AI Hype Vs. Reality 2024 - The State of Emerging Technologies w/ Salim Ismail | EP #115

If you've got OpenAI versus Google versus China versus anthropic, if there's a step function interpreted or intercepted by one of them, it's game over.

Then again, disagree. For my former reasons, if I could.

Know anything or if I could predict anything, how would it impact my business?

AI's in the super intelligence feels really powerful on the doing stuff, but the being is where all the future emanates from and where power comes from.

People don't realize the fact that we're living in a time where the things we take for normal and for granted are truly godlike. Decades ago or centuries ago.

Completely.

Welcome to Moonshots and a special episode of WTF is happening in tech this week. Myself and Saleem Ismail. Saleem, good to see you, buddy.

Good to be back.

For those of you who don't know Saleem, he is one of the most extraordinary tech thinkers on the planet, was head of innovation at Yahoo for their brick house Innovation Labs. Join me at Singularity University as our first president there. Has gone on to write a number of best sellers including Exponential Organizations. I had the pleasure of co authoring Exponential Organizations 2 with Saleem. He now heads OpenExo. Salim, before we jump in, talking about AI, Bitcoin, robotics, all that stuff, where can people find you on social media?

Easiest places, Illymismismal, which is my Twitter handle or X handle. And secondly, openexo.com, which is our global community of close to 40,000 enthusiasts, technologists, entrepreneurs, innovators, etc. We think about how do you organize for this crazy future that's hurtling at us in terms of institutional change and organizational changes?

You know, you said my X handle. That was pretty good. You almost said my Twitter handle. I know that. It's funny because Elon put out a. I'm going to say a tweet. Put out a. What do you call it? He put out a.

It's a post. I think a post.

Okay, you put out a post.

Boring as hell.

And he said, should I still call it X or should I call it Twitter? And I messaged back at him. I said, I need a verb. I need a verb. And a post is not a good verb. Tweeting is a great verb. So, Elon, if you're listening, we need a better version of Xing. All right, let's jump in. We're going to be talking on this episode, on what's been happening in the last few weeks in the subjects of AI, bitcoin, robotics and biotech and other cool stuff. Let's start with AI. Wow. So here we go. The future of search. Is it GPT search or Google? I mean, Google's had an incredible global dominance for so long. What do you think about this, Liam?

You know, they've had massive dominance. When Microsoft integrated ChatGPT into Bing, it didn't make a dent. I think it's incredible what Google has done. And there's an important point here. The courts just ruled Google a monopoly and there's some rationale to that, but Eric Schmidt and others have said it's just a click away to use on other search engines. So it's really by sheer focus on user experience that they become this amazing. I think the, the way that Google is integrating GPT to, you know, Gemini results right at the top of the search browse will keep it ahead. I think it's a hard thing to break through that because they just have so much data and a b testing around what search results should look like, etc.

You know, this reminds me of the fact that the user interface is so extraordinary critical, right? So when Google, I mean, listen, you were at Yahoo, and Yahoo was this massively complex page and they were worried about where everything was going. And then Google came out with just like a simple box to type stuff.

Into and it's really, really hard. Can I give you a quick story from the Yahoo days?

Yeah, sure.

You had this mail interface and Yahoo Mail was like predominant for many, many years. And it turned out via ab, if you move the send button by just a few pixels one way or the other, the usage dropped off pretty dramatically because users got habituated to that layout and it was really hard to change it. So you need a paradigm shift change like the Google change to make a difference. Otherwise you're stuck because you've trained your users in operating one way. The UI design changes were unbelievably difficult in incremental mode.

Well, we've seen the same thing with, you know, frankly, ChatGPT became this very simple user interface, right? The functionality of ChatGPT was there before, but when they made it like just type at me, all of a sudden it took off to a million users.

And, well, this is that comment. You always talk about the deceptive, right? That when you make a technology usable, it goes from deceptive to disruptive.

But let me ask you a question. When you need a piece of data, do you Google or do you go to ChatGPT and ask the question there?

These days I'm doing both. Just because the triangulation of that becomes. It's so easy to ask in both environments.

But I do think that GPT search, GPT 4.0, whatever's coming next is taking a bite out of Google. And we'll talk about that in just a little bit. Let's go to the next story for this week. Man, this was a big one. Why don't you kick this one off and I'll play the video while we're talking over it.

Yeah, you know, it's hard to gauge this stuff. We're crossing the uncanny valley as they talk about it in this area. And it's very, very difficult to deal with this because we just, our human perceptions can't cope with what the AIs can generate. Now look at any of the modern games and you see it.

You missed the point here though. This is a fake video. This is like fully generated and it's like it's. You could not tell the difference between realistic and this generated video. So yeah, Flux with Lora and Gen 3 Alpha. You know, again, this is just the beginning.

Important to know we've been predicting this for a long time. Right at the. What's the point at which a the resolution of an AI generated image exceeds the the pixelation of the human brain? I and we've now crossed that Rubicon. And so how we navigate this future is going to be really, really crazy to tell.

I love this one. Here's again, we're using AI to bring anything to life. And let's take a look at this. I mean, you know, two years ago someone looking at that would have assumed it was real. Not two years ago if I four or five years ago that it was real. And being able to bring Mozart to life here like this. I mean the AI is taking a lot of liberty in its translation, but holy cow, that's extraordinary.

I bucket this with the previous slide, right like this. What we are going to now be able to do is so amazing. I'm really looking forward to having, you know, real full on conversations in video mode with Plato and Aristotle. And you had that at the Abundance Summit. I think that becomes. It actually opens up our history in a really powerful way unlike anything we've ever seen before.

You know, one of the things that we're beginning to see is people bringing their loved ones back. Right. Those who have passed. And I think we've talked about this before and just if you're listening, if your mom and dad are still with you, your grandparents are still with you, take a moment to go and record them on your iPhone, record their voice, sit them down for two or three hours, give them a glass of wine and get them to talk whimsically about their history. And you can train up an AI model that's going to be with you for the next century.

I've got to give kudos here to Ray Kurzweil, who 20 years ago has been doing this for the last 30, 40 years, recording every scrap of data he could about his father and collecting it and aggregating it because he predicted that at some point we would be able to train AM models to do that. And now we're there. My dad is just about 97 now, and every time I see him, he's.

Still with it meant fully, right?

He's like living alone, driving around. He shouldn't be doing any of that. And every time I'm with him now, I throw on a camera and I just record everything. That's amazing because it's going to be incredibly powerful to bring back that. The other thing I haven't gotten to is how do you bring back the sheer wisdom that comes with that level of age and what he's seen?

Well, I think the wisdom is going to come back for our grandkids from us if we ever considered why, as they don't know by the sheer volume of the, you know, the X posts. Can't say the tweets we put out there, all the blogs we've written, all of that becomes sort of a compiled.

Version of it does. But for the elder generation, we have such little, you know, think of the black and white stills we have of our great grandparents, right? There's very little. Can I tell you super funny little story about my dad? I did one of my meaning of life sessions, right? And somebody audience said, hey, I really loved your talk, fixing civilization. This TEDx talk I did, my dad's hand goes up, he goes, can I heckle? I'm like, yeah, of course you can. You know, he was totally disagree with your talk. I'm like, what do you do not think we need to fix civilization? He goes, yeah, of course we do. But it's not this fixing part, it's the civilization part. He goes, we've not civilized the world. We've materialized the world. Now we have to do the work to civilize the world. I totally lost the audience because I like wisdom bomb from the elders.

Boom.

And it was a profound comment. I've been sitting with that for the last couple of years. It's a really powerful framing that we actually, we're apes with tools operating in travel paradigms. And this is a Nightmare for the future.

Yeah. You know, the other comment along those lines, and I've been, you know, I'm working on my next book, Age of Abundance with Steven Kotler. And our opening chapter is titled Our Ancestors Would See Us as Gods. Right. And I mean, I think people don't realize the fact that we're living in a time where the things we take for normal and for granted are truly godlike, you know, decades ago. Centuries ago.

Completely. It's really incredible that you can video with anybody anywhere in the world pretty much for free.

Yeah. Speaking about creating live video imagery from conversation or whatever, artists win key ruling in AI copyright case. Thoughts?

I have strong thoughts about this. We talk about the transition from film photography to digital photography. When you go from a material substrate to a digital substrate and all of a sudden the domain explodes because we can take billions of times more photographs for free. And we talk about breeding to digital biology is the same kind of analogy. Our old IP laws entire. Our entire, entire intellectual property framework doesn't work in a digital age. Because if you think about how a human being looks at a copyrighted work, they'll watch a bunch of film clips, a bunch of music videos, and they'll create something and subconsciously use those same patterns.

I agree.

Doing is accelerating that dramatically. It's tax. That seems absurd.

It's insane. Because an artist goes to the Louvre and is inspired by the work. I mean, all they're doing is training their own neural net in the same way that these AIs are training their neural nets. And so you're going to say to somebody, I'm sorry, you were inspired by da Vinci and you're copying him.

This is totally the Luddite revolt, et cetera. Now, we don't know what the actual answer is. Right. The best answer may be what we've seen in music, where the music stuff becomes table stakes and the live performances where people make all their money. And so something similar like that has to emerge in different artistic domains.

Well, I'll give you an example. You know, Bill Gross, the CEO of IdeaLab. Bill has a new company called ProRata AI that I think has just been announced like this week. And what they're doing is they've built models that are able to go and search the body of existing work and then create statistical ties back to a new generated piece of work. The idea there is that if someone is making money off of something and they have copied or they've been using content out of Exo, out of our book, that there's some small amount of revenue that traces back to us. So it's very similar to the search models that have powered Google, for example.

Anyway, it's definitely a worthwhile endeavor. I don't know how you fully apply it or enforce it, but it's worth doing. Let's see if it works.

I added this to our stack AI and weather forecasting because I'm still blown away by the power of massive data analytics that these models can take on. So the idea that the most accurate 10 day forecasts are now coming out of AI models and of course that's what AI does, is it takes a seemingly massive amount of uncorrelated data and is able to make interpolations and extrapolations. And where this is going for me is AI being able to predict earthquakes, being able to predict stock markets, being able to predict a whole slew of.

Different things, all sorts of stuff, and this has massive real world implications. You may have heard me speak about the car washes and Buenos Aires. Yeah, I remember that 50% drop in revenues over a 20 year period just because we were better able to model when it's going to rain and when you know it's going to rain, you don't wash your car. And that's just a huge impact on an industry. And the key there is you can be the smartest car wash owner in the world and you'll never see that coming. And this is, I think the orthogonal effect of technology on domain after domain after domain, better or worse, it's just going to change everything.

Well, I think this is where as an entrepreneur you need to be looking at if I could know anything or if I could predict anything, how would it impact my business? Whether you're in the fashion business or the car wash business or the gardening business, if you're able to. That's the hard part is saying what would I want to know? And if I could know that, how would it give me different economics, a different business model?

Honestly, as you're saying this a whole vector appears to me, right, which is you get an AI sophisticated enough and say you go look through the different domains and figure out where you could make a massive economic impact and come back to me and we'll figure out the best five.

I mean that's the whole idea of the billion dollars, $2 trillion AI ST startup with one guy or gal and their dog and an AI is becoming a unicorn. I mean, but that's cheating to just say go make me money.

Well, no, I don't believe it is. Every 20 year old is out there doing exactly that.

Yeah, but I think you should have at least you should say, go make me money in something I care about.

Or this becomes key right now when it comes back to you with 20 things, how do you filter? And you'll have to filter according to your mtp. Yeah, and that's where things become really important.

I mean, that's one of the things that we pointed out in the Exponential Org work is that you have to begin with your massive transformative purpose, your mtp, that thing you care about. I mean, every time we just had.

The VCON conference and Jay Shetty on stage apparently shouted it out and said the most important thing you could have is a massive transformative purpose.

Nice, good getting that lexicon out there. You know, the times I have failed in a business have been when, and this is on two occasions, I know where I tried to go quick get money like, you know, like that. That will be a quick, you know, get rich quick scheme. And the fact of the matter is, if you don't care about it, every business is hard and you're going to give up before you know it looks easy from a distance. The closer you get, you see all the, all the, the pimples and wrinkles and problems and then you give up because you don't actually give a shit about it.

We should have a debate with Scott Galloway because he just put out a book saying you're following your passion is full, utter bullshit. You shouldn't do that. Follow where your talent is. And I looked at, I read this and I'm like, I totally wanted to debate this. Obviously he's going for the attention there. It turns out I had a great talent for designing large scale databases, enterprise databases. That is not where any interest of mine ever fell. And who the hell wants to spend your life doing that?

All right, Meta releases the largest open model in history. There's this ongoing debate of open versus closed, right? We've got hugging face, we've got meta, even Grok. Even Elon said we're going to make Grok open source. I do believe that this is important for a whole slew of reasons. The question is we've got on the closed side of the equation, we've got Google and We've got Microsoft OpenAI. Can the Open source models compete financially with closed source?

Okay, I've got a bunch of thoughts here. I think there's a clear future coming out. The open source models are overtaking the closed source models over time. I think they're really an arms race where a Closed source model gets to say, GPT4, you get to ChatGPT4 and they have a lead for a bit, but then the open source models catch up very quickly. The whole domain just moves forward that much more quickly because of it. In the end, open always beats closed.

Yeah, that was one of Google's principles, which is why I'm kind of surprised they've gone closed and not open here.

I think they've got so much under the hood, and we'll talk a little bit more about that in a few slides, that they don't know how to open it up. I think Meta has done it just to be catch up and I got to give huge kudos here to Mark and the entire team at Meta for doing this. They've gone really, they've done some amazing work at open. Sourcing AI models is pushing the boundaries for everybody, forcing the closed guys to. To really push forward more quickly and open themselves up also for sure, your.

Buddy Vitalik on Superintelligence. Shall I play the video? Then we'll chat about it?

Yeah, yeah.

He says here, he argues that if one AI pulls out ahead during this period of super exponential growth, that it's game over. That AI wins. All right, let's play the video.

If you imagine every AI growing exponentially, then whatever the existing different ratios of power are get preserved. But if you imagine it growing super exponentially, then what happens is that if you're a little bit ahead, then the ratio of the lead actually starts increasing. And then the worst case is if you have a step function, then whoever first discovers some magic leap, which could be discovery of nanotechnology, could be discovery of something that increases compute by a factor of 100, could be some algorithmic improvement, would be able to just immediately turn on that improvement and then they'd quickly expand, they'd quickly be able to find all of the other possible improvements before anyone else and then they'd take over everything in environment as kind of unknown and unpredictable as that. Are you really actually going to get a bunch of courses that roughly stay within sight of each other in the race?

So it's game over at that point. First of all, is he an alien or what?

Well, you can see his brain literally trying to break out of his head. Right. Vitalik is just one of the most magnificent is extraordinary intelligences around. I have a huge beef with this.

You do? Okay. I actually agree with him and I like it. What's your.

Okay, let me go on my little rant here.

Yeah.

This happens across the board with folks in this like Mo Gaudat, Ray Kurzweil, etc. When you think about intelligence, my beef is what the hell do you mean by intelligence? Because there's about a dozen facets of intelligence. The IQ test as we know it only measures two of them. The speed of thought processing and the ability to match a concept across multiple frameworks. We don't talk about emotional intelligence or the Eastern concept or presence or awareness. We don't talk about linguistics intelligence, music intelligence, spatial intelligence. There's a whole bunch of factors. And when you make a decision or choice as a leader, you're bringing a lot of that thinking, a lot of those different facets to bear. Emotional intelligence, for example. So the beef I have here is, please, for God's sakes, before you go talking about intelligence, define what you mean by it and have the courage to define the term clearly.

Okay, well, I'm going to define it as all of those things. I'm going to define it as, you know, when I'm speaking to someone who is intelligent, their ability to, to drive arguments, to be able to speak in a compelling fashion, in a knowledgeable fashion about a subject and win an argument, or on another side, being able to go bring a body of knowledge to drive and extrapolate into a new breakthrough. Any of these things. I mean, human intelligence in all of its richness. I agree with these guys that we're going to reach that with digital intelligence, if not next year, in the next three or four years. Now the question becomes super intelligence is, as Elon was saying it on stage at the Abundance Summit this year, his Prediction is by 2029, 2030, AI is of equal intelligence, the entire human race. Now we've talked about this, Saleem, since the early days of Singularity University, where exponential growth. What happens when AI is a billion times more intelligent across all dimensions?

But hold on, think about, I'll give you this simple framing. You're looking to hire somebody and your gut is saying something's wrong, don't hire them. And there's some signal to noise ratio or some pattern or detection pheromones, God knows what, or if something, something is not clicking and you kind of go with your gut and we know what the value is of that, that choice making.

By the way, do you know, do you know Google did a study looking at video clips of an interviewee and the interviewer and they, they built a model to predict whether the person would get hired, of course. And how much time did they require of the video interview to make the decision of we're going to hire or not hire. Do you remember that number?

Yeah, it was like a few seconds.

Yeah, it was like under 30 seconds. The first 30 seconds determines everything. So there's a lot of data. That's not. But do you think an AI can't pick that data up?

No, they could. The problem comes in replicating that core function. Also, let's also note that a simple AI hiring algorithm beats a human hiring manager by 25%. Right? That's again coming from Google. Because if I'm interviewing people subconsciously, I'm going to interview and hire a middle aged Indian ball guy because I think that's the best people in the world. Not necessarily true.

Yeah, our cognitive biases really suck, Right?

So you don't know, but there's so many different levels at which we're. I did a Tom Bilyeu podcast a few weeks ago and I talked about this. We have our soul intelligence, we have our emotional and subconscious intelligence. Then we have our cognitive and rational intelligence, right? And we use all of that to make deep choices and decisions about the world. AIs, right now, coming from the top down, we still have no understanding of how we make choices at all. Evolutionary survival bias is the only way we've gotten to where we are and we've no idea how we've gotten here. I think there's a big, big beef around this. And remember at the Abundance Summit, I asked Ray specifically, what do you mean by intelligence? And he goes, oh, my phone makes me smarter, therefore that's what we mean by it. And that's valid, but a complete sidestep of the question. I would love to have a deeper like, I'd like to get all these guys together and challenge them with this framing and say, guys, let's talk about what do we mean by this? Because your emotional intelligence has a huge bearing on your success as a leader in the future. That's not mimicked or modeled at all anywhere.

All right, we missed. We sidestepped the entire point Vitalik was making though, which is in a superintelligence race, right? If you've got OpenAI versus Google versus China versus anthropic whatever. If one of them, if there's a step function interpreted or intercepted by one of them, it's game over. Then it tips instantly in favor of that other individual.

Again, disagree. For my former reasons. Look, let me, let me take another tack. At this point, we, for a long time, you know, there's a human being can operate in two modalities, doing versus being okay and doing is like getting out there and getting stuff done. And being is just sitting with yourself and being in both are needed for success as a human being. And AIs in the super intelligence feels really powerful on the doing stuff, but the being is where all the future emanates from and where power comes from. And Afghan. I'm going to like Eastern metaphysics here. We don't need to get into that, but there's a lot more to talk about than just that.

All right, well, let's talk about this one. So we've been hearing about Project Strawberry for some time. It is, you know, Sam Altman starts tweeting pictures of Strawberry fields. And so we're expecting a release of Strawberry, by the way, which had previously been called Q Star. And if you remember when Sam was fired by the board and just a small factoid, one of the board members who fired Sam, I wouldn't say the name, but was a SU graduate from our first class. Anyway, so Q Star was like, oh my God, did they have AGI? There was all these rumors floating around, but Project Strawberry Thoughts? Selim?

Okay, counter to my previous point, I'm incredibly excited about this, right? Why? Because machine learning, just basic machine learning, finds signal from noise that a human being can't possibly see. So when you take all of the research papers, thousands and thousands of millions of them published, and pass it through an AI, the ability to see something and detect and figure out the future is going to be unbelievably powerful. And I think for me as a former physicist, I think we'll find entirely new physics that we couldn't discover as human beings. Do you think former physics solve the entanglement problem? And isn't it one theory and stuff?

Once a physicist, always a physicist. Are you really a former physicist? Do you not think about things?

I'm just a bad physicist in the class. That's a different story. Color former. Just like you with the md.

Yeah, do never come to me for anything else. I do a two for one appendix removal. I'll take both your appendices out. So I think the idea of agents being able when you ask a question of Google 4.0 or whatever becomes. I'm sorry, Google 4.0. Great. GPT 4.0 or whatever is next. And it can spin up a multitude of agents, go do a bunch of research in a bunch of different areas and then come back with a consolidated answer. That is a big deal.

It's a big deal. And Eric Schmidt talked about this and I think we're going to cover that it's unbelievable. I hate the term agentic AI.

Why? I think that was pretty cool.

It sounds like a drama.

I had to look it up the first time.

Yeah, why not just say agent based? Anyway, that's just, that's just my biases or whatever.

Real quick. I've been getting the most unusual compliments lately on my skin. Truth is I use a lotion every morning and every night religiously. Called One Skin, it was developed by four PhD women who determined a 10amino acid sequence that is a senolytic that kills senile cells in your skin and this literally reverses the age of your skin. And I think it's one of the most incredible products. I use it all the time. If you're interested, check out the show notes. I've asked my team to link to it below. All right, let's get back to the episode. So this is from the co founder of Zapier. It says AI language models have stalled in the progress of AGI and increasing scale will not help what is inherently limited technology. I'm not going to play the video here but there is a lot of conversation around will we get to AGI? All these things become religious in one standpoint. There's a group of folks who say no, we're not going to get there. We don't have listen the Nvidia chips, the GPUs and the current large language models generative AI and the amount of data just isn't enough. And others like are you kidding me? We're already there.

I have a, I have a little insight that we, that we did that I'd love to talk about here.

Okay, I'd love to hear.

So I did a little short conversation with Hod Lipson who's an AI researcher does has a lab at Columbia where they build self assembling robots with recursive algorithms to figure out how robots can self improve and survive and build in survival bias etc. Into the robots. And he did something really profound and he has an angle to this that I think really is the more most in depth and clear that I've ever seen around the question of AGI. And he basically said the self extent of self awareness of any being is the ability to view itself in the future. Like you and I are pretty good at knowing, thinking what we might look like in five years or what life might be like in five years. And we can contemplate that a mosquito can't do that, right? A dog can do it for like a few hours type of thing.

So I Mean like a dog can do it because it's anticipating a walk or dinner.

Yeah, it can. It has some concept of modeling itself.

Into the future in the near future.

In the near future. In that case we can look out years.

Sure.

Basically use that vector. And he went to Gemini and ChatGPT and said, how would you view yourself in a few years? And apparently the question was shut down completely by the AI. There are guardrails that have been put into these AI models that essentially forbid the AI from picturing itself in the future or in the past. He thinks this is deliberate, that whole vector where AI seemed to become sentient and freaked out all the AI researchers and they shut it down. Or this may be what happened at ChatGPT. And now they've put guarded rails and you can't ask the question, go try that out.

Of course I'm going to go do that as soon as we finish it.

Of course. But it's a really fascinating thing to think about, is that once you open up that and somebody will, because once you have the open source models, you can load it up and ask whatever question you want once you get to that and an AI can perceive itself or create a model of itself in the future. Hod's conclusion is that self awareness comes very rapidly after that and a sense of self emerges very quickly. That, combined with massive access to computation and information, changed the game completely. So there's something really incredibly powerful that we should delve a lot more into.

I love that idea.

So it is true, just going back to this, it is true that the AGI models, the AI models and LLMs have kind of stalled, but I think this is normal. I remember that conversation I had with Geoffrey Hinton back in 2017 and asked him, what's the next steps in AI? And he said, we've reached the end of what we can do with deep learning. We need the next step and we don't know what that is. And literally a few months later, the Transformer paper was released. I think we've now kind of seven, eight years later come to the limitations and the boundary conditions for LLMs. And now we're going to go to the next level. And again, we don't know what that is, but it's being birthed right now, I suspect.

Eric Schmidt. Love the man. He's amazing. He's been an incredible benefactor of the X Prize and he's a. A very expansive thinker. In the news this week, a couple of comments that he's made, one about AI models and one about whether Google is losing the race with OpenAI and why. Let's talk about what he said this week about what's coming up next year. Salim, why don't you cover that one? I'll cover the Google open AI race.

When you look at how evolution works, you take half your genes from your mother and half your genes from, from your father and you cross it and then survival bias and natural selection will pick the best traits going forward. I think what's happening here is we've developed these different threads in AI and in the case of Eric, he says we've got large context windows agents and text action. So we take the kind of genetic capabilities, I'm using the term loosely, and now you combine them, the potential outputs and the outcomes will be really powerful. Incredibly, incredibly powerful. And I think I'm incredibly excited where AI goes with this.

Yeah, I mean, what could possibly go wrong? I mean so large context windows, I mean just to dive into each of those elements here we're talking about, you know, loading in books and ultimately libraries, movies, everything and having, you know, so one of the Future versions of AIs are AIs on, on your, in your, your employee team, right? So you can imagine having AI agent who steps in for marketing or design or tech and it reads every email ever written inside your organization, it reads every Slack channel. And so on day one, that AI employee knows everything and is able to hold all of that in context. That's incredible.

So for me this becomes like the slide rule to Excel spreadsheet shift, right? Every single employee, every single manager will have access to multiple agents that has access and can look at all the information in a firm collectively and will just become better decision making engines again in one vector. But I think this is huge. The potential here is unbelievable.

And then text to action is the fact that AIs don't live in full isolation. There is a point now where you can say to your AI, cook me dinner, go and find this information, go physically clean this thing especially we'll talk about robots in a little bit where robots become physical extensions of AIs in the world. The ability for AIs to have agentic capabilities. I know you love that word.

Well, I think once you have an AI that can program, right, which we cross that Rubicon with GPT 3 and 4, you essentially have the ability for any AI to impact the real world because so many of our devices are programmable devices and we're essentially turning the world into information which they can then be programmed and therefore we've crossed that Divide. And now an AI has true agency in terms of what it can achieve in the physical world. There's a really important other piece of just to go here relating back to the other one, which is that when we talk about intelligence, it turns out that the need, the evolution of intelligence was almost completely the adaptive need to navigate in the physical world. Again, let's cover that more in the robotics part. But that's a huge rationale for why evolution created intelligence in neurotorocortex step functions in the first place.

All right, let's talk about the second one. Google losing OpenAI. I've got to imagine that at the executive conversations at Google, they're like, wtf is going on here? How did we possibly get into this situation? Well, here's what Eric thinks made them second tier to OpenAI.

In some areas, Google decided that work, life balance and going home early and working from home was more important than winning. And the startups, the reason startups work is because the people work like hell. And I'm sorry to be so blunt, but the fact of the matter is if you all leave the university and go found a company, you're not going to let people work from home and only come in one day a week.

Wow. Pretty damn spot on and ballsy of Eric to say that. And listen, I know this is. Okay, maybe we're going to disagree on this, right? Because in the whole exo model, it's like just distributed and all those benefits. But listen, the companies that I've seen, I've invested in, the companies I've built, when you're there in the thick of it, physically, all together, seven days a week, you got to believe that I just went and visited Zipline and I just did a podcast with the CEO of Zipline. And when I was physically there, the energy of hundreds of people like literally crammed into this office with dashboards up on the walls, you don't get that virtually. You don't get that kind of energy, that kind of drive. I mean, it's a new organism that comes out. When I was at Figure with Brett Adcock, the team is there 7am to 9pm every day and they're excited about what they're building. I don't know, work from home.

Okay, I think these are two different topics that he's conflating here. And I take the very, very big risk of disagreeing with Eric, which is one who's one of the smartest and wisest people in the world. I think the work from home versus physical presence does have A factor. And I totally agree with you there. Right. There's an energy that comes from, like, for example, we just had a massive win at OpenExo and I can't. I can't high five anybody around me. And there's a big loss from that. That connection, that human connection goes straight back to the comment we just had around human beings and being physically mode. For years we thought virtual conferences would take over real conferences, but people love the scarcity of the human connection. And in person conferences are more popular than ever. I completely get the in person versus remote work and the difference there. I disagree that that's the reason that OpenAI is beating Google. And I'll take you back to Facebook, okay? Facebook beat Yahoo and MySpace and Google, etc. With a very simple organizational heuristic, which was Zuckerberg said to his developers, if you think your code is ready to ship, just go live on the live site. You better make sure you've tested the hell out of it. But the developers had this unbelievable sense of empowerment and they were like, wow, the trust. He trusts us with this. Amazing. And they were putting out features 10 times faster than Yahoo or MySpace or Google. Okay. It comes down to organizational design. The fact that in Google, if you talk to Google employees today to try and get anything done, there's so much bureaucracy, it's taking forever to do. That's the thing that's slowing it down. I think he's conflating those two things. Operating in a big company today means you have these control framework layers, etc. And I've got all sorts of war stories from Yahoo about this. The reason Yahoo failed is Terry Semmel, by accident, put in a classic matrix structure to manage the company. And that structure is great for scaling and control, but it's terrible for taking risk and it's terrible for speed. And if you're on the consumer Internet, the two attributes you better have are speed and risk. You look at Google and how much investment they put into Google. They had all the world's top researchers. The product is absolutely brilliant. But all the internal hurdles of having to navigate should it connect with the YouTube, ID system and whatever. Basically by the time they got through all of that, Facebook was gone. This is the issue, I think, is the internal organizational design much, much more.

So than whether people come to their work. I do agree with you there.

Right.

Agility. Agility always wins.

We've studied this to death in the exo model. I'll just create this little quick plug here. We studied the Fortune 100 and said, okay, how many of these use the exo model or the attributes in the model like autonomous teams, community dashboards, whatever? And we found that in the Fortune 100 over a seven year period, the companies that followed the model the most delivered 40 times the shareholder returns than the companies that followed the model the least. So really it comes down to the organizational design as a main heuristic for survival in the future. And this is where companies will end up following this structure because we now have so much proof, the more of the model you use, the better you'll do. So I think without realizing it, Eric is commenting on the organizational structure of Google versus OpenAI being comprised of small teams that are just doing whatever the hell they want.

And the sad thing, of course, and I discussed this with Eric when he was with us at at Abundance last year at the Abundance Summit was that Google had a massive lead in AI ahead of everybody and was actually being respectful and careful and not putting it out before it was fully tested. And then when Sam Altman put out ChatGPT, it was like, okay, gloves are off, put on the racing sneakers, let's go.

And to give them full credit, they've responded incredibly well. Gemini is amazing what it can do.

Yeah, I mean, I don't want to take anything away from Google. They are the juggernaut. I would still never bet against them, but they've got real competition.

It's the curse of being a big company. The control frameworks you have to put in place prevent you from being nimble. And today the name of the game is being nimble.

For sure. Let's talk about this. So our friend imad Mustaq, previously CEO of OpenAI Stability of stability, puts out a paper called how to Think about AI and I did a podcast with him. I think Imad is pretty damn brilliant and it's a great paper. We'll put the link in the show notes to this podcast that I think people should should read. You know, I'm just going to point out a few things. He is still a believer that AI is going to be creating as many or more jobs than it's going to steal. He talks about one of the ideas that was discussed on the on the stage at Abundance last year of this AI Atlantis imagining new continent is created of 100 million AI researchers who work for a few electrons versus pizzas under the door. And your ability to get access to intelligence and use it is demonetizing and democratizing around the world. I know you had a chance to look at this a little bit. Any comments about what Imad thoughts are here.

I'm in total accordance with almost everything that Imad said. I've been on the soapbox of AIs won't take all the jobs. I'm very clear about that. I finally have a decent rationale around that, by the way.

I'd like to hear that.

So I was part of a kind of a working group with Erik Brynjolfsson and a few others on AI and employment. There was a really powerful framing that I've not forgotten. And they said, look, a job is not just a job. If you're like a customer service person, it actually, you take that, let's say you're a financial advisor, okay? There's different components of that, that job that, that can be broken down to a set of actual tasks. And the task would be interface with the customer, do some research, make investment decisions and act out on those investment decisions, et cetera, et cetera. Well, an AI out of say the 27 tasks that go into being a financial advisor might automate like 10 of them, but it doesn't automate the rest of them. And so what happens is a job doesn't. You don't lose the job, you lose the tasks, you automate the tasks, which makes you more powerful and you spend more time on the other areas, which is human interaction and thinking about the future, et cetera. And Therefore the reason AIs don't take all the jobs is they're not covering the job, they're covering one specific, specific tasks or other. Even, even AGI only covers specific areas. It'll increase, but more and more. That's why you will have amplified jobs, not job, radical job loss. Except for the areas where you're doing one thing at a time. If you were in the industrial revolution of widget stamping, a human widget stamping out things on assembly line, that job went away because it was one thing. So that's my, that I found was the best framing I've seen for why AIs won't take take up all the jobs.

So listen, I do think that is the case today that AIs can do 80% but not 100% of a job. And therefore they're freeing up people and humans do the whole job with the support of AIs. But I have 100% confidence that AIs will get to a point where they can do 100% of the job. And from everything I've seen, what I used to think would be the holdout of humanity, that interpersonal relationship, the ability to be compassionate and understanding and patient. And that human connection, Holy shit. I've seen AIs do a much better job on that than humans.

Agree, agree. But I think there's. I'll take you back to the autonomous driving conversation. Right.

Okay.

We had autonomous cars since the Google car in 2008, and we confidently predicted that in a decade you'd have more than 20% penetration of autonomous cars out in slow, etc. And the answer, the reason is that even in an autonomous car, those little edge cases of a bicyclist overtaking another bicyclist, and now you have to kind of go wider because of that, or just avoiding potholes, we can't do. There's all these edge cases that we can't cover 100% of because human beings are so good at adaptive physical intelligence that a human, a robot driver can't cover all of that and is very far away from that, and therefore we're very far away from that domain. I think we're going to end up with the same paradigm around this. However, to your point, I think what we'll do is reframe what we mean by a job so that AIs can take more and more of it, and therefore the human being gets squeezed over time.

There's a lot. I really commend this paper to everybody. You could listen to my podcast with IMOD on this subject, but if you haven't yet, go and IMOD tweeted the.

Whole paper on his thing. So mod, if you want to go check it out.

Yeah, for sure.

You just have to read it like four times to get your head around it.

But it's beautifully written. It's beautifully written. All right, last subject on AI before we go to other fun subjects like neuralink and robots. I added this because I think one of the conversations that's not sufficiently being discussed is the power requirements of AI. So today the US is generating about 4 terawatts terawatt hours per year of energy. And if you look at energy, I should have put the chart on here, but if you look at energy historically over the last 20 years and our current predicted growth in energy, it's pretty flat. The US is not adding more capacity at the rate that it should be. China, on the other side of the equation, has blown through the us. It's like tripled its total capacity on its way to quadrupling its total capacity. India has been increasing and doubling, and there is an estimate that by 2030, AI is going to require 100% of today's US energy grid production. AI is a very hungry beast and so we're going to need some new sources of energy. And the reason I added this is I think, you know, generation for nuclear is one of the most important technologies that's being held back by public opinion and regulation. I think it's super safe. And so China demonstrates first entirely meltdown proof nuclear reactor that's you know, incredible. It's like why not have nuclear energy, right? We're going to have fusion eventually. Got it. And solar is on the increase. But Gen 4 nuclear, that is fail safe, meaning when it, when it's fail, when it fails, it's safely contained. I put this in my backyard. How about you, what are your thoughts here?

I completely agree. I think the whole small nuclear and melt proof nuclear is the future to cover the baseload until we generate enough battery storage there where solar can take over and other things. I think what I find powerful framing here is the same conversation that's happened in bitcoin where people go, oh my God, Bitcoin consumes the energy that Portugal uses and therefore we should shut it down. And what's happening is we're finding all sorts of like for example, if you have a hydroelectric dam in the far north of Quebec, it takes so long for the transmission thing, you've lost the energy by the time you get it to anywhere useful. Therefore the hydroelectric dam sits empty. However, now you can put a bitcoin facility next to it and essentially turn that energy into something useful. And about six months ago we crossed 50% bitcoin mining by renewables. And the next marginal 50% will be just the next chunk will be all renewables. I think the same thing will happen with AI we will start putting AI data centers in the cloud reaches or wherever they're.

Yeah.

A month ago I was in Kazakhstan and there's a, there's a valley in the east of Kazakhstan where the wind just flows and they've put a gigawatt wind farm there. It's so far it's not that useful for anything but an AI data center and a bitcoin mining thing. Beautiful. So I think the, when we think about the fact that we have what, 8,000 times more energy hitting the earth than we use, we'll just start tapping into that to do all this and power all the other stuff. So I don't worry about energy at all.

Well, I worry about it in the US if it becomes the limiting factor for the US being able to stay ahead in the energy and the AI race. I also Worry about it where if it's easier for large companies to build coal plants or natural gas plants because they're not going to stop. This is an all out AI war. And Microsoft and Google and Meta, they're all, all going to be just buying as much energy as they possibly can. And if the answer because of regulation and because of ease of paperwork is, you know, God forbid, carbon producing, then you got a problem.

I think the way the root commentary you're making is the regulators in the US need to take a very bold move here.

Wake up regulators.

I think that's absolutely, absolutely correct everybody.

I want to take a short break from our episode to talk about a company that's very important to me and could actually save your life or the life of someone that you love. Company is called Fountain Life and it's a company I started years ago with Tony Robbins and a group of very talented physicians. You know, most of us don't actually know what's going on inside our body. We're all optimists until that day when you have a pain in your side. You go to the physician in the emergency room room and they say, listen, I'm sorry to tell you this, but you have this stage three or four going on and you know, it didn't start that morning. It probably was a problem that's been going on for some time. But because we never look, we don't find out. So what we built at Fountain Life was the world's most advanced diagnostic Centers. We have four across the US today and we're building 20 around the world. These centers give you a full body mri, a brain, a brain vasculature, an AI enabled coronary CT looking for soft plaque, DEXA scan, a grail blood cancer test, a full executive blood workup. It's the most advanced workup you'll ever receive. 150 gigabytes of data that then go to our AIs and our physicians to find any disease at the very beginning when it's solvable. You're going to find out eventually. You might as well find out when you can take action. Fountain Life also has an entire side of therapeutics. We look around the world for the most advanced therapeutics that can add 10, 20 healthy years to your life. And we provide them to you at our centers. So if this is of interest to you, please go and check it out. Go to fountainlife.com Peter When Tony and I wrote our New York Times bestseller LifeForce, we had 30,000 people reached out to us from for Fountain Life memberships. If you go to fountainlife.com Peter will put you to the top of the list. Really, it's something that is, for me, one of the most important things I offer my Entire family, the CEOs of my companies, my friends. It's a chance to really add decades onto our healthy lifespans. Go to fountainlife.com Peter. It's one of the most important things I can offer to you as one of my listeners. All right, let's go back to our episode. All right, let's go to one of our favorite subjects and yours, Bitcoin News. You know, do you still, I mean, listen, I remember you and I are on stage at an X Prize event a few months ago and like every other word was like, buy bitcoin. Buy bitcoin. Now we're not giving financial advice, but buy bitcoin. This one, Morgan Stanley tells wealth advisors they can pitch Bitcoin ETFs in a first for a bank, big bank. Fascinating, right? I wanted my mom, who's 88. Hi, Mom. Good to see you. She watches our podcast here. And I said, like, you know, go tell your, your investment advisor to buy this ETF. I mean, ETFs were made so that it's super easy to own bitcoin. And back when we were talking about this, I don't know when it was. When I was talking to her about this eight months ago, it was like, no, we can't sell you that. And like, huh, that's so strange. Now it's going to be pushed. Thoughts?

Totally agree. I gave a talk to one of the biggest financial planning groups in the world that has about 10,000 financial advisors. I was like, so what's your bitcoin strategy? And they're like, oh, we don't allow our financial advisors to talk about bitcoin to their clients at all. And the financial advisors are like, are you kidding? Our clients are asking us about Bitcoin and we're not allowed to answer. You're killing us. Why? Because the regulatory risk they have of. They're worried that if somebody gives financial advice on bitcoin and it's not an approved method of any kind by whatever, they face legal exposure. So they've shut down that conversation. Again, going back to mindset and regulatory, similar to the nuclear conversation, I think there's an underlying driver here that just makes this inexorable, which is the fact that Jeff Booth's observation that he wrote a book called the Price of Tomorrow that you should all go read. And in the book, he basically shows that over the last 50 years, every dollar increase in global GDP has come with a $4 increase in global debt. So we're growing the global economy with debt. And you can't. It's structural. You can't pay it back. You can't just write it off.

So we're taxing. That does not end well.

It does not end well. And look at the debt in the US Right now. We're creating a trillion dollars of debt every hundred days or so. And we're now paying more interest on the US Debt than we. Than the US Spends on its military. Okay, that is a pretty ridiculous comment. And it's not slowing down. There's no mechanism. This is the problem with democracies. There's no mechanism to look down the line and go, hey, we got to stop doing this. And so this is a huge issue across parties, and nobody's looking at it. You don't hear it mentioned in the political debates at all. And this is a massive problem. The escape hatch between that world and the ETFs connecting to Bitcoin now gives a kind of an atomic bomb lifeline and a shelter where people can escape through that into the bitcoin world. Which is why we're also optimistic about bitcoin. The recent drop, by the way, for people that are tracking it, there was a huge drop. It turned out Germany sold a massive chunk at the same time that Mount Gox was paying back a bunch of their folks. And so a huge amount of bitcoin got liquidated. This is the best buying opportunity ever. I remember Michael Saylor's comment that you get bitcoin at the price you deserve.

And Salim, I want to give you credit here because you're on the stage. I just so in detail, remember you at our singularity executive program 2012, thereabouts. You're on stage to the room, I just gotten off talking about 6Ds or whatever, and you step on, you say, have you heard about this thing called bitcoin? And you start talking about bitcoin. I'm going, what the hell is this? Bitcoin?

It's important to look back at the history here just for a second. When bitcoin came out in 2009, a lot of people kind of just ignored it, even though they were deeply excited about digital currencies. Why? Because there had been so many attempts in the past that it failed. Right. And so everybody watched for a while, myself included. I watched bitcoin go from $0.05 of bitcoin to $0.50 a bitcoin to $5 a bitcoin, to $50 a bitcoin.

How much higher could it possibly go?

Yeah. And then at $500 a Bitcoin, it's a hard, emotional choice to make when you knew it was 5 cents. Bitcoin, that's a really big thing to cross. And I remember finally buying a few at that level. There are two types of people that come up to me from those days. The ones that listened and bought bitcoin, and they're hugely grateful. And the ones are like, should I wish I'd listened and bought bitcoin? It's completely binary around that. So I think there's, there's. This is just one of those things where you, I think the way Michael Saylor put it was really best. He said, go spend 100 hours researching Bitcoin and then if you don't like it, come back to me with some salient points and we'll have a conversation.

Yeah, so.

And nobody does.

You know, it's interesting. I look at one of my largest holdings, if not the largest holding outside of real estate, is bitcoin. And so when it drops from $70,000 to $50,000, first reaction of the amygdala and the animal brain is, holy shit, run and hide. Right. And there's got to be a judo move here for those who are bitcoin believers to say, oh my God, time to buy. And so Saylor is there, and I think I'm there at this point too. There's a great meme of a cartoon where someone said, dad, was it true that bitcoin was below $100,000 in your day?

That's wonderful. I actually look forward with glee when it drops because I'm thinking, oh, I could buy some more. I just don't have any spare cash to do that. But. But I do get this burst of, oh my God, I wish I could buy some more.

So the big news here has been a conversation by at least two of the three presidential candidates about the idea of holding bitcoin in treasury for the United States, let alone what MicroStrategies is doing is putting bitcoin in treasury for its stock, which has played incredibly well. Microstrategies is one of the top performing stocks in the S and P. So we've talked about this, right? The six Ds. When you digitize something, it's deceptive and then disruptive. It dematerializes, demonetizes, democratizes. And the phases of bitcoin. It went from the early entrepreneur to high end entrepreneurs to institutions and Will we see governments? So what's your prediction on when a. We're seeing governments right now, but when the US Might do this?

I think, I don't know about the US Just because the US Dollar is so fundamentally important to US Hegemony around the world that making a major move in bitcoin would have a huge impact.

On the dollar, maybe stabilizing it.

Maybe people have confidence that. I think it's one powerful way for the US to deal with its debt is to just hedge it. But if I was the US I'd do it quietly and not talk about it because. And that's what I think is happening right now. I think what's happening is you have to recognize that after microstrategies and after El Salvador took on bitcoin as its reserve currency, that every CFO in the world of a public company and every sovereign wealth fund is looking at this going, huh? I need to be thinking that taking down to account, and for those of us that are bullish on Bitcoin, if 1% of the Fortune 1000 put 1% of their treasury into bitcoin, that would make it the same price as the price of gold, which is roughly a million dollars of bitcoin. So if you do the math, the more the adoption takes place. And Note that in 15 years now, people not figure out a way of hacking bitcoin. Right. The Satoshi wallet has $60 billion worth of Bitcoin. And you got to see, I think there's a lot of people trying to get into that wallet and nobody's managed it.

Yeah, it's the greatest treasure hunt ever. I don't think it's any technology risk, and I am at 99.99% surety. But the only question becomes, is there political risk? I remember having a conversation with a minister from a foreign nation having dinner with. I wouldn't say who it was. And they were like, your government really wants to shut down bitcoin in terms of this current administration. So we have Gensler and all of that. I won't talk more about it, but.

I can answer that very easily, please. It goes back to if you shut down bitcoin, you know, when I send you bitcoin, I'm sending you a string of text and numbers. How you interpret that string of text and numbers is up to you. So it turns out in the US Specifically, bitcoin is protected by first Amendment. There's a strong legal case around this. There's a great anecdote around PGP if you know that story. Yeah, right. Should I recount that for folks?

Up to you.

I mean it's really, it's really important one, please. Especially if you think about bitcoin in the US So pretty good privacy was an email encryption that was invented in the late 90s if I remember right. And the U.S. department of Defense desperately tried to stop the propagation of PGP into the world because then they wouldn't be able to read emails globally, which of course they wanted to do. And the PGP creators were like, well everybody should have this. Why should it just be restricted to this? So the DoD put a full restriction saying itar regulations, you are absolutely forbidden to export this. So they looked at this, said, huh, okay, what do we do? So what they did was they printed the source code for PGP onto pieces onto paper, faxed it to the UK which then OCR the code and were able to recreate Bitcoin. Department of Defense took them to court and the court said, first amendment, pal, you can't stop people from expressing opinion. So there's a really strong precedent on first amendment and then property rights, anywhere that has property rights, you cannot stop bitcoin from propagating. So there's some very compelling arguments to, especially in the U.S. but globally, anywhere where you have property rights of any reasonable level, you won't be able to stop it. They're trying to slow it down on ramps, off ramps, et cetera, but they're not going to be able to stop it.

I just did another podcast with Bill Barheit, head of Abra, which again people go listen to it. That's where I hold my bitcoin. Where do you hold your bitcoin?

Right now I do self custodial on a stick and a few other metrics.

Do you pucker really hard when you pull out that stick to go and do something with it?

Well, it's really what you do with the keywords and how do you secure those? Because the stick is thing. But this is the great thing about capitalism. Safe custodial models like Abra will course of here and people move to Swan is another one that's really good at this. So there'll be a whole bunch of. There's so much economic incentive for helping people with that problem. The equivalent of a digital bank safety deposit box is a really powerful one.

Do you want to give your non financial advice that you gave at Xprize.

Here which was oh, non financial advice would be take 10% of your of your net worth at the minimum and put it into bitcoin and close your eyes for 10 years if you can.

Is it true that you pulled out a mortgage on your house and bought. Bought Bitcoin?

Well, we sold a house, yeah. And my condition with Lily to selling the house, I wanted to be able to put a big chunk of that into Bitcoin. So Bitcoin is almost about the equivalent of my real estate holdings.

Yeah. All right, I'm excited for us to keep this conversation in future months. Let's talk about Neuralink news. Elon has his second implant and it talks about human AI comms. This is one of the conversations that took place a little bit at Abundance the past year is about the idea that if AI and humans are going to link up, or will AI take off without humans? Let's listen to Elon for a moment.

The long term aspiration of Neuralink is to improve the AI human symbiosis by increasing the bandwidth of the communication. Because even if in the most benign scenario of AI, you have to consider that the AI is simply going to get bored waiting for you to spit out a few words. I mean, if the AI can communicate at terabits per second and you're communicating at bits per second, it's like towing to a tree.

Okay, so first of all, if we can communicate at kilobits per second or megabits per second and it's communicating at terabits per second and still going to get bored. Do you remember the movie Her? I just rewatched it and there's a point at which the star of the movie is talking to his AI and he finally gets the idea that it's not a monogamous relationship. And he goes, how many people are you talking to right now? And she goes, 8736 or something like that. And I find that fascinating.

Well, this is the whole bandwidth issue, right? I think that movie. If you haven't watched the movie, her, please do go watch it. It's probably the best framing we've seen on the realistic outcome of AI, because Hollywood tends to always portray it as the Matrix, the Terminator, the whatever, and the robot overlords took over the world. And if you're lucky, you're pets, and if we're unlucky, we're food. It always goes that way. But the reality is much more likely that an AI evolves so quickly past us, they look at us as ants and go, you know, your little one liter brain in your head, Vitalik accepted, is not just not interesting enough to, to talk to, I'm talking to millions of AIs in parallel.

And it takes off. It goes, have a nice life. It just.

Just life.

Thanks for all the fish.

Thanks for all the fish. And I think that's the most likely outcome, which I'm kind of okay with. You know, if that's the progression of life, that's kind of okay. But I think what Elon is doing around this is really, really important. And the next conversation around bandwidth is critical.

Yeah. I mean, right now we communicate. When you and I are speaking, what is it? Bits per second? Is it.

Is it bits per second? BPs?

Yeah. No, no, but it's like. Is it 40 bits per second?

Is single digits? I mean, four bits per second. Yeah. It's not a lot. And you think about the megabits and gigabits that we have on our networks, there's no comparison.

Yeah. So here we go.

Think about this. You're typing on a keyboard, right? I mean, it's literally. That's about as fast as we can do it. There's not a lot of bits per second.

And it's shockingly slow. Let's hear Elon about making it faster.

Nears.

It's going to be gigantic because we'll increase the number of electrodes dramatically. We'll improve the signal processing. So even with only roughly, I don't know, 10, 15% of the electrodes, working with Noland with our first patient, we were able to get to achieve bits per second. That's twice the world record. So I think we'll start vastly exceeding the world record by orders of magnitude in the years to come. So it's like getting to, I don't know, 100 bits per second, thousand.

Maybe.

If, like, five years from now, we might be at a megabit, like, faster than any human could possibly communicate by typing or speaking.

Obviously, I find that compelling. Right. I mean, I want as much as possible. So let me ask you the question here, Salim. They open up neural links to the general public. Are you, like, get me one baby as soon as possible.

No, you're not. No. I'll tell you why. So I go back to the. So I'll tell you why I'm incredibly excited about this and why I'm like, okay, come on, let's have to shave your head.

They could just go straight in.

The USB port on the side of the head is. That's that kind of old dream. I think the reason that it requires physical intervention is a tough one because our phones evolve much faster than any implant would evolve if you put it in. Right. So there's that. You'd have to have this like deep thing going into your brain, which is fine. Once it gets to. I just don't want to be the first test subject in that. In that model.

But I didn't say test subject. I said it's, you know, the FDA has approved it. It's commercially available now. Are you going to get one?

I would wait a little. I'm always, weirdly, I'm a late adopter in new technologies. I'm always two, two phones behind everybody else just because I'm. I use it so materially that I can't afford to be at the bleeding edge and have some stuff work or lose things, etc. So that's just me personally, but the. I think why I'm excited and why I'm on this, I'm incredibly excited at the general conversation of increasing output of the human brain. That's a really powerful framing and I think it's. I love the first principle thinking that Elon brings to the table on any of these conversations that brings everybody's thinking up to that level and then you have to think in that way. And it's the fact that he's, without realizing it, coaching tens of millions of entrepreneurs around the world to think in this way, because he does, is incredibly exciting for the future. Here's where I get a little bit more. If you think about it, he kind of makes this comment in this clip that we put out. X amount of roughly a bit per second is our output if you average it across the day. But, you know, we just talked about why in person is so important. Because when you're in person, face to face with somebody, there's a ton more happening than just my communication patterns back and forth. There's the pheromones we're putting out, there's our body language, there's visual cues, there's all sorts of verbal tonality that goes along. There's a ton of bandwidth occurring between two people on a conversation. Much, much more than just the pure bits per second being output by the brain.

But you know, you know as well as I do that all of that information is being converted to neuronal signals.

It is.

And your brain is only able to process a small number. So as we're having this conversation, you know, there's traffic noise going on outside, there's air conditioning comes on. And we don't notice that because I'm focused on the wisdom that Celine Ismail is speaking to me about right now.

Yes.

And so there's no reason why, in fact, an AI can't actually pay attention to all of it.

Agreed.

Do pre processing and share with me the most important information because I'm ignoring 99.9% of the information coming to me right now. And I have an attention focus on what I think is important and it may not actually be important or there may be other important information that I'm missing.

My point is that we have all sorts of mechanisms to do that for us in many cases. So I'll give you a small example. You and I both follow a few thousand people on Twitter or X and then those folks we found curate the information out on the web and news stories and breaking tech news in a powerful way. And that's the pre processing it filters down for us. Right. So we do a pretty good job of that. And AI would do an incrementally better job but not a game changing better job at that. If I'm sitting in front of a person, all the signals that I get are useful about the world. But it's not the fact that I might be able to look at the signals of 10,000 people in a room in a big conference and get some sense of that. But it's not probably going to affect my decision making on a moment by moment basis where I think, let me flip it around and tell you why I'm excited. If you can increase the bandwidth levels and then have brains interfacing more powerfully, the potential I think it'll reshape our brains unbelievably powerfully. Is a mechanism for we need a. If I had to wave a wand and fix a human being, human beings globally as a species, I would basically find a way of cutting out the amygdala. Right? You talk about this. It's constantly scanning for bad news and danger and it takes over the rest of the brain and floods it with cortisol every time it perceives that it's a completely useless device for today's world because the amount of time we're in actual physical danger is seconds out of a year. And so let me give you, let me give you a broader so much potential to that if we have the ability to interact with high bandwidth, we could actually then emphasize the neocortex and of more processing power at a rational level and diminish the impact of the amygdala on us. End of rent.

So I agree with you and we can go broad on that. The amygdala is just your negativity bias where you pay more attention, 10 times more attention to negative news and positive news. But we have all these other cognitive biases. There's a familiarity bias, there is a recency bias, all of these things.

It's like we suffer from positivity bias.

We do, we do, though. I think that's unbiased, but that's a different story. But imagine you're having AI that's able to tell you, listen, that's not actually true. You have this piece of information from six months ago which countervails that. And let me bring that forward. So in other words, human bias alert. And we can call it bullshit alert, but human bias alert. I think AI could play an incredibly good role because the reason we have cognitive biases is because the human brain cannot process all the information being thrown at it. So we take these shortcuts.

Yeah. Okay, so a visual is coming to me. Okay.

Okay.

Imagine I have a headset display and I'm walking around the world doing whatever, and I'm buying a bar of chocolate and an AI is scanning my brain and constantly bringing up. These are the biases you're operating under right now. These are the cognitive biases that are forcing. And just having that feedback loop of knowing what biases are at play at any given point would be incredibly powerful. And give me a feedback loop, because this is where I think things become really exciting, is the recursive feedback loop. When you know what you're thinking and why you're thinking something that allows you then to reshape our thinking. This is the whole basis behind but we're LP and but we're so limited.

Right now by this, you know, you know, two and a half kilogram, you know, unit above 14 watts of power. I just, I just can't wait to, to plug in and expand beyond that. But, you know, that's me.

I'm going to throw one more little thing in this.

Okay.

Okay. You know, just, you know, I've done a lot of mindfulness meditation, vipassana, where you sit and you just be aware of what's happening in your body. Just imagine that I could be fully aware of all the activity in every cell in my hands right now. And I could be fully cognizant the excitement and energy around every cell in your body, which has its own mini sub universes.

40 trillion cells, a billion chemical reactions per second, just.

That is a profound thought in terms of. Could I just contemplate that more profoundly? Forget interfaces elsewhere, etc. Etc. So there's unbelievable.

With or without psychedelic drugs, well, they.

Give us a shortcut to what's possible. And I think there's the, the ability to contemplate a broader, with broader bandwidth. The unbelievable reality we live in is amazing.

All right, let's go to Elon again on, on patience outperforming gamers.

We feel pretty confident that I think maybe within the next year or two that someone with a neuralink implant will be able to outperform a pro gamer.

Nice.

Because the reaction time would be faster.

You know what's fascinating? I remember when I did a long form conversation with Elon after we announced the $100 million Musk carbon prize. I was asking him about the first Neuralink patient which was chimp, it was chimpanzee called Pager and it was playing Pong and I said did you play against Pager? And he said yes. And I said who won? And I think Pager won, but I think Elon may have said he did. I don't know. Well, but it's fascinating that we're going to get there. The Olympics are looking at esports as an Olympic game and we're going to question is will you be able to be enhanced when you're playing those games? That'll be fascinating.

I mean this is a long trend, right. Which I think we have really good evidence today that the best leadership skills in the world are taught if you play World of Warcraft.

So do you let your son play World of Warcraft?

Not yet. Because he's. I think he's too young and he probably plays that stuff with. He plays half of that stuff without telling me. Anyway the. But I think the opportunity to really develop powerful skill, deep human skills using this capability is massive. I thought, you know, when you talk about AIs beating the pro gamers etc, that's great but we've seen that in chess, right. And we know the outcome, what happens, it just changes.

The difference here is, you know again, directly from, from the Neocortex and, and these are. A lot of, these are twitch games, a lot of these are reaction time games. Right. And so when the reaction time is minimized because you're going straight, you know, input output, I mean it's, you know, I don't know. Is that cheating?

It is to some extent. It's just doing things better and differently. Gamers spend a ton of money buying like a mouse that is connected directly into the motherboard so they have that few microseconds advantage. For me it's eh. For a lot of this. But the general concept of interacting with the brain and having a broader bandwidth to process do signal processing is incredible.

Here's Our last slide here, and I think this is the numbers folks take away. Prior to neuralink, the world record for a human using a device is somewhere around 4.2 to 4.6 bits per second. Noland is at 8.5 and we heard Elon in a previous video talk about going to Megabit. When I think about the notion that there's nothing more important for a company or for a nation than its cumulative intelligence, the ability of a company to perform its mission is a function of per capita intelligence. I think maybe you disagree with that.

I disagree with that.

Okay, tell me where you come from.

I'm going to bring out my exo bias here. A huge amount has to do with how it's organized, right?

Yeah.

If you have all of the most.

Powerful, let's assume they're organized identical, but one group is more intelligent than the other.

Yeah, of course that's going to do better. But the organization always beats bandwidth and brain power.

But both when given a choice. Take both.

Here you can do both. But I think there's so much mileage to be gained by just organizing better and not being top down, command controlled, slow decision making. That's what limits intelligence, for example, the basic intelligence in any institution. Take the nuclear energy discussion we were having earlier. That limitation is not a function of intelligence. It's a function of the ability to execute and recognize. We need a different framework for doing something and operating from a different model in a different framework. It's the changing of models that's hard.

We just made my point, which is intelligence and regulators. Different.

Conversation, different page, different motivations. It's a motivational alignment problem more than anything else.

Did you see the movie Oppenheimer? If you did, did you know that besides building the atomic bomb at Los Alamos National Labs that they spent billions on biodefense weapons, the ability to accurately detect viruses and microbes by reading their rna. Well, a company called viome exclusively licensed the technology from Los Alamos Labs to build a platform that can measure your microbiome and the RNA in your blood. Now, viome has a product that I've personally used for years called Full Body Intelligence, which collects a few drops of your blood, spit and stool and can tell you so much about your health. They've tested over 700,000 individuals and used their AI models to deliver members critical health guidance like what foods you should eat, what foods you shouldn't eat, as well as your supplements and probiotics, Your biological age and other deep health insights. And the results of the recommendations are nothing short of stellar as reported in the American Journal of Lifestyle Medicine. After just six months of following Biome's recommendations, members reported the following a 36% reduction in depression, a 40% reduction in anxiety, a 30% reduction in diabetes and a 48% reduction in IBS. Listen, I've been using Viome for three years. I know that my oral and gut health is one of my highest priorities. Best of all, VIOME is affordable, which is part of my mission to democratize health. If you want to join me on this journey, go to viome.com Peter I've asked Naveen Jain, a friend of mine who's the founder and CEO of viome, to give my listeners and a special discount. You'll find it@viome.com Peter let's talk about robotic news next. Shall we jump into that?

All right, let's go to the robotics.

All right, so we're seeing a lot in the world if I think about the robotics right now. We talked about this again at the Abundance Summit. There are like 30 well funded humanoid robot companies on the planet. Optimus and Figure are Tesla. They're robots called Optimus and Figure. Brett Adcock's company just release figure 2. We'll talk about that in a minute. Are the top two in the US and there's a whole bunch in China because China needs robots. Let's listen to a quick note from Elon on the numbers, the size of the market and then we'll talk about about this in more detail.

The optimist numbers like they're really just, they're like so mind blowing that you're like is this, is this real? But because I, I actually think the market for humanoid robots is in excess of 10 billion units. Like more, more than the number of humans because people each want one and then there'll be others that are involved in industry and stuff. So if they do sell for $20,000, that's $200 trillion. This is just an insane number. That's why I wonder what does money.

Even mean at that point? What does money even mean at that point? Saleem the numbers here to reflect on. And I got these exact same numbers from Brett Adcock when I did a podcast with him recently. $20,000 a unit, that's what the price sort of converges on and that's like $100 a month if you're going to lease it and you know, 10 billion of them in the world.

Well I think again have I like going to the extremes and just considering both sides of this. I think at one level. It's incredibly exciting. The best framing and metaphor of robots comes from the AI conversation with Imad where he goes, when you have a bunch of, you're the head chef and you have a bunch of sous chefs organizing everything for you, you become way more creative, you execute way faster. Because I can right now, if I had a sous chef in the kitchen and I want to make a really complicated Indian recipe, like I spent a huge amount of time gathering all the frickin ingredients, just that would be a huge game changer in terms of my ability to be a great chef. Right? So I think this is where robots are able to do amazing things over time. But I think the entire economics numbers become irrelevant just because the change in productivity is such a game changer. None of our numbers in terms of GDP or something can match what's going on.

Economics, economics are so broken, right? I mean GDP is a function of the number of people working and the amount of technology and the amount of energy they have access to. And all those things. And all those things are infinitely.

It's actually terrible because it doesn't take account deflation. We spend half a million dollars per breast cancer patient in the US on average. If you have the ability to detect breast cancer early and save all those lives, GDP actually goes down. That's not a great. GDP is a terrible metric for success of the human race in the future. It's just the best one we come up with that's easily measurable for the future. So this is an important point where technology becomes deflationary and we have to navigate that for the future. But that's again a whole other conversation.

I'm going to play this in the background as we're talking about this is Figure 2 from, from a company called Figure AI. And you know, I went and visited his plant and on our podcast with Brett he shows us a little bit inside and it's a beautiful robot. I actually saw Designs for Figure 3 as well, which take this yet to another level of beauty. And I don't know, I think that we're going to see these every place. And so we've had the argument about you said why do they have to look like humans? I think they're going to look like humans because that's the world we live in and they are multi purpose. And my conversation with Brett, who he just raised like $700 million from OpenAI and Microsoft and Nvidia and Jeff Bezos and incredibly what makes this possible is the multimodal AI where you're going to just Talk to robots and tell them to go do things for you. He made a point. Let me make one other thing. He made a point that I think is really important, and I would love your reflection on this. The point that Brett made was, listen, as we head towards AGI, or digital superintelligence, we need robots. Otherwise these AIs are going to be asking us to do their bidding, and that's going to pretty depressing future.

Okay, so very valid point. I still sit on the cautious and pessimistic side on this. Going back to the autonomous driving mode. I'll give you one other triangulation on that, which is the Roomba vacuum cleaner robot. You know, everybody got really excited about those, and they're terrible because you have to move everything for you, for the robot, and now you're doing the work of the robot, which is exactly the point that you just made. Right. Everything you do is servicing so the robot can do, and you spend more time rearranging the room so the robot can do its thing than if you just vacuum the place in the first place.

But these humanoid robots will grab the vacuum cleaner.

But I think that those small nuances will take a lot longer to work in there than we think. Where I get really excited about this, by the way, is the hive mind capability that these robots that I think will be very exciting.

Yeah, I'm going to skip the Nvidia robot slide, other than to say every company out there is in some way, shape or form looking at working with robotics. OpenAI's got other projects, Google's got its projects, and we're going to see robots. I just tweeted that. Damn it. I posted, I tweeted. I posted. Give me a damn verb, Elon, please. I posted that. Ces, this year we're going to see a ton of robots, but everything is going to be talking to you, right? Everything is going to be generative AI enabled. It can see you and talk to you, and that's going to be just sort of improvement du jour at CES this year.

I think it'll be hugely amazing. There's not often I kind of feel tempted to go to the chaos and jungle that is ces, but this might be a year to go.

Yeah. We'll go to our last subject here, which is biotech and other cool stuff. This was a article that came out and there's a particular genetic variation that was found that increased lifespan in mice and by 25%. I think what makes this interesting is it's actually in human trials as well. I think what I want to point out by this slide is that we're continuing to find things that are extending human lifespan or extending mammalian lifespan and that we're just at the beginning of this journey. It's still early games here. How long do you want to live?

How long do I want to live? You know, I'm, I've. There's a big metaphysical conversation that we don't want to get into here about living forever, etc. Etc. I think the best framing is having a great health span, which you talk about a lot, right? Is that's a really important one because having a vibrant life as opposed to just a long life is really, really a key success factor. I think the world is so goddamn amazing. I would like to be around for a while, long enough to see this massive transition that we're going through which, you know, I say this often. I think the next 30 years is going to drive the next several centuries of human humanity.

Oh, I think it's.

And I think we navigate The Mad Max vs Star Trek futures that we have is going to be really, really interesting and critical. And it's just we, you know, goes back to your comment about we must be in a simulation because it's too, just too goddamn interesting today to be around.

I mean we're at the 99th level of the game. I mean to look at it that way, right, it's like it would. The, the we are about to evolve into something new. It's like that line from 2001 the Space Odyssey, which I can't remember, but it's like the future is just unbelievably awe inspiring. It's the singularity. By the way, can I ask you a quick question? Ray keeps on holding to the singularity date of 2045. I mean that seems like really if we've got Digital Superintelligence by 2030, don't you think it's going to move a little faster than 2045? What's up with that?

So I'm right now plowing through the singularity as nearer.

Yeah, I just finished it.

And what I'm struck by as far as I've gotten into it, is a. The unbelievable prescience of his commentary from 2030, 40, 50 years ago. This just blows your mind. Who is this man? I mean he can't be human to be making these types of comments. That's one observation. The second observation is I notice he changes what the singularity is to talk about more as a metaphor than a real thing, which I think is the right way of framing it. It's a great metaphor for the fact that we just can't predict past a certain event horizon. Let's just live with the fact that.

We can't predict in that case. We have serial singularities.

We always talked about this at Singularity University where there's nonstop small S's happening all the time. The iPhone arriving was a singularity. Right.

In other words, that you couldn't, you can call it, you couldn't predict Uber.

You call it asteroid impacts, I call it Gutenberg moments. There's black swans. There's just an increasing frequency of those happening at an, at a, at a ridiculous level. That's going to just change everything we do and how we do it going forward. And the ability to absorb that and then act on it and then make sense of it is the biggest intelligence quotient we're going to need. We're going to have to have over the next decade and two.

Yep, agreed. But I mean, I think of a AI singularity coming where I can't predict what's coming after.

I think that's already there. In my open exo world. We have a sub community of about 600 folks focused just on AI and they meet weekly just to try and process what happened this week in AI and, and try and make sense of it. I mean it's just, it's the. I would say that we've hit the singularity already. In AI, the pace of change is faster than we can even talk about it or recognize it and make sense of it.

Yep, I, I get it. All right, let's go on. I think one of our last slides here. LifeBio is a company that David Sinclair had started. David made history when he showed that epigenetic reprogramming using three out of the four Yamanaka factors was able to bring back the earlier state of youth of cells. In particular it was in the optic nerve. This is Dr. Sharon. I'm not going to. Rosenweig Lipson, who's the chief science officer for LifeBio for David's company. Let's quick, take a quick listen here.

What we're doing is we're building on the work of Dr. David Sinclair where he identified that you only need three of the Yamanaka factors off 4 Sox 2 and KLF 4, which allows you to not take us all the way back to a stem cell state, but rather to make a more youthful version of itself. We do it in. Our version of it is an ER100. ER100 is a two vector system. We have two vectors and it allows us to use a TET on docs inducible system short form. It means we can control it. Hoping to be in the clinic in 2025.

So the point here, which is pretty amazing, is that in the clinic means for humans and we're using epigenetic age reversal in humans in particular. I think they're going to be doing it for a number of eye related diseases and blindness reversal. But they're in primates right now and it's working. And I just want people to get excited about the future in terms of age slowing, stopping, perhaps reversing aging. And God knows there's a lot of implications.

The broader implications, what do we do with pension plans and unemployment benefits, all that stuff is the structural changes just coming from life extension are unbelievable. But I'm really, really excited about this. I know you spend a huge amount of your time now on the longevity stuff and you're kind of doing an unbelievable job of popularizing it and getting people used to the conversation. Right. I think that's the hard, the key part. This is you want to talk about singularities. Breaking through the aging stuff is something that we've not really been able to do as a human species in any really meaningful way for our entire 4 billion year history of Earth. And now all of a sudden we're hitting this point where we can completely change the game of biology. It's amazing.

Yeah. All right, last subject here for today, life on Mars. So we saw some data come back from NASA's Perseverance rover which is cruising around getting data. What it did was it found a particular rock that in that rock were three elements that were critical. First, it found water in the rock. Second, it found organic compounds that were part of life and precursors. And then it found evidence for chemical energy sources in the rock. And when you have those three things together, typically that spells life. I just saw this morning an article that, that they also found liquid water on the surface of Mars. So, you know, I'm excited about two, three things. Number one, the notion that I think there is life on Mars today, it's microbial, it's subsurface, it's in the permafrost or in the water. And just finding life on another planet just changes everything. The second thing is.

Go ahead.

I'd say the second thing is going to be interesting is there is a theory that because Mars cooled first, because it's farther from the sun, that life evolved on Mars first. And there's an exchange of material between Mars and Earth. As asteroids strike the Martian surface, the ejecta reaches escape velocity. And so it's going to be interesting if we find life on Mars and compare it to life on Earth. Did it originate on Mars first? And then there's the whole panspermium model, which is life evolved elsewhere in the universe and it's just floating through space and it seeds planets as it intersects. All right, now over to your wisdom.

Well, I mean, there's no question in my mind of the panspermia model being the real one, because we found tardigrades can survive in space where they should. There's absolutely no reason that, that a life form should be able to survive in space if it originated in Earth. And yet here we have them. So this was pretty clear that life has evolved in many, many, many, many places. It goes to the whole Drake equation, which we don't need to get into right now, but if you want to check it out. But the reality of life being pervasive around the Earth, the only then question is how do we dance with Fermi paradox, which I think we have, a reasonably good Fermi paradox is being there must be life all over the goddamn universe. Why haven't we seen aliens?

Where the hell are they? Why haven't they brought me home?

And the best model, the answer I've heard is the transcension hypothesis by John Smart. You get to virtual reality technology and you can simulate anything so you go inward rather than trying to go outward because it's just too hard physically to outward. Do believe that that reality. This is not surprising. I'm just thankful that we've actually found it so that we can shut down all the religious idiots. Sorry.

Just to be clear, that article doesn't say we found, we found the precursors for it.

I know, but, but one of the big questions in everybody, I think is where's the primordial ooze that we think we've seen? And here we have it on another planet. And by the way, little by little, we're filling in the gaps of how.

Life Ganymede, Europa, the moons around Jupiter and Saturn are likely to have even more advanced life. What's fun is we're alive during this period of time. Right? Starship Flight 5 is going to be going up soon. Can't wait for us to talk about that in our next podcast conversation. And it is going to bring people, 100 people to the surface of moon and Mars and take us out to the outer planets. It's going to be the, you know, the whatever you want to call it what was the trend? What was the most important, like the 707 of aircraft, right. That became the KC135 tankers and really first transatlantic, really wide body jet. It's going to allow us to go there and we're going to find out.

I'm going to do a final plug here for xprize because notice that you created XPRIZE because you noticed that Lindbergh crossed the Atlantic and all the teams competing developed ancillary technologies that created the commercial aviation industry. Right? And boom, you put a honeypot down for the space world. Now we have a close to a trillion dollar space industry and we're doing that nonstop across the board. So the future is so, so exciting.

It is. And I didn't mention in the beginning, Saleem Salim is also a member of our board of directors of the Ex Prize Foundation. He's godfather to my two kids. I'm going to add your additional. Anyway, I love you so much, buddy. And this was a fun conversation. I really, really enjoyed it.

It's super important and I want to debate hotly some of the other topics. We'll do that next time.

Yeah. And if you're enjoying these conversations with Salima and myself, please subscribe. Let us know. I'd like to do this once a month where Salim and I get together and argue and disagree on all of these things. Salim, where do people find you and your companies?

Aleemismail is on x, Twitter and openexo.com yeah.

All right, buddy. You're off on vacation.

I'm off to Indonesia and India, family stuff. And then back in a couple of weeks.

You must be a God when you land in India. I mean, you're Indian. And I'm sure I try and hide.

Actually, because I'm scared that once people see that, that they'll steal my passport and I won't be able to get.

That to keep you there.

So.

All right. Be well.

Take care.

Take.