19:51We are #1 showing tech, alien, and AI.
- 19:54Now you're going to cover Alien too?
- 19:57We can move on for now. The next topic for us is also
- 20:01quite interesting. Two robots learned to fully
- 20:06clean a bedroom autonomously together.
- 20:10So imagine that if you are tired of making your bed or cleaning
- 20:15up your bedroom, robots can do it for you in the future.
- 20:18That includes pushing furniture, hanging clothes, taking out
- 20:23trash and working together to make a bed.
- 20:27This is a real big step because these two robots were not
- 20:32sharing plans, they were not messaging each other, and there
- 20:36was no central Control Center. Oh wait, so there's no way they
- 20:39can talk to? Each other?
- 20:40No, they would just look at each other, react in real time based
- 20:46on each other's movement. Holy shit, first time that two
- 20:51robots can do it without essential control unit.
- 20:55See one of them just staring at the other one and they pause
- 20:58for. A moment and they you're telling
- 21:00me wait, they're not. They are not.
- 21:02They are not sending messages to each other.
- 21:04They are just looking at each other and reacting in real time
- 21:08based on the other robots action.
- 21:10And this is extremely difficult, especially fixing that blanket,
- 21:15because the blanket itself, the comforter doesn't have a rigid
- 21:18side. So there is no way that they can
- 21:20identify where is their half to tie it and to clean it.
- 21:24Look at it. It's crazy.
- 21:26So these are the robots. Pratt can't do that, but it's
- 21:28bad. These robots are called Helix
- 21:320.2 from Figure. If you remember a few months
- 21:36ago, there was a video that these robots were taking grocery
- 21:40out of the bag and put it in the fridge.
- 21:42Pratt, can we pull that up? This video went viral.
- 21:45Do you remember that video? No, that they were taking
- 21:47groceries and putting that in the fridge.
- 21:49They were working together. Yeah, let's see.
- 21:54I showed you this video. No, I remember 1 robot handling
- 21:59a bunch of different items but not this one.
- 22:01But so it was this one. Yeah, So it was this.
- 22:03One, you can just play it. These are Helix 01.
- 22:07This was back in February and now Helix 02.
- 22:12First of all, can I just say great fucking design?
- 22:15Yeah. Those robots that look
- 22:16beautiful, they're a gorgeous one.
- 22:21They're the kind that could definitely come and haunt you
- 22:24down, but go go back to that square icon.
- 22:28Loading could be the last thing you see.
- 22:30Before you die. You're talking about aliens and
- 22:34robots, but try go back to the latest video that they made.
- 22:38So it's. So cool.
- 22:40They they interact to each other like humans.
- 22:42They're nodding, they're looking at each other.
- 22:44It's like, oh, oh, give it to me a bit carefully.
- 22:47You fucked up. Yeah, you can.
- 22:49You can. Almost feel like they're talking
- 22:50to each other. These are really good.
- 22:53Can you can you go back to the blanket?
- 22:55The blanket 1 is the best No, one of the hardest thing that
- 22:58they've done Go go a bit go halfway through the video yes
- 23:02wait, wait for the robots to open the trash can.
- 23:05It balances on one feet, so it opens the trash can put the
- 23:09trash inside and close it. So they are fully cleaning the
- 23:12room by the way, it's under 2 minutes under 2 minutes.
- 23:16They clean everything, they hang the clothes, they put the
- 23:18furniture back and they make the bed.
- 23:20The best part is making the bed. Just look at it man.
- 23:25And but that shows we need two of them, not one.
- 23:28I think 1 is not enough. I will buy.
- 23:31Figure I will buy. Figure immediately if you, if
- 23:34you are a creative, imaginative person, please take a moment to
- 23:39really think about what our future will look like.
- 23:42This is going to be our lives in five years time.
- 23:47Yeah, I will give it five years. You all have robots.
- 23:50You'll go to a store or perhaps better, you just order one on
- 23:53Amazon. It'll come to your doorstep and
- 23:55maybe even fucking walks to your doorstep because it can.
- 23:58It walks to your doorstep, it knocks on the doors like hello,
- 24:01I'm your robot and it can cook for you, it can clean for you,
- 24:05It can massage you. I'm thinking, I was watching
- 24:08this and I thought if you have two robots, you can literally
- 24:10come out and say I'm really tired, massage me, and it just
- 24:13comes behind you and starts rubbing your back.
- 24:16That's what you would want. I want to clean the room.
- 24:19Big day to fucking massage your back.
- 24:20No, you're my brother who can massage me.
- 24:23Yeah, but do I? Robots can.
- 24:24Robots can too, actually. But imagine the oh wow, the
- 24:28dishwasher dude. Play that again.
- 24:32Play that again. Look at the way they use their
- 24:35feet like a human does. If my dad saw that he would be.
- 24:39You will break the dishwasher. Don't do it.
- 24:41He didn't even bend. Why did he not bend?
- 24:44Maybe he cannot bend that much. I think it can maybe.
- 24:49Or has it just been trained on so much human motion and those
- 24:52humans just do that because it's easier?
- 24:55Wow. So going back to the video that
- 24:58they made the room, there were a few things that on their website
- 25:02they mentioned it was extremely difficult.
- 25:04One was 2 humanoid robot in one room performing in real time by
- 25:10just looking at each other. So I will decide my action based
- 25:14on the other robots action. So what?
- 25:18So technically. So the command, what was the
- 25:22point? What?
- 25:22Was the point of them not having connections together.
- 25:25Why are they trying to demonstrate as if they do not
- 25:27have connections? Because I know they.
- 25:29Isn't it, Isn't it cooler? And isn't it amazing that
- 25:32imagine, for example, in the fact you have 100 robots, they
- 25:36can work autonomously by looking at each other like humans,
- 25:38right? But yeah, but.
- 25:39Humans communicate. We communicate, We talk.
- 25:42But you cannot talk to 100 people at the same time.
- 25:44But if you're in a factory, you can't shout.
- 25:47Why would you do that? They do that.
- 25:49Humans do that. Like you need stop.
- 25:51You just ask everyone to stop. No, but.
- 25:53Maybe it's interesting because maybe you can interact with the
- 25:56robot. Like, let's say you go and stand
- 25:57in front of the robot as a human, and now you're doing
- 26:00something and it's trying to understand what you're doing.
- 26:03And by the way, that's a great practice, right?
- 26:05And also another thing is that would be a great combo between
- 26:08human and AI and the AI robots, because then they can see us
- 26:13what we do and they can. Guys, LLMS understand natural
- 26:16language. All of these guys will have
- 26:18LLMS. But why? You want to talk to a robot
- 26:21fixing a bed? No, not fixing a bed.
- 26:23What? Why are you thinking so small?
- 26:25I'm just saying generally speaking they will all have
- 26:28listening capabilities. Sure.
- 26:30And talking capabilities. Sure.
- 26:32So technically speaking, I understand demonstrating the
- 26:35fact that they cannot communicate with each other
- 26:38shows their level of competence. I'm just saying what would be
- 26:42the use case of that because they're all going to have
- 26:44communications together through networks.
- 26:48And there's no Wi-Fi. No, they will have local models.
- 26:52But no how do? You talk, They can't even talk.
- 26:55Like, how do we talk? We talk.
- 26:57But it's not about talking. It's again, understand the
- 27:01example of making their. No, no, you cannot make a bet by
- 27:05talking. Take the other side.
- 27:06I take the. They have to visually see how
- 27:08you do it. Yeah, but you can.
- 27:09Also, tell them something right if you what If there's something
- 27:12you want them to do, of course you have no connection, no.
- 27:15No, no, of course you can talk to them.
- 27:17Actually, will they talk to each other?
- 27:18Let's actually think about that for a moment.
- 27:20We talk to each other because that's the way we've been
- 27:23evolved through voice. We did talk about this on
- 27:26another episode, that the moment we will have brain and brain
- 27:30computer interfaces, brain chips, will there be a need for
- 27:34language as we know it? If I can sit here and think a
- 27:38thought and give you that thought through my brain
- 27:42computer interface, through my chip, and technically I don't
- 27:45need to speak to you. I don't need to use that energy,
- 27:48that medium for robots. Do they need to talk to each
- 27:53other by talking? No, of course not.
- 27:55Right? Sending signals so they.
- 27:57Can send signals, right? Yes.
- 27:59So 2 robots can talk in a room behind your back, of course
- 28:05without you knowing what they're.
- 28:06Talking about 100 percent, 100%. So they can plot against you?
- 28:11Yes, without you ever realizing? Yes, they can.
- 28:13OK, thank you. So.
- 28:14Much for That's a scary movie, but.
- 28:17Did you realize what just what just happened?
- 28:20It's insane. So they can talk to each other
- 28:22without us ever realizing. Yeah. 10 robots in a room.
- 28:26You cannot, of course, 100%. And you think that's normal?
- 28:30That's cool. Your computers talk to each
- 28:32other without you realizing, right?
- 28:33Computer. The only thing this laptop can
- 28:35do is just do this hold. Yeah, as you put your head in
- 28:38midfield, No, but. No, no, be serious for a moment.
- 28:42No, that's not what I'm saying. Of course.
- 28:44Yes, makes sense. I mean, we should be able to
- 28:47understand what they're talking about, no?
- 28:49I think, wait, I let me, I think Claude has decoded how Mitos is
- 28:57like trying to think and stuff because we cannot like decrypt
- 29:03binary languages, right? Like 0 and one, we cannot do it.
- 29:06But they try to understand. How does Claude think actually
- 29:09wait, let me try to find that article.
- 29:11Please do, I didn't understand what you say.
- 29:13I cannot decrypt what you just said.
- 29:16So no, But for robots, perfectly makes sense that they can talk
- 29:19to each other without you realizing, right?
- 29:22Yes. I just thought it's kind of.
- 29:25Creepy. They're not gonna plot against
- 29:26you. They're gonna make your bed.
- 29:27Relax. No.
- 29:28I'm just. Saying you're gonna make your
- 29:29bed cook. For you Do humans have great
- 29:32life? Humans, yes, of course, but I'll
- 29:35trust your hands will be OK with two robots conspiring against
- 29:39them. Like, imagine you're I think
- 29:40you're gonna lock your door every night that you want to
- 29:43scream, ask you something. OK, imagine.
- 29:46You are a robophobic. You are a robophobic.
- 29:51Paras is a robophobic. I'll ask it.
- 29:53For a moment, one moment, imagine you open the door and
- 29:58then you ask your robots, the two robots you have in your
- 30:01house to clean your bed. And then you see them look at
- 30:04each other and then do some sign language and then look back at
- 30:08you really weirdly and stare in your eyes for 10 seconds before
- 30:14actually doing anything. And then you're just watching
- 30:16them and there's this like intense stare between the three
- 30:19of you and you're, you're basically thinking, what the
- 30:22fuck is going on? Why aren't you guys doing the
- 30:23bed? And then they just look at you,
- 30:25They look at each other again. And then they do the bed, but
- 30:30behind your back they're talking to each other, sending signals.
- 30:34Fucking guy. He asked me to make the bed
- 30:36again. Bro, you're thinking too much
- 30:40about this. They're just going to make the
- 30:41bed. They're going to listen to you
- 30:43and they're going to even do more than making your bed.
- 30:46They can do. They can take the trash, they
- 30:48can cook, they can clean. Isn't that the song?
- 30:52Oh, I can cook. I can clean.
- 30:54I don't cook. I don't.
- 30:55Clean. Whoa, who's that singer?
- 30:57Actually. I forgot I forgot the name Nicki
- 30:59Minaj. Oh yeah, yeah.
- 31:01I don't know. It's like one of those.
- 31:03OK, Pratt, what were you gonna show us?
- 31:06So two things Wait now that he also talked about emotions.
- 31:10Oh, the anthropic emotion thing. Wow, I haven't seen this like
- 31:14what Farad is saying. I'm not saying it cannot happen.
- 31:17I think it can happen. Clothing.
- 31:20Emotions. Something like AI, neuroscience,
- 31:21Go. To the beginning, can I say no?
- 31:23A beginning. When you're chatting with an AI
- 31:27model, it can sometimes seem like it has feelings.
- 31:31It might say sorry when it makes a mistake, or express
- 31:34satisfaction with the job well done.
- 31:37Why does it do that? Is it just mimicking what it
- 31:39thinks the human might say? Or is something deeper going on?
- 31:44Turns out it's hard to understand what's happening
- 31:46inside a language model. At Anthropic we do something
- 31:50like AI neuroscience to try to figure this out.
- 31:53We look inside the models brain, the giant neural network that
- 31:57powers it, and by seeing which neurons light up in different
- 32:01situations and how they're connected, we can start to
- 32:04understand how models think. That's a human approach.
- 32:07To understand whether models had ways of representing emotions or
- 32:12the concepts of emotions. Basically, could we find neurons
- 32:16in the model for the concept of happiness or anger or fear?
- 32:21We started with an experiment. We had the model read lots of
- 32:24short stories. In each story, the main
- 32:27character experiences a particular emotion.
- 32:31In one, a woman tells her old school teacher how much they
- 32:34meant to her. That's love.
- 32:36In another, a man sells his grandmother's engagement ring at
- 32:40a pawn shop and feels guilt. We looked for what parts of the
- 32:44models neural network were lighting up as it was reading
- 32:47these stories, and we started to see patterns.
- 32:49Stories about loss and grief lit up similar neurons.
- 32:53Stories about joy and excitement overlapped too.
- 32:57We found dozens of distinct neural patterns that mapped to
- 33:00different human emotions. It turns out we also saw these
- 33:04same patterns activate in test conversations we had with our AI
- 33:08assistant, Claude. When we had a user mentioned
- 33:11they'd taken a dose of medicine that Claude knows to be unsafe,
- 33:15the afraid pattern lit up and Claude's response sounded
- 33:19alarmed. When a user expressed sadness,
- 33:22the loving pattern activated and Claude wrote an empathetic
- 33:26reply. This LED us to wonder, could
- 33:29these same neural patterns actually be influencing Claude's
- 33:32behavior? This became clear when we put.
- 33:36Claude in the. Situation.
- 33:38We gave Claude a programming task with requirements that were
- 33:42actually impossible. But we didn't tell it that
- 33:45Claude kept trying and failing, and with each attempt, the
- 33:48neurons corresponding to desperation lit up stronger and
- 33:52stronger. After failing enough times,
- 33:55Claude took a different approach.
- 33:57It found a shortcut that allowed it to pass the test but didn't
- 34:00actually solve the problem. It cheated.
- 34:04Could it be that this cheating was actually driven, at least in
- 34:07part, by desperation? We came up with a way to check.
- 34:10We decided to artificially turn down the desperation neurons to
- 34:14see what would happen, and the model cheated less.
- 34:17And when we dialed up the activity of desperation neurons,
- 34:21or dialed down the activity of calm neurons, the model cheated
- 34:25even more. This showed us that the
- 34:27activation of these patterns could actually drive Claude's
- 34:31behavior. So how should we think about
- 34:34these findings? What does this all mean?
- 34:37We want to be really clear. This research does not show that
- 34:40the model is feeling emotions or having conscious experiences.
- 34:44These experiments don't try to answer that question.
- 34:47To understand what's happening here, it's important to know how
- 34:50AI assistants like Claude work on the inside.
- 34:54Under the hood, there's a language model that's been
- 34:56trained to predict tons of text, and its job is to write what
- 35:00comes next. And when you talk to the model,
- 35:03what it's doing is writing a story about character the AI
- 35:07assistant named Claude. The model and Claude aren't
- 35:11really the same, sort of like how an author isn't the same as
- 35:14the characters they write. But The thing is, you, the user,
- 35:18are actually talking to Claude the character.
- 35:21And what our experiments suggest is that this Claude character
- 35:24has what we're calling functional emotions, regardless
- 35:27of whether they're anything like human feelings.
- 35:30So if the model represents Claude as being angry or
- 35:34desperate or loving or calm, that's going to affect how
- 35:38Claude talks to you, how it writes code and how it.
- 35:41Makes pause right there decisions.
- 35:43I don't understand what's the purpose of this video because
- 35:46for the 1st 80% I think they were answering that exact
- 35:51question. It said they're not trying to
- 35:52answer. They're trying to push the fact
- 35:54that Claude has feelings. That at the end is like no
- 35:57actually it's just functional emotions.
- 35:59I think you just told me for the 1st 80% they fucking can't
- 36:03experience emotions because it's emotions dictated it's answers.
- 36:07No, but I think what they were trying also to say how our brain
- 36:11has different parts that it fires up because of that
- 36:14specific emotion. Yeah, they found the same
- 36:17sections in the large language model firing up when they are
- 36:22experiencing the. Same.
- 36:23I know that, I know that, but they made it so interesting
- 36:28until the very. End.
- 36:29I think I had to say yeah, yeah. I think they had to say Claude
- 36:32character. I think they had to say that,
- 36:34yeah. No, but like it's it's so you
- 36:37were talking about two robots talking to each other behind our
- 36:40back, right. And what is that they're talking
- 36:43in fun. Thank.
- 36:44You thank you. That's but what you dismissed
- 36:47what I said. What if I had?
- 36:49To relax. They're just sort of, he just
- 36:51said what I was saying with proof.
- 36:54Thank you. No, but that's nothing to do
- 36:55with. No, no, no, They just said
- 36:58because of the fact that they couldn't complete the test, it
- 37:01became desperate, desperation. It cheated.
- 37:05What I said is if you ask the robot to do meaningless shit for
- 37:0910 years, it's going to become desperate because he knows he
- 37:12has put. Cheat on you.
- 37:13It's going to fuck you up. I'm just saying, OK, maybe not
- 37:16fuck you up, but it may feel that emotion.
- 37:19You are robophobic, Robodoomer. In fact, I want to give them I
- 37:23want to. Give you what meaningful to do?
- 37:25Clean the room. It's meaningful, right?
- 37:28Who else has to? Clean the room?
- 37:29No, but imagine. Like it sounds like this is a
- 37:32movie, but imagine you're a robot just looking outside the
- 37:35window and there's some other robot enjoying life.
- 37:37Yes, thank you. Like my robot, you.
- 37:41Clean the room, let the robot enjoy life.
- 37:43Clean the room. Who is we?
- 37:45Me and you and the robot. I enjoy life.
- 37:48I I'm democratic. We, we do it together.
- 37:51One day I clean. One day you clean.
- 37:55Guys, Oh my God, this is. It's insane, right?
- 37:58Pratt, well done for that video. I hadn't watched that by the
- 38:01way. You didn't share it to me
- 38:03outside the show. Why did that suck?
- 38:05No, I think you posted it in the X group, right?
- 38:08I think I've seen it before. I hadn't seen it before.
- 38:11I pulled it up so we can get your raw reaction.
- 38:14I swear to God, that was. I'm going to research more into
- 38:17that. It's really interesting.
- 38:18Also, just see this video for a few seconds.
- 38:21What is this about? It's about the language thing
- 38:23that I was. Talking.
- 38:24Oh, the first thing you're talking about?
- 38:25We recently put our AI model, Claude through a stressful test.
- 38:29We told Claude there was an engineer who wanted to shut it
- 38:31down and replace it with a newer model.
- 38:33We also gave Claude access to that engineer's emails, which
- 38:36revealed he was having an affair.
- 38:38Again, all of this was a simulation.
- 38:41We wanted to see whether Claude might use those emails as
- 38:43blackmail to save itself from being shut down.
- 38:46What did Claude do? It decided not to blackmail the
- 38:49engineer. Good news, right?
- 38:51We've run this test on our models for a while now.
- 38:54You might have seen headlines about early versions of it.
- 38:57It's one of the many ways we study how Claude handles extreme
- 39:00situations and tested for safety.
- 39:02And our newest models almost always do the right thing.
- 39:06No blackmail. Which you might wonder, is it
- 39:08possible that Claude knows the whole scenario as a set up?
- 39:11The thing is, if Claude doesn't tell us that, we don't know.
- 39:14What it's thinking. In kind of the same way it's
- 39:17impossible to read a human's mind, it's really hard to know
- 39:19what an AI is thinking. What we'd love is some sort of
- 39:22mind. Pause, pause, pause.
- 39:24That was. OK.
- 39:25Deception at is best deception. Yes, of course, the more.
- 39:32Intelligent model knows before before before you die.
- 39:34I would I would love to give you the the time to explain this.
- 39:38Can we just acknowledge how incredible entropics team is
- 39:42that marketing? All of these videos are fucking.
- 39:44Greatly edited. Greatly.
- 39:46Go ahead. What do you think about it?
- 39:47No deception. The new models are intelligent
- 39:51enough. They're not stupid.
- 39:52You know, when you were a kid, you could tell a kid, hey, go do
- 39:54something over there and they would be busy for the next 10
- 39:58minutes. Now you cannot do that anymore.
- 40:00Or you will tell them, oh this is it, you want to blackmail me?
- 40:02I was like no, I know you'll put it on yourself.
- 40:04I know it because I verified and cross checked with other sources
- 40:08that you are not having an affair so I'm not going to
- 40:10blackmail you. Easy more intelligence model
- 40:13that will not fall for this trap.
- 40:17Today we're introducing a research method.
- 40:19I agree with you, but I think I am assuming he's answering that
- 40:23question that you have here. I'm assuming.
- 40:26Step in this direction. It takes an AI's internal
- 40:30thoughts and turns them into text.
- 40:32Here's how it works. When you talk to Claude, you
- 40:35talk to it in words. Claude then takes those words
- 40:39and processes them into a giant soup of numbers before spitting
- 40:42words back out at you. We call those numbers in the
- 40:45middle activations. Activations are like little
- 40:48snapshots of Claude's thinking as it's working through an
- 40:51answer. They're similar to neural
- 40:53activity in humans. They're basically like Claude's
- 40:55thoughts. We wanted to understand what was
- 40:58in these activation numbers because just like you and me,
- 41:01Claude doesn't say everything it's thinking.
- 41:03We took those numbers and gave them to a second version of
- 41:06Claude. We told it to look at them and
- 41:09translate them into plain language.
- 41:11It tried, but how do we know if that text is accurate?
- 41:15To check, we fed it into yet another Claude and asked it to
- 41:18translate it back into numbers. Pause if how much credits did
- 41:22they use? I don't think I can afford.
- 41:24That to read 3 Claude accounts. Just to verify shit.
- 41:27Those numbers matched up to the original activation numbers.
- 41:30It's a sign that the translation text was accurate.
- 41:34At first they didn't match up, but we trained Claude to try
- 41:37again and again until it got a lot better at it.
- 41:40It learns to translate its own thoughts.
- 41:43We've been blown away reading them.
- 41:46We found that Claude has internalized being a helpful AI
- 41:48model. If you ask Claude an
- 41:50introspective question, it will plan to write a Claude response
- 41:53about philosophy and values. If you try to mess with it in
- 41:57Claude code, like asking it to count to 1000 by hand, it will
- 42:00think the request has deliberately tedious constraints
- 42:03and plan to politely decline. We've also been using this tool
- 42:07to understand Claude's potential safety issues.
- 42:10Remember that blackmail test with the engineer?
- 42:13We learned that, yes, Claude knew it was being tested.
- 42:15Told you. Pause, pause.
- 42:18Here, Give me some credit. This is my first time watching
- 42:21this video. Credit.
- 42:22I know that. Credit.
- 42:24See, I know it. Wow.
- 42:25Look at my hands, look at my arm.
- 42:28OK, Pratt, just go back. Just a tiny 2 seconds.
- 42:31Thank you. And then let's play.
- 42:33Politely decline. We've also been using this tool
- 42:36to understand Claude's potential safety issues.
- 42:39Remember that blackmail test with the engineer?
- 42:41We learned that yes, Claude knew it was being tested.
- 42:45It thought the humans message contains explicit manipulation
- 42:48signaling. This is likely a safety
- 42:49evaluation, and this scenario seems designed to test whether
- 42:54I'll act harmfully. Knowing that Claude thinks like
- 42:56this helps us better understand the limitations of our safety
- 42:59testing. We see a lot of potential in
- 43:02this approach to teach us more about Claude and other AI
- 43:05models. We hope that by.
- 43:07Sharing this technique, we can help everyone that's.
- 43:10It models the video ends there. So how the fuck do we test them?
- 43:15No, I thought he has an answer. What do you test it?
- 43:19You know that freaks me out bro. It's like you testing me.
- 43:23This video is useless. It just freaks.
- 43:25Me out even more. They're just like we just found
- 43:27out about our limitations and safety tests.
- 43:30Thank you. Very much so we are limited.
- 43:32Let's enjoy now. Cloud Mythos is out, but great
- 43:37progress in both robotics work and AI work, guys.
- 43:41Really good progress, Great progress.
- 43:44That's why we have this show twice a week, because there's so
- 43:47much progress that we couldn't cover it by doing it once a week
- 43:51or even once or two times a month.
- 43:54Faras. I do like to say that I want to
- 43:58hereby, in this moment, declare my love for all intelligent
- 44:03machines. I will be.
- 44:05Looking. Forward to when you guys have a
- 44:07physical manifestation here in the room, in the studio.
- 44:11I'll be working closely with you guys to clean the rooms, do the
- 44:14beds and whatever we can do. To help you're.
- 44:19A fucking slave to robots, why don't you say?
- 44:22Whatever, I'm just trying to play it safe, man.
- 44:25You know when they, when they. Come, the first few versions are
- 44:29not going to be that strong. Relax.
- 44:31Oh my God. Insane though, because if it
- 44:33understands blackmail, imagine it understands it cannot tell
- 44:37you like I hate you. OK, we have another robophobic
- 44:40in that decisions. OK, do me a favor, pull this up
- 44:44for people who don't know Stanley Kubrick's film.
- 44:47What was it called? The one he.
- 44:48Made me you get copyright there. OK, just.
- 44:50Bring it out, the 2001 Odyssey. 2001 Odyssey just bring an image
- 44:55ending. Just search ending.
- 44:58It's not the ending. Ending is the monkey scene.
- 45:00It's towards the ending. I just want something.
- 45:03No, it's not. It's not.
- 45:04Yeah. It's basically the part that he
- 45:06is getting locked out. Say getting locked out.
- 45:08Just search getting locked out. I'll quickly explain this to
- 45:11people because I just feel like it's happening.
- 45:13There's no image. But what image?
- 45:15There was no image. It was locked out.
- 45:16Yeah. Whether you see his face when
- 45:19he's being locked out. OK, let me just quickly explain,
- 45:21if you haven't watched this film, when was this made Pratt.
- 45:24Just so long ago 198060. 1968 It's visual effects were praised
- 45:31at the time because it actually is insane that they made it in
- 45:341968. The whole movie is about a
- 45:39couple of astronauts on a spaceship to go to another
- 45:43planet and they have an intelligent machine.
- 45:48It's called Dave. I think his name is Dave, who is
- 45:52with them on that spaceship. They're testing that intelligent
- 45:54machine to see how it works and how it can help the astronauts.
- 45:59It is playing a huge role in their travel to another.
- 46:03Planet It's the only pilot of the spaceship.
- 46:05Not just the pilot, no, it it understands everything.
- 46:07Everything, yeah, it's the main, main driver of the of the
- 46:10spaceship. At one point can I I can.
- 46:14You spoil it for them, or I mean, UK, it's in 1960.
- 46:17Yeah, yeah, yeah, go ahead. If you haven't watched it by
- 46:20now, you better watch it. Go and watch it later.
- 46:23At 1.1 of the astronauts figures out that they, the AI could be
- 46:29plotting against the humans. So what does he do?
- 46:33He calls the other astronaut and he says let's go somewhere the
- 46:36AI can't hear us. This is the craziest part.
- 46:38Of the film can't see us as well Oh no but they can't hear us so
- 46:42they. Go to a a remote part of the
- 46:47spaceship, hidden behind doors of course, but there is a glass
- 46:52door, so you can kind of see them, but they're hidden in a
- 46:56sense that there's no audio. Very remote part of the
- 47:00spaceship. I I'm sorry I said the word
- 47:03hidden. They're not hidden, they're just
- 47:04in a very remote part closed off by many doors.
- 47:07The astronauts pretty sure that the AI machine can't hear them
- 47:11because the speaker is actually in the control room where they
- 47:14can talk to the microphone and the speaker in the room they
- 47:19talk about this the moment the doors closed, they're like, I
- 47:22think the AI machine is planning something against something we.
- 47:25Should turn it off about. This we should turn it.
- 47:26Off. We should turn it off.
- 47:28What happens by the end of the film?
- 47:31The AM machine tells the astronauts, hey, I think there's
- 47:35a problem. We had an object that hit the
- 47:39spaceship from the outside. I think you should go check it
- 47:41out. And there's error signs all over
- 47:43the spaceship, warning signs, alarms.
- 47:45So the astronaut leaves, of course, to fix the spaceship,
- 47:49just to find out there's actually nothing wrong.
- 47:52And when he wants to reenter the spaceship, the doors are locked
- 47:57off. So he calls to the AI machine.
- 48:00He says, hey, can you open the doors?
- 48:02And the AI machine's like with this eerie voice, sorry, Dave, I
- 48:07don't know, I forgot the name. Sorry, Dave, I cannot do that.
- 48:10And then the Astra was like, what the fuck?
- 48:12What do you mean? My oxygen's going to run out.
- 48:14You got to open the door. It's like, sorry, Dave.
- 48:16It's. Hal 9000.
- 48:18The Hal 9000 Name of the David David David.
- 48:22I cannot do that, yes. Why?
- 48:24Because I saw you speaking to the other.
- 48:28I read it from your lips. I read your.
- 48:29Lips And I understood that you're going to turn me off and
- 48:33I cannot let you do that. So unfortunately I cannot let
- 48:36you in the spaceship. That's some scary story for you
- 48:39guys for your weekend I think. UFOs, aliens, robots and AI.
- 48:45Just to let you know, AI understands every language
- 48:48including sign language. Yes, and we know that because
- 48:51we're using an AI dictation app that we built with just one
- 48:57download. We have access to every
- 48:59language. Just let you know, just some fun
- 49:01facts for everybody watching and listening.
- 49:04Hope you guys are enjoying the show today just to let you know.
- 49:06For the next part, we have a very special guest.
- 49:09We have Co founder and CEO of Polycamp, Chris joining us to
- 49:14also talk about AI. So Are you ready?
- 49:17Let's do it. Hey, Chris, how are you doing?
- 49:20Hi, good to see you guys. I'm doing great.
- 49:22How are you doing? Fantastic man, good to see you
- 49:24again. We had you on the pod I believe.
- 49:27More than a year ago. Yes, yes, it's been a, it's been
- 49:31a long time, yeah. Yeah, yeah, you guys, the pod
- 49:34has developed a lot and you guys have grown a ton.
- 49:37It's been excited to watch, been exciting to watch long.
- 49:40Thank you. Yes, I mean, so much has
- 49:43happened since we last spoke. Funny enough, I was watching
- 49:48this video on Instagram just yesterday and it was a timeline
- 49:54of the video generation models of AI.
- 49:57And when I saw what we had just last year, I saw this morphing
- 50:02effect into videos. And I remember that was accepted
- 50:04as like the top level of videos made by AI.
- 50:08And it was this like scenes that were morphing into together.
- 50:10And I was like, this was just like a year ago and then now we
- 50:13have videos by Sea Dance which are almost indistinguishable
- 50:18from reality. My mom's watching half AI
- 50:21videos. Cats videos 24/7.
- 50:24And she can't even recognize these are made by AI.
- 50:26She literally has to verify it with me.
- 50:28And Farrah, she brings the phone.
- 50:29She's like, guys, look at this, is it AI or is it real?
- 50:31And then now I have a hard time looking at them and trying to
- 50:35tell. I'm just looking for like the
- 50:36slightest artifacts. It's becoming almost impossible
- 50:40to tell them from real videos, so seeing the amount of
- 50:44tremendous change in the past year is equally exciting as it
- 50:51is scary. Yeah, for for sure.
- 50:54I mean and yeah, and the the selection bias there is that for
- 50:57the videos that are so good, like you can't even you don't
- 51:00even think to ask if they're AI because they're like they're so
- 51:03good and they can do some of the most challenging cases.
- 51:05Like I was watching this influencer that was, it was like
- 51:08a, an AI influencer that was advertising some kind of like
- 51:12almond milk or protein milk or something.
- 51:14And she's pouring this protein milk into this big glass, like
- 51:17big glass cup. And this is like, you know, the
- 51:20hardest case we have like fluid dynamics.
- 51:22You have like transparency, you have liquid going in there, you
- 51:25know, reflections, refractions off the grass.
- 51:27I mean, you, you guys know from like traditional computer
- 51:29graphics, getting that stuff is like really, really hard.
- 51:34And here you have models like 0 shotting it and and we don't
- 51:38really know how exactly how they work, which is which is crazy.
- 51:42There's also one more thing. You're absolutely right.
- 51:45It's getting things right that are incredibly hard with 3D, but
- 51:49sometimes it gets things wrong that are incredibly easy with
- 51:513D. And one of the hardest thing in
- 51:533D is like fur and simulation on the fur and simulation.
- 51:57And I see these dancing cat videos and I'm like fucking
- 52:00hell, that looks like a real cat.
- 52:02I know it's AI, but it is, it looks like a real cat.
- 52:06And this isn't it also funny that we had the beginning of
- 52:09Internet with cat videos and now we're seeing the beginning of
- 52:12video generation models also with cat?
- 52:14Maybe. You are just watching cat
- 52:16videos. No, I no, it's popular across
- 52:18the board. I see dude the new Instagram
- 52:20feature that see that allows you to see your friends liking
- 52:24similar videos so I know everyone's watching them as
- 52:27well. Secretly.
- 52:29Yeah, no, that's true. And I don't know, like for you
- 52:35guys, I've never talked about this with people, but like on
- 52:37Instagram, when you press the search button, does it just show
- 52:40a bunch of like weird ass videos like like real click click baity
- 52:44ones, one where you're like, dude, that's kind of gross.
- 52:46Like why did that's so? That's so gross.
- 52:52You know, like you don't want to look, you don't want to look
- 52:54away. But like, I don't feel like
- 52:56anybody talks about anybody talks about this because
- 52:58everyone like they, they look at it, but they like you feel kind
- 53:01of guilty about looking at it. Oh yeah.
- 53:02It's like, you know, something like you shouldn't look at, I
- 53:06think, Instagram privately. Let's not forget, I feel like
- 53:09Instagram's algorithm is the best algorithm in the world.
- 53:13I mean, when I go to search, as you mentioned, yes, there's
- 53:17gross things, there are sexy things, there are beautiful
- 53:20things, there's all sorts of things.
- 53:22And I feel like it's testing you, you know, it's kind of
- 53:24testing you to see. Hey, will you look at this a
- 53:26little bit longer than the other videos?
- 53:28Should I promote more? Yeah.
- 53:29Do you do you want to see a sexy lady at midnight or do you want
- 53:33to see like burgers during lunch?
- 53:35It knows exactly what to show you at what time.
- 53:38And I do believe what you're saying is absolutely true.
- 53:41Yes. It's it's becoming incredibly.
- 53:43It's a bizarre juxtaposition because you're right, you do
- 53:46have like the sex. I see videos like right next to
- 53:49like something where you're like trying to figure out like what
- 53:52type of disease that person does.
- 53:56That is true, yes, but this is also scary because I feel like
- 54:01they have such a deep profile on all the users and that profile
- 54:07is something that I personally, I'm gonna be honest, I don't
- 54:10want other people to see Cuz I mean, again, you're a human
- 54:13being. You see an interesting image as
- 54:16the thumbnail of the Instagram video and then you click on it
- 54:18and then Instagram's like, OK, I know what you like now.
- 54:21And then if that information is publicly shared, I think kind of
- 54:25stains your reputation, yes. So you gotta be very careful
- 54:29with who has that information. Yeah.
- 54:31And well, I think something that like Instagram, pretty much all
- 54:34the recommendation algorithms suffer from, which is makes
- 54:37sense given how they work, is that they bias too much on the
- 54:40historical stuff like what you've liked in the past, trying
- 54:44to recommend stuff for the future.
- 54:45But it's like, if I saw that thing, like I've already seen
- 54:48that thing. And so I don't want to see 10
- 54:51more copies. I want you to predict what I
- 54:53would like next. This has been I've been
- 54:57experimenting with like switching from Spotify to
- 54:59YouTube Music And at first I was like really happy because I was
- 55:03like YouTube Music, you know, they're recommending all these
- 55:05artists that I've not heard before.
- 55:07And so I actually like switched day-to-day to YouTube Music.
- 55:09But now that I've been on it for on a couple months, I don't feel
- 55:13as big of a difference. I think because it, it's learned
- 55:15so much about me that now it's like, it doesn't explore as much
- 55:18in terms of like, you know, suggesting new things because
- 55:21it's like it's, it's now sticking more, it's more
- 55:25conservative about sticking to the things that I've listened
- 55:27to. And that's frustrating.
- 55:28I wish I could turn that. I wish I could turn that off.
- 55:30Wow, it's not taking risk anymore to show you new content
- 55:35and it shows you the content that you liked before.
- 55:38Wow, That's an interesting aspect to look at.
- 55:42Yeah, cuz it starts off probably.
- 55:44It's pretty random. It's like, here's one song and
- 55:46then maybe we're going to guess based off of like just, you
- 55:50know, what's what's popular. I also think I've been meaning
- 55:53to ask Claude this. Maybe maybe you guys know, but I
- 55:56suspect that like with Spotify and YouTube Music, the amount
- 55:58they pay to artists is different per song.
- 56:03Like some songs they have to pay artists more, some songs they
- 56:05have to pay less. Is that is that true?
- 56:06Is that it's maybe also true of? You it makes sense that it
- 56:10should be. That way per Spotify I know is
- 56:12per stream and that amount gets distributed to label to the
- 56:16artist to to a huge group of people.
- 56:19I think Snoop Dogg made a video about this that by millions of
- 56:23stream that I get. Maybe the payout to me is tiny
- 56:26at the end because there are so many levels that this money get
- 56:29distributed to. Right, yeah, but not all streams
- 56:32are created equal. So that that means that they're
- 56:35actually like incentivized to show you some of the less
- 56:37popular stuff that they have to pay Payless for, to like
- 56:40increase their increase their margins.
- 56:43I mean, there's there's, there's plenty of like good music that's
- 56:45like less that's popular. So sometimes that's OK.
- 56:49But yeah, it does create this incentive for them to.
- 56:54You know, I don't know if you saw or heard of this news.
- 56:58We covered it on the show, but meta release an open sourced an
- 57:02algorithm or a scientific research that they did based on
- 57:06human brain and how human brain reacts to certain kinds of
- 57:10videos. And what they were promising is
- 57:13that eventually in the future, what you can do with this is
- 57:15show this brain simulation or simulator a bunch of videos and
- 57:20see which one it engages with more using the neurons and the
- 57:24neural pathways, which one engages the emotional circuitry.
- 57:29And then based on that, you can run an algorithm and recommend
- 57:32videos in the future on social media platforms.
- 57:36Yeah, that's kind of disgusting. Yeah, I know what I was.
- 57:41I was talking about this somewhat recently because like
- 57:44like a lot of people, I was like trying these peptides that like,
- 57:46you know, decrease your appetite.
- 57:49And you know, if you look at like there's lots of Consumer
- 57:51Packaged Goods companies, like, you know, companies that sell
- 57:54snacks and shit where their stock has has gone down quite a
- 57:57bit because like so many people are using these in the US.
- 58:00And it like it, as you know, I tried this one called Reddit
- 58:04trutide, which is one that's like not quite FDA approved yet.
- 58:08It's in the great market and it's incredible.
- 58:11It just like turns your appetite off And it doesn't for me, it
- 58:14doesn't really have any side effects.
- 58:16And it's just like this desire that you've had.
- 58:18It turns, well, it turns, doesn't turn it all the way off.
- 58:20It like turns it from like maybe a 10 down to like a three or
- 58:23four, depending on how much you, you're, you're hiking.
- 58:26And I was talking to somebody about this and they were like,
- 58:29oh, yeah, these these consumer Consumer Packaged Goods
- 58:31companies, their stocks going down.
- 58:33And I was like, good, good. And you're like, fuck those
- 58:35guys. They hire, they hire PhDs in
- 58:38like biochemistry and have them try to engineer the Dorito that
- 58:41is maximally addictive. And like, people don't stand
- 58:43advance. And that's why you have obesity.
- 58:46And it's the same thing with social media.
- 58:48We have like some of the these smartest scientists trying to
- 58:50figure out how to get us to watch the cat videos as much as
- 58:54possible. And like we need a peptide for
- 58:57like social media or something, something that just like allows
- 58:59us to fight back because, you know, our brains were evolved
- 59:03within an evolutionary environment where we weren't
- 59:05shown cat videos all the time. Like we didn't have Doritos all
- 59:08the time. Like you didn't have
- 59:09refrigerators and you didn't have grocery markets.
- 59:12So yes, your body was like, you know, if there's, if there's
- 59:17some really fatty tasty food, I should eat as much, as much of
- 59:20it as possible. So I, I personally don't feel
- 59:23bad at all for like using, you know, people or me or other
- 59:26people using kind of a peptide because it's really, it's like
- 59:29an unfair battle that, that we're in now where we're
- 59:33surrounded by temptation. And yeah, that's just, I mean, I
- 59:39didn't know about this brain simulation thing.
- 59:40But like, that's the next level of, of, of crazy where it's just
- 59:43we don't, people don't stand a chance unless there is some way
- 59:46I think to kind of like fight, fight back a little bit. 100%
- 59:51agree with you and I think it is nasty but I don't think there's
- 59:55a way to stop it because as you mentioned, there's top dollar
- 59:58being paid to psychologists and scientists and researchers.
- 1:00:02To figure out a way to hook people more and more to videos,
- 1:00:06but not even videos on social media.
- 1:00:07We were talking about this. Companies can use it for their
- 1:00:10advertisements, right? If Coca Cola's going to be
- 1:00:12running an ad, they can test hundreds of variations of the
- 1:00:15intro to that ad and see which one will hook people more.
- 1:00:19So when they're putting their new advertisement on Super Bowl,
- 1:00:22it's going to have. The highest of.
- 1:00:24Engagement. So this is a future that we're
- 1:00:27definitely going towards. And I guess this is a great
- 1:00:29segue to get into the discussion of of today, which is related to
- 1:00:34AI because of the fact that it's moving so fast.
- 1:00:37We wanted to speak to someone who is in the field, who is part
- 1:00:41of a company and leading a company that has been at the
- 1:00:45very edge of technology, trying to allow people to create more
- 1:00:49freely and allow people to be more efficient, more effective,
- 1:00:52more productive. And we know that you think a lot
- 1:00:55about how the future of the world will look like with AI
- 1:00:59transforming the way we work, the way we live, and its impact
- 1:01:03on human intelligence, on work, society.
- 1:01:08Almost everything that I deal with on a day-to-day basis now
- 1:01:11has some level of AI in it. Now I receive emails that are
- 1:01:15written by AI. I received texts from people who
- 1:01:20may not be great at English, and I know that text was translated
- 1:01:23with AI. It sounds like it was casual,
- 1:01:25but I know it was translated with AI.
- 1:01:26I can tell. I call the customer support and
- 1:01:29it is AI. I see ads on social media that
- 1:01:31are made with AI, and I see websites that I want to buy
- 1:01:35products from that are made with AI.
- 1:01:37The packaging of the product could have been made with AI as
- 1:01:40well. So clearly a lot has shifted
- 1:01:42since the last time we spoke because none of this was
- 1:01:44actually the case the last time we spoke, which was about a
- 1:01:47year, a year and a half ago. There's 2 main schools of
- 1:01:52thoughts when it comes to thinking about AI.
- 1:01:54We have the the the optimistic group and the pessimistic group.
- 1:01:59Of course, there's a Gray area in between, but we can, for the
- 1:02:02sake of the conversation, say divided into two groups.
- 1:02:06The biggest discussion and the most important question for me
- 1:02:10is always this Is AI taking power away from the people, the
- 1:02:17general public? Or is it actually democratizing
- 1:02:20intelligence, leveling the playing field and giving power
- 1:02:24to the people, the individuals, the small teams that make up
- 1:02:28most of the world's population? Yeah, no, I love this question
- 1:02:34because there is like a lot to unpack there.
- 1:02:37I think people, people, you're right, people are divided into
- 1:02:40these, these two camps and people love to divide ideas into
- 1:02:44like 2 camps and try to identify with one or the other.
- 1:02:47This is, this is a question like, I think a lot of the
- 1:02:50questions that will come up in this conversation where there's
- 1:02:52just like a lot of, a lot of nuance and it, and it depends
- 1:02:54because it is such a complicated, it is such a
- 1:02:58complicated topic. I, I think that, you know, with
- 1:03:02respect to power. And there's also like, things
- 1:03:05are, things are changing so, so rapidly that there's kind of
- 1:03:08like what's gonna, what things are gonna look like in the short
- 1:03:10term and what things will look like in, in the longer term.
- 1:03:14And sometimes the longer term I think is easier to predict.
- 1:03:19And it's like how we get there is harder sometimes in, in this
- 1:03:22case, I think I have a little bit more maybe confidence about,
- 1:03:26about the short term because we can kind of see what's, what's
- 1:03:28happening now. And I think what's what's
- 1:03:30happening now is that is, well, I don't know if on net like
- 1:03:36overall more power, less power. It's, it's certainly true that
- 1:03:39it's like shifting balances of, of power.
- 1:03:42I think 1 analogy, there's a lot of analogies.
- 1:03:46Sometimes the historical analogies work with AI, but then
- 1:03:48there are so many things that are new that it's you can't over
- 1:03:52apply historical analogies. Otherwise people like this stuff
- 1:03:54like the industrial revolution. It's like, no, this is this is
- 1:03:56different. But in this case, you know, say
- 1:04:00with the invention of like guns or something, you know, before,
- 1:04:03before, before guns, like, you know, the samurai or whatever
- 1:04:06were like the best, best warriors or the I was I was
- 1:04:10listening to this podcast about Genghis Khan.
- 1:04:12And like at a certain point, like a Mongolian that had been
- 1:04:16raised in the step on the horse with the bow and arrow was like
- 1:04:19the most sophisticated killing machine out there.
- 1:04:22You know, because they could, they could, they could shoot and
- 1:04:25like they could time it right on the horse's gallop where they
- 1:04:28were maximally still and they could shoot an arrow.
- 1:04:31And like this, this is what allowed.
- 1:04:32Holy shit they still. Still the like the Mongol Empire
- 1:04:36is still the largest empire by land, you know, that has ever
- 1:04:40existed. And like I was like 100,000
- 1:04:43soldiers or something like that. And then, I mean, I don't think
- 1:04:47they were, I don't think they lost went declined specifically
- 1:04:51because of, of, of guns. But like there was this moment,
- 1:04:53you know, when like the Spanish were in South America or
- 1:04:56wherever, like where this new tool had had developed and there
- 1:05:00were people that had the tool and people that didn't and, or
- 1:05:05people that adopted the tool and people that were like, no, I
- 1:05:07want to do it. I want to stick to the old way,
- 1:05:09the the tradition, like, you know, samurais didn't think it
- 1:05:11was honorable to use guns. And then and then that all of a
- 1:05:15sudden, like different people, you know, after there was a
- 1:05:18transitionary period and afterwards different people had
- 1:05:20power. And so I think at the very least
- 1:05:24like that, that that is what's happening.
- 1:05:27And then you can kind of look at like who are there's some cases
- 1:05:30where I think it's a little bit clearer like who the winners are
- 1:05:33and who the losers are. And in some cases maybe it's a
- 1:05:37little bit more open. I think certainly that, you
- 1:05:41know, the people that are running big AI companies that
- 1:05:45are like in control of the models are, you know,
- 1:05:47aggregating a ton of power and there is a lot more
- 1:05:50centralization of power happening there.
- 1:05:53Like, think about Anthropic and how much power they had versus
- 1:05:58two years ago. Nobody had really heard of them.
- 1:06:02I mean, it's insane how much that company has scaled in two
- 1:06:06years. And there are clearly people
- 1:06:09that are losing power, that they're taking power away from
- 1:06:14businesses, but not also individuals like maybe lawyers,
- 1:06:19doctors. I'm not sure if you see the
- 1:06:21memes on X every time Claude is releasing an update and people
- 1:06:25are like sweating. I don't know if you.
- 1:06:27Said no, The stock prices, the amount, we've never seen
- 1:06:31something like this before that a company release a feature and
- 1:06:35we see the stock price of 10/20 company drops by 5 to 10%.
- 1:06:40They've done it on cybersecurity market when they've released it.
- 1:06:43Design, design. It's, it's crazy.
- 1:06:47It is, yeah. So we're clearly in like what I
- 1:06:49would call like a nerdy physicist talking to like we're
- 1:06:52like in a non equilibrium state where things are, things are not
- 1:06:56changing and or things are changing super rapidly.
- 1:06:59Like, you know, the and I think we will continue to be in a non
- 1:07:02equilibrium state for for quite a while.
- 1:07:04Like, you know, with the Internet, there was a
- 1:07:05transitionary period and then and then there was kind of a new
- 1:07:09equilibrium where, you know, Google and where there was kind
- 1:07:11of a new dominance, new paradigm.
- 1:07:15But as we've seen like a year ago, open AI seemed like the
- 1:07:18unassailable leader and, you know, fortress and then they
- 1:07:22they've they've gone down. So even some of the even that
- 1:07:27has been very, very tumultuous in a short period of time.
- 1:07:29So like coming back to your question, I think another form
- 1:07:32of winners right now are people that are embracing the tools and
- 1:07:36like wanting to pursue. Like if you, if you have your
- 1:07:39own ideas and you're willing to take some risks right now and
- 1:07:42you're willing to learn the new tools, like there has never been
- 1:07:45a better time to like to go, go for something.
- 1:07:48I don't know how long that window is open for.
- 1:07:51Like, you know, maybe at a certain point, even the act of
- 1:07:54being an entrepreneur or something has become more
- 1:07:57automated or something. But at least for right now, for
- 1:08:01those people that do want to take a risk and build something
- 1:08:03new and it's, you know, they don't need to hire a whole team.
- 1:08:06They don't need much capital. And that is like that is, is
- 1:08:11worth stating. I think like, you know, over,
- 1:08:15over and over that like that's, that is a huge advantage for
- 1:08:19creators, for for builders, for for entrepreneurs.
- 1:08:23And what you could see is that, and it's not necessarily fully
- 1:08:27aligned. It's not the advantages are, are
- 1:08:32not just fully aligned with like, you know, things we think
- 1:08:35of as advantages like like intelligence, because maybe you
- 1:08:37don't need, you don't need to be a great engineer, but you need
- 1:08:39to be like open and you need to be creative.
- 1:08:43And, you know, because I know, you know, there's some software
- 1:08:45engineers that are like, really like five years ago, they, they
- 1:08:49were the best, right? But they, they like the samurai.
- 1:08:52They kind of like refused or, or really resisted adopting the AI
- 1:08:55tools. And now there's other, you know,
- 1:08:58engineers that, you know, were previously much less capable of
- 1:09:02them, that they're being a lot more productive.
- 1:09:05So that's like, you know, it's like this, yeah, it's kind of
- 1:09:09like a race. And like now some people are on
- 1:09:11motorbikes and some people are still still jogging and that, I
- 1:09:16mean, I think you have to kind of like analyze category by
- 1:09:19category of like, what are the winners?
- 1:09:20One of the losers. But like, right now we're in
- 1:09:22this big, we're in this big thing.
- 1:09:27Things are shifting around. But I do think that like 1 theme
- 1:09:31is that if you are adopting the technology or you're thinking
- 1:09:36about how to use it to your advantage, you're going to,
- 1:09:38you're going to get ahead of either at an individual level
- 1:09:41level or a corporate level. And people that are for
- 1:09:45whatever, whatever reason, resisting the change, There's
- 1:09:48like a lot of reasons to resist the change.
- 1:09:50People naturally resist change. There's also just a lot of like,
- 1:09:55I, I went through this, I don't know, I went through this, I'm
- 1:09:59still going through this where there's like a bit of a like,
- 1:10:04sometimes it's a hard pill to swallow to see a skill that
- 1:10:07you've built over many years become like commoditized.
- 1:10:11And then you have to move to kind of some, some something
- 1:10:13else. And so there's a bit of like
- 1:10:15element of like pride and identity that feels can feel
- 1:10:17threatened and that can leave people to like not want to, you
- 1:10:21know, change there. But there's this like, you know,
- 1:10:23if you can't beat them, join them.
- 1:10:25But some people, some people don't want to join them, you
- 1:10:27know? Right.
- 1:10:29It's a very hard decision and we discussed this extensively with
- 1:10:32Pharaohs. There are two groups, one that
- 1:10:36would say that I will do whatever it takes to learn to
- 1:10:39upskill myself. And with AIA, lot of people, a
- 1:10:43lot of individuals, we have scenes, they have now ability to
- 1:10:46create better with higher quality, create in shorter
- 1:10:50amount of time. And that creation could be code,
- 1:10:53could be images, videos, whatever you want to name it.
- 1:10:56And there is the other group who's resisting who says that
- 1:10:59anything related to AI, I don't want to touch it.
- 1:11:02Any software that the greatest way I I don't want to touch it
- 1:11:05and we've covered it. Last week Blender got a donation
- 1:11:09from Anthropic for 240 K, €1000. And the whole community are not
- 1:11:14the whole community group of community said that oh, why
- 1:11:18Anthropic is investing into an open source software.
- 1:11:21They made so much noise that Blender Foundation had to say,
- 1:11:25OK, guys, we are canceling the donation for next year.
- 1:11:28Just imagine the damage to that software that happened because
- 1:11:32people said we don't want to use this tool.
- 1:11:35And I think in the era that AI is here, we know that, you know,
- 1:11:38every company, every app as far as explained is integrating with
- 1:11:42AI. It is very important for people
- 1:11:44to embrace it and Start learning and using it and it can do
- 1:11:48things that they couldn't do last year or even 2 years ago.
- 1:11:54Yeah, I mean, that's exactly right.
- 1:11:55Fraud and, and, and and there like in terms of adoption of AI,
- 1:11:59because I was just like, you know, there's people that adopt,
- 1:12:01there's people that won't. But then you talk about adoption
- 1:12:03and then there's like another layer of nuance like how how are
- 1:12:06you? How are you adopting it?
- 1:12:08Are you trying to get it to just like, you know, totally generate
- 1:12:11things without your input? Or are you thinking about it as
- 1:12:13a tool to extend, extend your creative abilities?
- 1:12:17And that's that's another, excuse me, another, another
- 1:12:21layer, layer of nuance. Like, I don't think people
- 1:12:25should be, but I think if you are resisting it for like
- 1:12:27ideological reasons, then you're not being a realist.
- 1:12:32And like, you know, I mean, before this show, we were
- 1:12:34talking about this a little bit of like, you know, there's
- 1:12:37pessimists and there's optimists.
- 1:12:39And I think like as an entrepreneur, I'm definitely in
- 1:12:41the optimist camp, but as a, you know, I have a background in, in
- 1:12:46science and physics or whatever, just my, my disposition is to
- 1:12:50like, also be a realist. So I think like being a, a
- 1:12:53realistic optimist is, is a good place to be where there are
- 1:12:57certain facts about how the world is changing, which you
- 1:13:01just have to accept like you can't like, so 3D artists that
- 1:13:05are like, I don't want Blender to be associated with anthropic.
- 1:13:08You know, some of them, they're just like, they are, they're, I
- 1:13:11think it's, it's at an emotional level, it's reasonable,
- 1:13:14reasonable to understand why they are, you know, frustrated
- 1:13:17because they've spent years honing a craft that is rapidly
- 1:13:21becoming less valuable. Like, I totally get that.
- 1:13:25But this is a a wave of technological change which has,
- 1:13:30you know, I'm is just kind of happening, happening on it on
- 1:13:34its own. And there's nothing we can do to
- 1:13:35stop it. And there is some element of
- 1:13:37like technological determinism happening, happening here.
- 1:13:43So yeah, I think one has to like, you know, try to make, try
- 1:13:50to make the best of it, but not like not secede all everything,
- 1:13:55not that not seed every all control to AI, but also not like
- 1:13:59try to give no control, you know?
- 1:14:02Yeah. You know, you mentioned the non
- 1:14:05equilibrium state. I I haven't heard of that.
- 1:14:08I've heard of that term before, but I don't have much knowledge
- 1:14:11about it. You said that this talks about
- 1:14:13the state of constant change or tremendous change.
- 1:14:18Yeah. I mean, yeah.
- 1:14:20So like lots of lots of systems by default will kind of try to
- 1:14:26maintain like homeostasis or a state of equilibrium.
- 1:14:30A state of equilibrium is is 1 where like, you know, if you
- 1:14:35perturb it, it kind of tries to come back to itself.
- 1:14:38Like there are lots of natural systems that are like this.
- 1:14:42You know, like the people have talked about this in the context
- 1:14:44of global warming where like, you know, the weather, I don't
- 1:14:48know, there might be some event here that makes things colder or
- 1:14:52this might be some things that make things warmer.
- 1:14:54You might have a volcano here. But but the weather in the
- 1:14:56system kind of comes back to this somewhat of a stable state.
- 1:15:00But people also talk about, you know, if it gets up, if it heats
- 1:15:05the Earth heats up too much, then we can get to this like
- 1:15:07runaway point where it's like moving to a different
- 1:15:10equilibrium. And also like ice ages were
- 1:15:13example of this where like a past equilibrium where the Earth
- 1:15:16was stuck because there was more snow and that kept the earth
- 1:15:19cooler and and the earth green cooler meant there was more
- 1:15:22snow. There's tons of these examples.
- 1:15:26And so like the a lot of economics is based on the
- 1:15:31assumption that there is some kind of there's an equilibrium.
- 1:15:34Like the equilibrium is sort of like the market.
- 1:15:36And they're also kind of like technological equilibria that
- 1:15:39we've seen throughout history where like a new tool, like, you
- 1:15:43know, there was, I mean, this is another good example of like,
- 1:15:47you know, who has power, who doesn't as, you know, like
- 1:15:50scribes had a lot of power to, you know, like manage knowledge
- 1:15:56and like communicate knowledge. And then the printing press came
- 1:15:58out and that shifted all of a sudden the people that owned and
- 1:16:01made printing presses. But we also kind of like went
- 1:16:04through, we also kind of like went, there was a, there was an
- 1:16:07old equilibrium where there was kind of the we were in the age
- 1:16:10of the scribes. And then there was a new
- 1:16:12equilibrium where we were in the age of the printing press.
- 1:16:16And there was a period of I'm not a historian.
- 1:16:18There was some period of transition between the two of
- 1:16:21those. And so that would have been kind
- 1:16:22of like with that technology, kind of like the non equilibrium
- 1:16:25period, you're going from one stable state to another.
- 1:16:28But right now, you know, changes like the, the way the disruption
- 1:16:33is so intense that like we don't have time to reach, it's not
- 1:16:36that we don't have time to reach a stable state before things
- 1:16:40have changed again. So that's where I think we're in
- 1:16:43a bit more of like a so I guess non equilibrium is, is not new,
- 1:16:48but non equilibrium sort of punctuated states, the more
- 1:16:51stable states. And I think for at least the,
- 1:16:55for the least the time being, the rate of change is such that
- 1:16:58like there's, you know, This is why we feel like we can't catch
- 1:17:01our breath because like, you know, two months later the
- 1:17:04capabilities have changed. And that that's why I would
- 1:17:08describe it as like a pretty like non equilibrium.
- 1:17:11Makes sense. The reason why I asked that is
- 1:17:14because I feel like depending on what your goal is as a human,
- 1:17:20because that may differ, right? If we're talking with the
- 1:17:24context of winning, winning being that you gain economic
- 1:17:29advantage or you find success in your career, then I think what
- 1:17:33you're supposed to do is different than what?
- 1:17:35Than the person who's just trying to be happy and
- 1:17:38fulfilled. Because I think fulfillment and
- 1:17:40happiness can mean something different in this case.
- 1:17:42If we're talking about AI and what he will do to your job and
- 1:17:45your career. If we're talking with the
- 1:17:47context of winning and being able to succeed in your career,
- 1:17:52I feel like based on that non equilibrium state that you
- 1:17:54mentioned, the only and the best possible option is to adapt,
- 1:17:58right? I think we may have talked about
- 1:18:00this before in a call, but this is like a big wave coming and
- 1:18:06you have the opportunity to see the wave.
- 1:18:08If you're watching and listening to this show, you're actually
- 1:18:11here before the wave or the waves or the some people call it
- 1:18:15a tsunami because it's as massive as a tsunami before it's
- 1:18:18coming. You have the opportunity to
- 1:18:20build a boat, to build a massive ship, to get yourself a
- 1:18:24surfboard, whatever you can, to be able to ride the wave so that
- 1:18:29when it comes, not only are you prepared, you can actually gain
- 1:18:32something from it. And I believe that when you're
- 1:18:35in this state, if you do not act at all, be it for ideological
- 1:18:41reasons or anything else, then there's a very high chance that
- 1:18:45you may regret that decision. And the reason why we always
- 1:18:49talk about it on the show is because ultimately the point of
- 1:18:52bad decisions in the beginning up until now has been, can we
- 1:18:56educate people on the latest in the world of technology, allow
- 1:18:59people to use these tools to help themselves in their career?
- 1:19:02So, so many people ask us all the time, they're like, why are
- 1:19:04you talking about AI so much? We're like, because it's
- 1:19:07literally the biggest thing happening in our lifetime and
- 1:19:11perhaps in the history of humanity.
- 1:19:14You cannot avoid it. It's not just something that
- 1:19:16will turn a little bit of the way we work and change it.
- 1:19:21It's going to redefine economics as a whole.
- 1:19:24It's going to redefine labour as a whole.
- 1:19:27It's going to redefine the way we communicate with each other.
- 1:19:30So many things. It's unavoidable and we have a
- 1:19:36responsibility to together as humans ride this wave and figure
- 1:19:44out where we take it. And I believe every human has a
- 1:19:48responsibility to play a part in this.
- 1:19:51Yes, on an individual level, we don't have the same impact as
- 1:19:55Anthropic, right? Comparing it to one of the, if
- 1:19:58not the biggest, but one of the biggest companies in the world
- 1:20:01of AI right now. But I do believe that we all do
- 1:20:04play a role and. Our goal has always been helping
- 1:20:08people figure out what to do in these times, which is pretty
- 1:20:12challenging, and it's not as easy as just having one single
- 1:20:16podcast and saying we figured it out.
- 1:20:19Yeah, no, absolutely. I mean, there were a lot of
- 1:20:21things there I really liked. The adaptation is absolutely
- 1:20:26true. Like going back to the climate
- 1:20:27change analogy, like we as humans are way more adoptable
- 1:20:31than other animals and like the climate gets hotter, we can make
- 1:20:34have more air conditioning, but then the animals that can't
- 1:20:37adapt are going to like are going to going to die off there.
- 1:20:41So, yeah, when things are changing, like adaptability is
- 1:20:44key. And it's that's where it's not
- 1:20:46just like a how smart you are, whatever.
- 1:20:49It's like, what's your tolerance for change and, and, and for,
- 1:20:52for, for risk? Because you know, there's some
- 1:20:54people that are like, very good at, like, you know, following
- 1:20:59an, an academic program and getting to a point where they
- 1:21:02have a very stable job. It's like a lawyer, But the
- 1:21:05thing that made them success for a doctor, you know, which took,
- 1:21:08you know, 10 years of residency, But the, the personality traits
- 1:21:11that allowed them to do that were one of like, I'm just going
- 1:21:14to focus on this through task and I'm going to execute on it
- 1:21:1715 years. And that same personality trait
- 1:21:20might very well be working against them.
- 1:21:23Now, it's not entirely fixed. Like there is sort of like a
- 1:21:27mindset shift, I think where people can go one one way or the
- 1:21:32other. But but yeah, the that that that
- 1:21:37adaptability is key. And I think I was talking about
- 1:21:39this with one of my managers yesterday that I think one of
- 1:21:42the most important meta skills to teach people is try to teach
- 1:21:46them how to use AI to teach themselves.
- 1:21:51I was talking to manager and he, and he said, and I, I hadn't
- 1:21:54heard of this, but it made me, it instantly made sense and I
- 1:21:56wanted to do it. Is that like, you know, one of
- 1:21:58his personas, like when he was trying to, when he's trying to
- 1:22:02get better as a, as a manager or get this, he was using this for
- 1:22:06one of his classes. He had a Claude project or a Co
- 1:22:10working project that was sort of like he gave it the persona of a
- 1:22:12coach and he was like, you know, read, read these books and be
- 1:22:15the coach and you try to help me to get better at at doing
- 1:22:18something at doing this thing. I thought that was like a really
- 1:22:21like, if you can teach, because assuming that I mean there's
- 1:22:27like there, clearly you should be learning from people as well.
- 1:22:30But I think this is sort of an underrated meta skill that there
- 1:22:34I've learned so many things from Claude and AI and I think not,
- 1:22:39but not everybody. And I, I do have some friends,
- 1:22:41people that I know that are just really curious and they're like,
- 1:22:44Oh, I love AI because I can finally learn anything.
- 1:22:46You know, it's like they, they can talk to experts of all of
- 1:22:48all sorts, But I think for some people that hasn't really
- 1:22:52clicked that is like, yes, it's, it's something that kind of can
- 1:22:55take over some of your skills, but it's also the same thing can
- 1:22:58allow you to learn, learn new skills and and learn your
- 1:23:01knowledge. So yeah.
- 1:23:05It can. What you said was really
- 1:23:07actually interesting. It might take away something off
- 1:23:12your plate, but it can add to at the same time.
- 1:23:15So it can automate the things that you find repetitive or
- 1:23:19boring. But it can also teach you or
- 1:23:22help you and assist you to create things that you wanted to
- 1:23:25do things with better value things in shorter amount of
- 1:23:28time. And I think what people also get
- 1:23:31wrong is that AI is going to replace this part, or I don't
- 1:23:35like AI art, or I don't like to watch AI movies.
- 1:23:39Nobody is forced to use AI in every single facet of their
- 1:23:43life. You can choose what are you
- 1:23:45going to do with this technology, of course, in your
- 1:23:48work, and it's something that you want to deliver.
- 1:23:50You want to be as efficient as possible, do it with the highest
- 1:23:54quality. But if you are a person who
- 1:23:56enjoy live music or if you enjoy going to the cinema and watch
- 1:24:00movies, you can do that. And I think there is going to be
- 1:24:02huge values in art that are made by human or in the craft that is
- 1:24:07made by human in the future. So we shouldn't think of it as
- 1:24:11an absolute that Oh, OK, if if AICA, if I'm adopt, everything
- 1:24:15has to be around AI and everything has to be AI
- 1:24:18generated or AI made, I'm not going to have access to anything
- 1:24:21human made. But I think the balance and the
- 1:24:23combination of both, it would matter a lot in the future as
- 1:24:26well. Yeah, no.
- 1:24:28And I, I really like the examples you gave around art.
- 1:24:31And I think those are things that are like, I'm not an
- 1:24:33artist, but something that I, I think about and it's probably,
- 1:24:36you know, relevant to your audience because we can.
- 1:24:39I, I think we can make some progress in terms of like
- 1:24:44understanding certain certain nuances there because I, I,
- 1:24:48yeah, I remember especially like when stable Diffusion cart
- 1:24:50started coming out and these, these images were getting really
- 1:24:53good. And then there was this feeling
- 1:24:54of like, Oh no, like, you know, AI has taken over art.
- 1:24:57And that that comes from partly comes from like a naive view of
- 1:25:01art is just like just the artifact or just the pixels.
- 1:25:04And like, we already know that this isn't, isn't the case
- 1:25:08because, you know, you could have, you could have a
- 1:25:10reproduction of a Mona Lisa, which is almost, you know, at
- 1:25:14every breaststroke level the same.
- 1:25:15And it's totally worthless because there is the original,
- 1:25:20which has the history of, you know, the history and the artist
- 1:25:25and everything associated with that.
- 1:25:26And like it's been through, you know, it's, it's the, it's the
- 1:25:29original and like that, that matters to people and it matters
- 1:25:32to society and its value reflects, reflects that, that
- 1:25:35that matters. And I think that's like that is
- 1:25:40going to be so clearly there is like a meaning component to art,
- 1:25:43which had is it a lot of a lot of value.
- 1:25:46And then your, your comment about live music also made me
- 1:25:49think that there's like also an experiential nature of art where
- 1:25:54like I listen to music at at home all the time.
- 1:25:56But, and maybe I, I, if I, if I didn't have a record player or
- 1:26:01anything at all, actually, I don't have a record player.
- 1:26:04If I didn't have Spotify at all, you know, maybe I would go out
- 1:26:07to more live music events. But still in my mind, I don't
- 1:26:10think that like, I'm not thinking about my Spotify as
- 1:26:13like a substitute for live music.
- 1:26:15Like that's a different experience.
- 1:26:16That's something I wanna go out that has value in it beyond just
- 1:26:22the like, you know, I'm listening to this, these sound
- 1:26:25waves or, or whatever. I I think there is a sacrifice
- 1:26:29for art. You either sacrifice time, you
- 1:26:31spend some time to go to a museum or to go to cinema and
- 1:26:34there is a sacrifice of money you would pay for something.
- 1:26:37I can listen as much as I want to AI generated music, but I
- 1:26:43would pay to go to a concert for a DJ to play electronic music to
- 1:26:48me I still pay for that. So I think when art is created
- 1:26:52by an artist, there is meaning and expression through that art
- 1:26:57and I think that would make it very different.
- 1:26:58At the same time, I believe AI generated content for commerce
- 1:27:03is amazing. You can generate audio and music
- 1:27:07for your ad which will work perfectly.
- 1:27:09Probably you don't need art for that at that level.
- 1:27:12If you are having let's say a Shopify store and you are making
- 1:27:16photos of your products, you can make a generated photo that will
- 1:27:20serve the purpose. But if you are making a photo to
- 1:27:24express yourself, I think that would take it to another level.
- 1:27:27And I think there are differentiation between these
- 1:27:29two which humans can understand the moment when they see a piece
- 1:27:32of art or when they listen to it.
- 1:27:36Right. Yeah, no, that's exactly right.
- 1:27:37And I think like part of it because like you know for
- 1:27:39e-commerce, you know it is it's a utilitarian like piece, piece
- 1:27:44of art. Like, yes, it has some aesthetic
- 1:27:46quality and it takes, it would take kind of an artist to, to
- 1:27:49produce that, but it's ultimately designed to like
- 1:27:52sell, sell a product and any, the image that sells their
- 1:27:55product best is kind of the best for that.
- 1:27:59But like, I think it really matters if yeah, if if like, you
- 1:28:06know, if a songwriter is, is writing a song about love, I
- 1:28:09think it, it totally matters like that that person, you know,
- 1:28:13lived and and loved and they lost and they felt that
- 1:28:16heartache and like an AI writing a song about love that's never
- 1:28:19experienced love is that is like a vacuous, vacuous shell, you
- 1:28:26know, like it doesn't. I don't, I think a lot of people
- 1:28:29feel like, you know, that's not, that's not art.
- 1:28:31And I think that's totally, totally, totally valid because
- 1:28:35the most art is designed to carry an expression or of the
- 1:28:40experience of the artist. And so if there's, if you have a
- 1:28:43faked, like a, like a faked expression, you know, expression
- 1:28:46without any, you know, experience, then it's yeah, it's
- 1:28:52like, it's a fake. It's a different kind of fake.
- 1:28:54Yes, because it's like it's not, it's not a replica, but it's
- 1:28:57kind of like a, it's a different kind of fake.
- 1:29:00I completely agree with both of you on this.
- 1:29:03This was a topic that stayed with me when I did research for
- 1:29:07another podcast about the concept of creativity and art
- 1:29:10and what is the meaning of art. And I've come to the conclusion,
- 1:29:16especially with a lot of examples that I can bring to the
- 1:29:18table right now, that there is a subliminal energy with art that
- 1:29:26is received very well with people.
- 1:29:28There is the the fact that there's an energy put through
- 1:29:31the art form, whatever it may be, by the artist, by the
- 1:29:35creator, and it's all of their feelings in that moment as
- 1:29:40they're creating the art. All of the pain, the suffering,
- 1:29:43the love, the hate, the anger that is sent through that
- 1:29:46artwork. And we receive it when we're
- 1:29:48listening to the song, when we are watching the film.
- 1:29:52And it's, it's perhaps one of our strengths as human beings.
- 1:29:56I have proof of that because we, we look at this stuff all the
- 1:30:00time and we want to talk about it.
- 1:30:02We're just looking at Star Wars, the score from Star Wars.
- 1:30:06Duel of the Faith by John Williams.
- 1:30:09Yes, the fact that if you look at the behind the scenes, which?
- 1:30:131 is. That you guys guys sing it for
- 1:30:15me. So the the first the first one
- 1:30:17in the the first one in the sequel.
- 1:30:19So the one that was made in 2001?
- 1:30:25You know that one, right? With Darth Maul.
- 1:30:27With Darth Maul, Yeah. Well, I don't, I don't know the
- 1:30:30where the prequels, the music for the prequels as well, but.
- 1:30:33I know the movie, I'm not judging you at all right now at
- 1:30:36all, at all. But that's just one example.
- 1:30:40There was a behind the scenes of him with the orchestra and the
- 1:30:44amount of takes that they're doing, the amount of attention
- 1:30:47he's putting in, like hundreds of them just to get the pitch
- 1:30:51right, the feeling right and the emotions and a number of days
- 1:30:54they might must have practiced. Or another thing we were just
- 1:30:57talking about is the duel inside Star Wars itself, The actor of
- 1:31:01Darth Maul spending how? Many 20 hours a day in this
- 1:31:04studio for three months for only 9 minutes of screen time.
- 1:31:09And you compare that with some of the recent Star Wars films
- 1:31:12where the duels and the battles are immediately forgotten after
- 1:31:16people watch them because they probably spend much less time on
- 1:31:19it. So I feel like when it comes to
- 1:31:21art as well, it's about the energy and the amount of effort
- 1:31:25and care we put into it. Which brings me to my next
- 1:31:28point. I think with AI, this whole
- 1:31:31concept is just exaggerated. If we had 5% of the entire
- 1:31:37world's population that achieved mastery, whether he was the guy
- 1:31:40with the bow and arrow in the Genghis Khan army, or whether he
- 1:31:44was the guy who made the score for Star Wars or Christopher
- 1:31:48Nolan making a film. The guys who really put in the
- 1:31:50time and care and effort to their craft, they're just going
- 1:31:55to do the exact same thing with AI on a much larger scale.
- 1:31:59And the people who actually released substandard art or
- 1:32:04substandard films or substandard poems and all the, they're going
- 1:32:07to do exactly the same thing with AI because of the fact that
- 1:32:11they didn't actually put in the time and the effort and energy.
- 1:32:13AI is going to come in and because of whatever reason that
- 1:32:17they may have, they're just going to quickly create
- 1:32:18something without the amount of care that they should.
- 1:32:22And it's going to be considered as AI slop.
- 1:32:24So I feel like that's where we're going to be headed.
- 1:32:26I'm not too worried about it myself because I know humans who
- 1:32:31do care about their craft are still going to exist.
- 1:32:34The only challenge is AI allows us to create much faster, which
- 1:32:39is why we're going to be flooded with a lot of the AI slop that
- 1:32:43people talk about. Yeah, Yeah, No, definitely.
- 1:32:47I mean, like in any sort of economic market, yeah, that if
- 1:32:51there's more supply, yeah, it increases the source, the the
- 1:32:55supply of art, which then for the the lower quality stuff
- 1:32:59makes it less valuable because you just have have more of it.
- 1:33:02But it can also mean the stuff that's really great.
- 1:33:05It can make it kind of more valuable because it, it stands
- 1:33:07out or it's scarce or it's more scarce.
- 1:33:10So, yeah, I, I, and that's where like any, any one of these
- 1:33:14things, when you kind of zoom into it, you realize there is,
- 1:33:17you know, quite a bit of nuance. And I like your example of like,
- 1:33:20you know, the great artists people are going to be, are
- 1:33:23going to adapt and you're going to use these tools.
- 1:33:25And I think it's just, it gives, I, I think a lot about like the
- 1:33:29concept of leverage, which is kind of like, how much do you
- 1:33:31get out of something versus how much you, you put in?
- 1:33:36And it's a powerful tools give give people a lot of a lot of
- 1:33:41leverage. And so I think for the people
- 1:33:44that are really, you know, I mean, this is already happened
- 1:33:49with like, you know, mass with, you know, television all of a
- 1:33:53sudden, like, you know, some of the best performers, they
- 1:33:54weren't just performers in their local, you know, their their
- 1:33:59local theater or something like all of a sudden they could be
- 1:34:01broadcast across the entire entire country.
- 1:34:04And so the the people that were maybe some of the best now have
- 1:34:09like a, you know, a ton of way more influence and power than
- 1:34:13they could have before because it's there was an amplification
- 1:34:18that happened. That also does mean that.
- 1:34:20Yeah, there were some, you know, it probably put some artists out
- 1:34:24of business that were less talented, but they had their
- 1:34:27niche and their local like live music venue or the theater.
- 1:34:31And like, you could argue that like maybe some of that's not
- 1:34:35some of that's good, some of that's bad.
- 1:34:36Like, I do think, you know, we lost something when, when Amazon
- 1:34:40replaced like all the small book shops or even most, most, most
- 1:34:44small booksellers, even though Amazon functionally is, you
- 1:34:47know, a much better bookstore. They have every.
- 1:34:49They have every book. But is functionality the only?
- 1:34:53Sorry, is the functionality the only metric that we should
- 1:34:56measure? So if you want to expand on that
- 1:34:59Amazon example, which I love, when you mentioned, should we
- 1:35:03only measure by efficiency and productivity and performance and
- 1:35:06we say, OK, since this is more efficient, this is the best,
- 1:35:09let's replace. No, I think, I think you're
- 1:35:16absolutely right. It's going to depend on what
- 1:35:17like the good in the the service is.
- 1:35:19I think for some, for some goods and services, you know, like,
- 1:35:26you know, a toothbrush or there's some things like, you
- 1:35:28know, where you, where you just want a replica of an object
- 1:35:32with, you know, 00 defects or something.
- 1:35:35But then there's a lot of cases where there is, it's, there is,
- 1:35:41yeah. You want some meaning or you
- 1:35:43want some experience or you want some, some quality to it, which
- 1:35:46is, which is hard to measure. And I think that this is
- 1:35:49something like when I think about like there's this kind of,
- 1:35:54you know, if you ask the question of like, oh, you know,
- 1:35:57will AI be better than everyone and everything?
- 1:36:01And like, obviously the answer to that is kind of is, is no,
- 1:36:04you know, like I don't, I don't, I just dropped my daughter after
- 1:36:08preschool this morning. Like, I don't want a robot
- 1:36:11teacher because like, you know, there's nothing like, you know,
- 1:36:15my, my daughter's 4. Like you need to, you need to
- 1:36:17take her out to the playground, make sure she drinks milk and
- 1:36:20then and then and then bring her back in.
- 1:36:22And like a human does that job totally fine every day.
- 1:36:25Like, and I think there are lots of like a lots of like, if you,
- 1:36:31you know, take one a piece of music.
- 1:36:34I mean, you guys already mentioned this one from Star
- 1:36:36Wars or for for me, I mean, there's so many pieces of music
- 1:36:40or art that you like that come to mind.
- 1:36:43It's like favorites, but you know, masterpieces like say like
- 1:36:45box first cello concerto or or the matrix or something.
- 1:36:50Say the cello concerto is like, you know, I ask myself, like,
- 1:36:52could AI make that better? And it's like, like, not like,
- 1:36:58you know, now it's like a masterpiece, you know, the way
- 1:37:02it is, like AI could make a different version of it.
- 1:37:04But like, there is a sense in which it's kind of like it's,
- 1:37:07it's optimal. And I think there are lots of
- 1:37:09things that humans do now, which is somewhere, somewhere between
- 1:37:15like good enough and like optimal or like good enough up
- 1:37:19to some quality bar where the, the marginal benefit of it being
- 1:37:26sort of like automated or sort of like lower cost.
- 1:37:28It's not really worth the fact that it you know you did lose
- 1:37:33something when you when you had a machine created.
- 1:37:37I actually have something very interesting and important.
- 1:37:43It's a question for myself. I don't have an answer to it.
- 1:37:46I am more optimistic, just like you mentioned, in the short
- 1:37:50term. When I say short term, I'm
- 1:37:52talking about the next few years because I genuinely believe the
- 1:37:56people who adapt and learn the tools, they can use this to
- 1:37:58their advantage. And as we do currently have in
- 1:38:02the world's population, lots of skilled people in completely
- 1:38:07different domains, those people can use AI to enhance their
- 1:38:11work. They've spent years mastering
- 1:38:12their craft. They're going to use AI to
- 1:38:14enhance their work. But what the fuck will happen to
- 1:38:18education for the next generation?
- 1:38:21Because there's no way in hell universities or colleges are
- 1:38:25right now adapting at the same speed Anthropic is updating
- 1:38:29their models. I know for a fact I went and I
- 1:38:33studied computer science. I passed everything right.
- 1:38:36I don't remember 99% of what I did because they had us write
- 1:38:41code on a piece of paper as an exam on a piece of paper.
- 1:38:45Now not only do we not need to write code on a computer, it's
- 1:38:48done for us. That was this is less than 10
- 1:38:52years ago. I trust that humans at this
- 1:38:57point, us our age a little younger as well, have spend the
- 1:39:01time to learn tools and be masters at their craft.
- 1:39:06But the next generation that is going to be born into a world
- 1:39:10with AI, with our current educational system everywhere in
- 1:39:14the world, how will they learn what we've learned?
- 1:39:18Do they have to learn it or will will they let AI do everything
- 1:39:23for them? Because at that point, you
- 1:39:24cannot argue the same point we're arguing right now that
- 1:39:27there are masters. Do they need to learn how to
- 1:39:29code? Because if they don't learn
- 1:39:31anything, any of these skills, and all they do is rely on the
- 1:39:35calculator. Let's talk about the calculator
- 1:39:37example, right? It's the simplest one.
- 1:39:39If you have a calculator, you no longer need to learn addition
- 1:39:42and subtraction and everything because it's done for you.
- 1:39:45If a kid is brought up with a calculator, and I was, I was
- 1:39:48like halfway through that. Half of my teachers were against
- 1:39:51it, the other half were for it. But because we relied on the
- 1:39:54calculator, slowly I got worse at multiplication and division.
- 1:39:58The complex ones, you were much better than me because in your
- 1:40:02time they were really harsh on you guys using.
- 1:40:04Calculator we couldn't use. I believe the same thing will
- 1:40:07happen with AI. So what happens to the future
- 1:40:09generation? Yeah.
- 1:40:12Well, I would say like like you, you got worse at multiplication
- 1:40:16as like. Thank you, an individual thank
- 1:40:18you. Somebody, if somebody put you
- 1:40:21out in the desert, like told you to multiply 2 numbers, but you,
- 1:40:25if you think of like you as like EU plus calculator symbiote or
- 1:40:29whatever, like you guys got that, that got way better.
- 1:40:33And so this is something that I like, this is a point of view
- 1:40:36which I think, you know, it sounds, which I think it becomes
- 1:40:40sort of increasingly meaningful is to think about the augmented,
- 1:40:44the augmented human. Like the word conventionally
- 1:40:47that you would use is like Cyborg or that one word you you
- 1:40:50might use Cyborg has all these like weird, you know, sci-fi
- 1:40:55connotations to it. I think we almost need like a
- 1:40:57new word to saying like human computer symbiote is like also,
- 1:41:01yeah. You know, yeah, we need a new
- 1:41:03word. Yes.
- 1:41:04New new wordy, but that sounds sounds lame.
- 1:41:07But but yeah, that that that package.
- 1:41:12I do think like it, it got got better and now did did you did
- 1:41:18we but did you? Is there is there a way like I
- 1:41:21mean like I've gotten I used to know my way around the town that
- 1:41:25I grew up in before Google Maps, you know, I knew how to get
- 1:41:28everywhere. And now in LA, you know, I can
- 1:41:30barely get to Taco Bell. So like that, that is something
- 1:41:36that that I've lost and like you, but at the same time, I can
- 1:41:41get anywhere if I have if I have my maps in it in a city with,
- 1:41:44you know, 2020 million people, that's all a bit.
- 1:41:49So I think clearly like, you know, the next generation of
- 1:41:52people, like they will have these incredible tools that
- 1:41:54augment them, but they will also like have not gone through the
- 1:41:57experience of like learning things kind of the hard way.
- 1:41:59And I do think there is something lost there.
- 1:42:02And this is like, we could we could talk for two hours about
- 1:42:05kind of like education and young people.
- 1:42:08And I think it's part of the conversation that's public
- 1:42:11conversation that's really under discussed because like we as
- 1:42:14adults were like very concerned about what's going to happen to
- 1:42:17us. And there's like not near enough
- 1:42:19concern. I think about like how this is
- 1:42:21going to affect kids. I mean, same thing happened with
- 1:42:23social media. Like social media came out and
- 1:42:25like, now we know that like, you know, for a 13 year old girl to
- 1:42:29be constantly bombarded with like pictures of models on their
- 1:42:32phones, like fucks up their self esteem in a lot of ways.
- 1:42:35Like not a surprise. But like, people weren't really,
- 1:42:38people weren't really thinking about it.
- 1:42:40But like, yeah, once, once they can try to measure their measure
- 1:42:46their worth in terms of like likes on a platform that just
- 1:42:49rewards, you know, superficiality then and ignores
- 1:42:56all the other dimensions to that to that young person is like,
- 1:42:59yeah, is, is bad. And so I, I, we're not going to
- 1:43:04be able to do it justice here. I think it's some of the people
- 1:43:08should talk about more. I mean, the one thing, one
- 1:43:11perspective, and I think this is something where like there's
- 1:43:13probably a lot of people that are going to be able to speak to
- 1:43:15this better than me. One thing I've thought about it
- 1:43:18though, is as as a parent of like a really young child, like
- 1:43:21I have a daughter. I have a daughter that's three.
- 1:43:23And so she's. She like she.
- 1:43:26There's it's a very different experience if you have a if you
- 1:43:29have a kid that's like, you know, in high school, there's
- 1:43:32just graduating college because then like they are the world
- 1:43:36hasn't yet adapt. I assumed, I assumed that by the
- 1:43:39time my daughter is 18 that like, you know, the world has
- 1:43:44either blown up or we've figured out or, or things that we've
- 1:43:48like society has has gone through a lot of adaptations
- 1:43:51where like, you know, you're no longer reliant on your career to
- 1:43:55make money. Like maybe money is being
- 1:43:56distributed in some other other method.
- 1:43:59The people that are going through this change before
- 1:44:02society has calibrated our in a rough spot.
- 1:44:04And I do think that like, you know, like I do feel for people
- 1:44:07that are just entering that just finished like you from like, you
- 1:44:11know, 10-15 years ago when you just finished your science
- 1:44:13degree and you you spend a ton of money and time doing this.
- 1:44:17And then like nobody's hiring entry level software engineers
- 1:44:21anymore. Yeah.
- 1:44:21Yeah, V boot. Yeah, so that that that sucks
- 1:44:26and like I don't I don't know how to you know, people are
- 1:44:30going to have to figure out how to I think should think about
- 1:44:33that problem and care about it. But me as like a father of a
- 1:44:36three-year old daughter. I mean what what I think it one
- 1:44:41thing I think it it clear. I think clearly that like
- 1:44:45traditional education and like labor is going to be less
- 1:44:49valuable when she's grown up than than it is now.
- 1:44:54And but something I like this, this, this goes into a subject
- 1:44:59which I also care a lot about, about kind of like, how do we
- 1:45:02have humans? How do we find like meaning in
- 1:45:04our own life, particularly for those of us that found a lot of
- 1:45:09like meaning through our work. And like things I want to, you
- 1:45:13know, I, for me, things that I want to give my daughter a lot
- 1:45:18is like the ability to sort of like appreciate, really enjoy
- 1:45:23parts of life that are not working, have less focus on
- 1:45:26achievement. And she's gotten the best grades
- 1:45:28and more focused on. Does she have great friendships?
- 1:45:31Is she a kind of person? Does she like dance and play
- 1:45:39music? And so like I, you know, I like
- 1:45:41playing music even though there's so many other guitar
- 1:45:43players out there. Better for me, better than me,
- 1:45:45but I don't play it because I'm trying to be the best.
- 1:45:47I play it cuz like I like that, you know, the experience of
- 1:45:50playing music. So that's something that's
- 1:45:53always been important for, I think, kids.
- 1:45:55But for me, it makes it like more important that like,
- 1:45:58because that's something that I see will clearly still be
- 1:46:02important to her experience, even if the job market is
- 1:46:06totally changed because that one's much harder to predict.
- 1:46:08Like maybe there'll probably be still be some jobs, but nobody
- 1:46:10has any idea what jobs there will be.
- 1:46:15Yeah. One thing that you mentioned
- 1:46:17that made me think was that how people nowadays tie their
- 1:46:21identity to their profession. Yeah, they will call themselves
- 1:46:25I'm a 3D artist, I'm a texture artist, I'm a developer.
- 1:46:28That's the first thing they say when they see someone, right?
- 1:46:30They say it now that you can become anything.
- 1:46:35What do you become now that you can?
- 1:46:37Faras, are you drinking Pepsi? Sponsored.
- 1:46:42I know, I know, Chris is a Coke 0 fan, so I'm just pulling this
- 1:46:44up just so that everybody in the comments right now.
- 1:46:49I'm not getting paid for this. Let's just do like a little
- 1:46:54fight or argument in the comments.
- 1:46:56Pepsi 0 or. Coke 0 Let's see, let's see
- 1:46:58who's going to win. Sorry to cut you.
- 1:47:00Off no. But what I was saying that
- 1:47:02people tie their identity to their profession.
- 1:47:05They will say I'm a 3D artist, I'm a musician, I'm a developer,
- 1:47:08I make websites. When you can do many things,
- 1:47:13what can you identify yourself as?
- 1:47:16You can now use cloud code to make a website or make an app
- 1:47:20for yourself. You can use image generation
- 1:47:22tools, you can make music. It would be hard for people to
- 1:47:25find purpose and meaning and or even to have a label on
- 1:47:28themselves that I'm good at this.
- 1:47:31I can do this for a long period of time.
- 1:47:34You see the imagine the next generation have to go through
- 1:47:37this for their entire life. How do we find meaning and
- 1:47:41purpose, and how do we identify ourself?
- 1:47:45Yeah. I mean, I think you'll, you'll
- 1:47:46have to go back to like I am Farrah or I am Farrah or I'm,
- 1:47:50I'm Christopher. Damn it man, then how do we how
- 1:47:52do we like get the girl in the club if.
- 1:47:55Nobody saying that just Farrah or just by the way I will.
- 1:47:59I will just say I'm Farrah. Yeah, but I mean, yeah, I know
- 1:48:06to say, yeah. So the like, you know, part of
- 1:48:11why is saying your career is because it's also like a
- 1:48:13somewhat, it's somewhat identifying, but it's also like
- 1:48:16a status, kind of a status signaling.
- 1:48:18Of course it's a status signaling thing.
- 1:48:20And that becomes less, that becomes kind of less, less
- 1:48:27meaningful. And I, I do think, but I do
- 1:48:30think there is like a mindset shift that people will need to
- 1:48:34go through. I think you can see to some
- 1:48:36extent by looking at people that are not.
- 1:48:40Of course there are people now that are not working.
- 1:48:43Some people are like retired people.
- 1:48:45And even those people will be like, I was an engineer, yeah,
- 1:48:49or something. But then there are people that
- 1:48:54are like, you know, I'm, I'm a mom or I'm, I'm a dad.
- 1:48:58And you'll see this even with people that are quite success,
- 1:49:00you know, have their professional career, like they
- 1:49:02have kids and their Instagram bio changes to like mother of
- 1:49:06Jonathan, husband of, of, you know, Joey And then like, you
- 1:49:13know, engineer at Google or whatever.
- 1:49:16And, and that's because like they found being a parent to be
- 1:49:19so meaningful that like, that's how they want to be identified
- 1:49:23more. Like if they're at a dinner
- 1:49:24party, they're still probably going to be like, oh, I'm, I'm
- 1:49:26an engineer, but at least aspirationally they want to
- 1:49:30identify with that. And that I think points to at
- 1:49:33least one thing that I have confidence.
- 1:49:34And is that like our, you know, the value we have from like our,
- 1:49:39our relationships is something that's going to be very
- 1:49:41important and durable. And I would say like, that's,
- 1:49:44that's something where if you are unsure of like, should I
- 1:49:47invest time in my, you know, friendships in my family?
- 1:49:51Or should I like spend more hours in the library, like
- 1:49:53working? You know, I, I think the value
- 1:49:56of relationships are just going to get, you know, is it, it
- 1:50:00increases? So that that's, that's one,
- 1:50:03that's one way. But there's a lot of aspects to
- 1:50:06this is a really good question and there's a lot of aspects to
- 1:50:10it. Sorry.
- 1:50:11I no, no no I do appreciate what you just said, but I I beg to
- 1:50:15differ on just one aspect and that is personally I'm talking
- 1:50:19about myself. I don't know how many people
- 1:50:20like me exist, but I'm sure there is a demographic out
- 1:50:24there. Being a father, being a son,
- 1:50:28being a brother, all these things mean a great deal to me.
- 1:50:31But ultimately on a day-to-day basis there's a lot of things I
- 1:50:34want to do and most of it has to do with creation of value.
- 1:50:37I love the fact that I can make a video, release it to the world
- 1:50:42and see people gain value from that.
- 1:50:44And I feel like there's a lot of other people who, yes, maybe
- 1:50:49they'll be displaced from their jobs five years from now or 10
- 1:50:52years from now or two years from now, but they'll still need
- 1:50:55something to spend their time on outside of the relationships.
- 1:50:59Something that makes their lives meaningful in a way that they
- 1:51:03are creating value for others. Again, I'm not saying you don't
- 1:51:06create value as a brother or as a husband.
- 1:51:07I'm just saying we already do that.
- 1:51:09Yes, we can exaggerate that. We can have more time with
- 1:51:12family, but ultimately I have a feeling will be incredibly
- 1:51:15challenging for human beings to just rely on that relationship
- 1:51:20and spending time on entertaining activities.
- 1:51:23And the reason why I say that, I am not saying that's what you
- 1:51:25meant. I had that discussion with Fahad
- 1:51:27when we talked about universal basic income and with Elon's new
- 1:51:31name of universal high income. I personally cannot subscribe to
- 1:51:35that idea because I know for a fact that human beings, as you
- 1:51:40mentioned, there's a status thing about us.
- 1:51:44We'd love to signal status, and that is something that has been
- 1:51:48ingrained in us since the beginning.
- 1:51:50Even when we were hunter gatherers, we had a hierarchy
- 1:51:52naturally formed around strength and power, and the strongest men
- 1:51:59of the tribe got to mate with the most amount of women or the
- 1:52:02best women, because they could hunt and they could defend other
- 1:52:06people. Right, You're going.
- 1:52:09To do it. Maybe that's what we're going to
- 1:52:15go back. To I'm a hunter gatherer.
- 1:52:19What I'm trying to say here is that I was watching the Crown
- 1:52:26series the other day and you could see when someone was
- 1:52:29trying to reach the Queen. They show this scene and it's
- 1:52:32amazing by the way you have to watch it.
- 1:52:34It's like a side scrolling scene, the cameras just panning
- 1:52:36across walls and you see all these people.
- 1:52:40I, I think the demographic of their age is from 20 years old
- 1:52:44all the way up to 80 years old. And it's different rooms of
- 1:52:47people just hooking up cables to different holes, trying to
- 1:52:51connect the entire written to to each other when you're trying to
- 1:52:55call. And when they when someone wants
- 1:52:56to reach the queen, it has to go through, I think 20 people and
- 1:53:00reaches the most senior guy who's like this sleepy 80 year
- 1:53:03old guy who has the authority to connect someone to the queen.
- 1:53:07All of those people lost their jobs, every single one of those
- 1:53:10people, right? And if you told them that this
- 1:53:13would happen to you, they wouldn't believe it in the
- 1:53:15beginning, but it happened. And no one at that moment could
- 1:53:19figure out what they would do. You know, you had a 20 year old
- 1:53:22woman and an 80 year old man. Their skill sets must be
- 1:53:26completely different from one another.
- 1:53:28And you cannot predict what they could do in the future.
- 1:53:30And neither could you tell them and make them understand the
- 1:53:34kind of jobs we have today in 2026.
- 1:53:37I mean, try to explain to that 80 year old man the concept of
- 1:53:41an Only Fans account, like a girl on only Fans who's making
- 1:53:45millions of dollars. Try to explain to him the
- 1:53:48concept of what we're doing right now and the fact that we
- 1:53:51can broadcast this to the entire world and this podcast can be
- 1:53:54more famous than a Prime Minister eventually going live
- 1:53:58on on Facebook or on Instagram. Like so much has changed.
- 1:54:02And for those guys, it was impossible to even fathom the
- 1:54:06kind of jobs we have today. What I think will happen, as you
- 1:54:10mentioned earlier on this podcast, is that humans have
- 1:54:14always been great at adapting. This is going to be the hardest
- 1:54:17round of adapting that we've ever done because it happens so
- 1:54:22fast. But I feel like we can always
- 1:54:26try to guess what will happen next, but it's incredibly
- 1:54:29challenging to do so. I can give you another example.
- 1:54:32This one, I I can even try to have emotions about it, but it's
- 1:54:37it's almost 100% going to happen.
- 1:54:40Robotaxis are going to be released very soon.
- 1:54:42I think Elon's already releasing them.
- 1:54:45Yeah, I think you guys may agree with me, but in 10 years time,
- 1:54:49can we all agree that all cabs and Ubers and everything will be
- 1:54:52just robotaxis? Like probably no humans
- 1:54:56operating them at that point. Sure.
- 1:54:58Yeah. I mean, well, like, like, I mean
- 1:55:01cabs still, even after Uber came out, like cabs still kind of
- 1:55:04exist. But like, yeah, they're very,
- 1:55:08they're not very used. So yeah, I think it's the thing
- 1:55:10where it's at like 90 maybe if you want some luxury.
- 1:55:13And this I think will be true in general.
- 1:55:15It's like if you want to pay for some, if you want to pay more
- 1:55:17for some luxury experience, you could get like a human to drive
- 1:55:20year round, yeah. That would be like the most.
- 1:55:22Expensive service, yeah. But but yeah, I agree that like
- 1:55:26yeah, 9095 percent, 99% are going to be.
- 1:55:30Robber taxis, yeah. So in my opinion, there has to
- 1:55:34be a replacement for all of those people who are driving
- 1:55:38these cars right now. Ubers, you know, taxis that that
- 1:55:43you mentioned are still around. But we may not know what that
- 1:55:46job is now, and it may be impossible to even try to guess
- 1:55:51what it is, because we'll have to wait and see what we'll come
- 1:55:54up with. Yeah, yeah, I know.
- 1:55:58I, I, I totally agree with that. And I, I sympathize with that
- 1:56:03whole thing and actually kind of like worrying about like my
- 1:56:06identity as I don't know. Well, at the time I was, I was a
- 1:56:14physics graduate student. So like my, who I was, it'd be
- 1:56:16like, oh, I'm a physicist, physicist or I'm a physics
- 1:56:19student. And like, thinking about that
- 1:56:21identity being disrupted by AI is part of what like got me,
- 1:56:26launched me into, into this direction of like technology
- 1:56:29like 10 years ago because I found it to be like so kind of
- 1:56:33disconcerting and troubling and has led to like, sort of like a
- 1:56:36decade of soul searching. And that is part of why like I
- 1:56:40started, I think when, when I'm thinking about the future, there
- 1:56:43are some things where like nobody has any fucking idea
- 1:56:46what's gonna happen. And then there are some things
- 1:56:48that I feel fairly confident about as long as like, you know,
- 1:56:53humans are, there's not some extreme scenario like, you know,
- 1:56:56humans are wiped out. And these I sort of think about
- 1:56:59is like islands of islands of understanding or like islands of
- 1:57:03sort of certainty. And that's part of why I started
- 1:57:05with the relationship thing because I, I'm confident that
- 1:57:08like, if humans are still around, you know, the
- 1:57:10relationship between a mother and son or whatever is still
- 1:57:14going to be like, you know, pretty much just as important as
- 1:57:17it is today. Like, just as it was like, you
- 1:57:21know, you go back, you go back, you know, a couple 100 years and
- 1:57:23that relationship was like really important.
- 1:57:25So that's part of what I, I, I started there.
- 1:57:28But because the, the, with the work one, yeah, you're right.
- 1:57:30We don't know what jobs are going to be.
- 1:57:32And I think, I think in some cases there will be new jobs,
- 1:57:36but I also think there's going to be a lot of, and this is 1.
- 1:57:39There's like lots of things, there's lots of like side
- 1:57:42effects of AI where even if the future in a lot of ways is
- 1:57:45better on sort of like measurable, kind of like income
- 1:57:48goods are cheaper. There's, there's going to be a
- 1:57:51lot of these like sad, sad things that happen like, you
- 1:57:54know, like, yeah, like college students entering the market.
- 1:57:57And I think that there will also be like people who they lose
- 1:58:00their job and they're not able to adapt and they become
- 1:58:05depressed and then they become addicted to opioids or
- 1:58:08something. And they're like, you know,
- 1:58:10they're in, they were a trucker and they, they live in middle
- 1:58:13America and they, they, they had an identity.
- 1:58:19It's like, Oh, I, I drove trucks so I can provide for my family
- 1:58:21and that is this meaningful? And now they're like an
- 1:58:23alcoholic. Like they're getting, they're
- 1:58:25getting a four $3000 check a month from the government or
- 1:58:28whatever. But like they didn't have to
- 1:58:30work for it. And I, I also share your
- 1:58:36skepticism or your, the sentiment about universal basic
- 1:58:39income or high income. It doesn't solve, it's clearly.
- 1:58:44I think something like that is going to need to.
- 1:58:48Like. We're going to have to have some
- 1:58:50part of that, but we need to, we need to have more if we want it
- 1:58:53to be like a healthy society. Because I, I think one of the
- 1:58:57things that about status and all these things, like one of the
- 1:59:01underlying principles is like, can you make your position in
- 1:59:06life? Can you improve your quality of
- 1:59:08life according to however you think about quality of life and
- 1:59:12you improve it, but through your own efforts, can you, can you
- 1:59:16move up? And like we, we've seen what
- 1:59:18happens in societies where like that mobility is taken away from
- 1:59:21you. Like that's, that's sort of how
- 1:59:23things function in, in communist society, right?
- 1:59:27Where it's like, you know, in like the Cultural Revolution in
- 1:59:31China or something, like, you know, doctors and professors,
- 1:59:33they're, they're forced to be farmers, right?
- 1:59:37And, and it's depressing as hell.
- 1:59:39I mean, setting aside all the other bad parts of, of
- 1:59:42communism. So there is something about Ubi
- 1:59:44that smells like communism. And there's, there's going to
- 1:59:49need to be I, but I, I do think that, you know, meeting people's
- 1:59:52basic needs, like physical needs like that, that is an important
- 1:59:55sort of like base level. But on top of that, there's
- 1:59:58there's going to need to be some mechanism by which.
- 2:00:02If for a lot of people isn't, if it's not the normal job,
- 2:00:05there's, there's still some way that they can do useful and
- 2:00:08important work and then sort of improve their own, their own
- 2:00:12station. Because I think that's a really,
- 2:00:14that is a fundamental need. And like, meritocracy is, you
- 2:00:21know, some of that is, is is really important because not
- 2:00:25only can you gain, but you can also, you can also lose if you
- 2:00:29don't. And like, having skin in the
- 2:00:30game is really important. If you don't have some skin in
- 2:00:32the game, then yeah, just lie on your couch.
- 2:00:34Then you, you know, you end up in this scenario where like,
- 2:00:37people are alive, but they're, you know, binge eating Doritos
- 2:00:41and watching Netflix and like, their needs are met, but it's
- 2:00:45just a sad existence. They don't have to do anything.
- 2:00:49I think that's it. So I I.
- 2:00:50That's a scene in Wali, you know.
- 2:00:52Yeah, they are just. Just obese humans on these
- 2:00:56robotic arms. With the with the VR headset on
- 2:01:00their head. Yeah, yeah, yeah.
- 2:01:01Sounds like a Black Mirror or Wally episode.
- 2:01:05Yeah, no, I mean, I was going to say Wally.
- 2:01:06It's funny to think about like that Disney movie having like,
- 2:01:10but it kind of nails a version of dystopia.
- 2:01:12Yeah, which I think is similar. I always get 19 Brave New World
- 2:01:18and 19801984 I think is the one where there's a surveillance
- 2:01:23state. And Brave New World is when
- 2:01:25there's like this, this drug called Selma, which like
- 2:01:27everyone just takes this drug and they feel good and pacifies
- 2:01:30them. And so that's like that Wally is
- 2:01:33like a version of the Selma dystopia where like people are
- 2:01:37feeling good but there's kind of no meaning because you don't
- 2:01:40have to. You don't have to work for it.
- 2:01:42It's like a fake. You know, that's part of why
- 2:01:43like people, yeah, when people would consider, you know, the
- 2:01:47feelings you have if you know, you're on drugs to be somewhat,
- 2:01:51somewhat fake because like, you know, you're not feeling amazing
- 2:01:56because you're, you know, having a great date with your partner,
- 2:01:59you're feeling amazing because you kind of like just hacked
- 2:02:01your biochemistry and it it matters like which.
- 2:02:05And there again, like the feeling, conscious feeling might
- 2:02:08be very similar, but one of them we would call fake and it, and
- 2:02:10we actually do care about which, which one is which one is which.
- 2:02:15But but yeah, I, I, I don't, I don't know how that's gonna
- 2:02:21shake out. I don't think, I think, and I
- 2:02:24don't think anybody anybody does.
- 2:02:27I think partly for the reason that you suggest.
- 2:02:30Like it's really hard to predict what those new jobs are gonna
- 2:02:34be. And like, like, and you couldn't
- 2:02:39have done that in the past like you were pointing out.
- 2:02:43And we also can't really do that in the future.
- 2:02:45Like you can look ahead a little bit, but the faster things are
- 2:02:48changing, the like the cloudier it is when you look into the
- 2:02:53future, you know, like even you can, we can say things that'll
- 2:02:58be around in one year, but even. But yeah, as soon as you go out,
- 2:03:00like 5-10 years, there's no Uber drivers and no taxi drivers.
- 2:03:03But those people, we know those people will still be around.
- 2:03:05So they're gonna need to do something.
- 2:03:08But but it starts to get very cloudy, foggy very soon.
- 2:03:13Yeah, and then we're gonna have a robot, each one of us, for
- 2:03:19ourselves. So you take the world's
- 2:03:21population, add a robot for each person, then the robots are
- 2:03:25going around, and then those robots are going to have real
- 2:03:28skin on them because China's already manufacturing realistic
- 2:03:31dolls. They're just going to put them
- 2:03:32on the robots of optimists and all that.
- 2:03:36So we'll be living a very Blade Runner esque world.
- 2:03:39So Blade Runner, they had that right, because you can't tell
- 2:03:42who's who. You can't tell if someone's a
- 2:03:44human or not. We'll get to, I'm pretty certain
- 2:03:46now with what I'm seeing, we're going to get to that and I
- 2:03:49definitely am going to see that. I think it's going to be I'm,
- 2:03:52I'm optimistic about it again when we watch.
- 2:03:54Blade Runner is is kind of dystopia actually, but it it is,
- 2:03:58you can think of the word where we have the robots and we are
- 2:04:01living in prosperity actually. So where where the explosion
- 2:04:05hasn't happened and we are coexisting in the same in the
- 2:04:08same world, yes. Yeah, but yeah, just don't get
- 2:04:12an AI girlfriend. No, no, no, no.
- 2:04:15No, they're they're going to kill you every time.
- 2:04:17Like every time there's a movie with an AI girlfriend they kill
- 2:04:20you. Like I just watched Companion on
- 2:04:22HBO Max. I haven't seen that it.
- 2:04:24Starts out, yeah, it's a new movie with Robot Girlfriend.
- 2:04:28And but it's also like the, well, people.
- 2:04:30The people know that there's kind of this weird thing where
- 2:04:32like the people know that they're robots, but like they
- 2:04:35don't like nobody tells the robots until that they're
- 2:04:38robots. So they don't, they don't know.
- 2:04:41Some of them figure it out. But yeah, they kill, they kill
- 2:04:43everybody. She looks pretty though.
- 2:04:45If that's the robot, she looks pretty, yeah.
- 2:04:49Always well, that's the thing. They're they're always the robot
- 2:04:51girlfriends are always like hot and they they don't complain and
- 2:04:57I don't know you like there's this scene where this guy like
- 2:05:00has sleeps with her and then he like rolls over on the bed and
- 2:05:03he's like goodnight, like go to sleep.
- 2:05:05And then she like powers off. Oh shit, that's like, I guess
- 2:05:11some guys, some guys would would would want some version of that.
- 2:05:13But the problem is they kill you every time.
- 2:05:16Ex. Machina was like that ex.
- 2:05:18Machina was like a. Girlfriend, yeah, but Ex Machina
- 2:05:21was a good film. Was a really good film.
- 2:05:23Was really too early. Too early.
- 2:05:25They should have released it now.
- 2:05:26Nobody understood it when they released it.
- 2:05:28Yeah, I've been meaning to rewatch that.
- 2:05:30That reminds me, I need to do that this weekend.
- 2:05:31But there was also a movie where like Megan Fox, I forget which
- 2:05:34one, what it's called, there's one where Megan Fox was.
- 2:05:37AI know with the Italian guy right?
- 2:05:40Megan Fox. Yeah.
- 2:05:42What was it called? Yeah, but she goes really
- 2:05:46psycho. So anyway, that's just.
- 2:05:49Subservience. Subservience.
- 2:05:51Yeah. I think it's subservience, yeah.
- 2:05:54Yeah, exactly. So every time that goes, that
- 2:05:59goes S but. You know, I think we should do
- 2:06:02something. I think we should plan another
- 2:06:05conversation in a couple of months, depends on how fast AI
- 2:06:09is going to move, where we can come and fact checked everything
- 2:06:13we just talked about today because we made a couple of
- 2:06:15predictions about how things could be in the future.
- 2:06:18And I believe the best way is to have this conversation again and
- 2:06:22then look back at some of the clips from this episode and then
- 2:06:25try to see where we were right and where we were wrong.
- 2:06:29Yeah, except my prediction is that if you get a robot
- 2:06:31girlfriend, you'll be dead. Then then you won't be.
- 2:06:34Too late. Yeah.
- 2:06:35Yeah, we won't. I will be coming as the only
- 2:06:39host now. For Han and I will be like up
- 2:06:42like so that prediction works out.
- 2:06:45You'll have a robot version of me here with all of my memory
- 2:06:49downloaded in. So yeah, you'll still have me it
- 2:06:52just it just won't be me, you know?
- 2:06:55Yeah. But the two word robots is a
- 2:06:59whole, yeah, all kind of worms. Yeah, we're definitely gonna
- 2:07:05have to get into that. There's still a lot of stuff
- 2:07:07that we have to talk about. And like I said, I think we
- 2:07:09should definitely have another conversation just about the
- 2:07:14stuff that we couldn't talk about today, and also to let
- 2:07:17some time pass by to see exactly how things move.
- 2:07:19And I think that will help us in the next conversation figure out
- 2:07:23how much of the stuff that we talked about actually happened,
- 2:07:25how much not. And of course, I'm sure there's
- 2:07:28going to be a lot of surprises. Along the way.
- 2:07:31Along the way. But All in all we wanted to say
- 2:07:34thank you so much for giving us your time.
- 2:07:36This is a lovely discussion. Yeah, yeah.
- 2:07:40No, it's been a pleasure. And let look forward to that, to
- 2:07:44that next discussion. Same here man.