#407 – Guillaume Verdon: Beff Jezos, E/acc Movement, Physics, Computation & AGI
03:00:12 Share

The following is a conversation with gill were done the men behind the previous and anonymous account based of J S on x. These two identities were merged by a dozing article in forbes titled who is based by, as the leader of the tech elites, E. X. movement. So let me describe these two identities that coexist in the mind of one human identity.

Number one, give is a physicist, applied mathematician and quana machine learning research and engineer, receiving his P, H, D, quana machine learning, working at google, a quantum computing, and finally launching his own company called extort pic six to build physics based computing hardware for generative A I identic. Number two, love jesus and x is the creator of the effective acceleration as a movement, often breathe as e that advocate for property rapidly cho logical progress as the ethical optimal course of action for humanity. For example, its proponents believe that progress in A I is a great social equalizer, which should be pushed forward.

Ex followers see themselves ves as a conor weight to the cautious view that A I is highly unpredictable, potentially dangerous and needs to be regulated. They often give their opponent the labels of costumers or short for declaration, as both himself put IT. Iac is a memetic optimism virus.

The style of communication of this movement leads always toward the memes in the laws, but there is an intellectual foundation that we explore in this conversation. Now, speaking of the mean, I am to a kind of aspiring carnesi of the epsom IT is not an accident that I spoke to. Jeff, baths and death, jesus, back to back as we talk about, both admired.

Jeff is one of the most important humans alive, and I admire the beautiful of certainty and the humor of IT all. Now a quick you second mention of sponsor jack out in the description is the best way to support this podcast. We ve got element for hydration, the thing i'm drinking right now.

Notion for team collaboration, inside track of four biological data, at least two year well being, and A G one for my daily nutritional health. Choose wise than my friends. Also, he went to work with our amazing team.

Where was hiring at the next experiment that cause as hiring, or if you wanna get in touch with me, for whatever reason, got a extreme that can flash contact. And now onto the full ad rates. As always, no eyes in the middle.

I try to make this interesting, but if you must skip them, friends, please. So check out our sponsors. I enjoy this.

Maybe you will too. This episode is brought you by element. Element ate drink. Mix is got something patashie magnesium. I drink IT so much, so many times a day.

It's really the foundation of my one million day life style, like he almost always one million a day in the evening. So I fast, and I really enjoy that everything he does for me. I recommend everybody least try IT into my fasting, taking to this of daily extreme of, you know, fasting for twenty three, twenty four hours, whatever IT is.

And for that you have to get all the electrized right. You have to drink water, but not just drink water. To drink water, couples with sodium, and sometimes getting the magnesium import the potassium part right is tRicky, but really the important, so that if you good, and that's what element does.

And IT mixit, delicious, my favorite flavorous watermelon salt. Get a simple back for free with any purchase. Tried a drink element to cos lashes. This shows also brought by a notion, a note taking, a tim collaboration tool.

I ve used them for a long, long time for no taking, but it's also very useful for no taking and all kind of collaborative no taking in the team environment. And they integrate the whole A I think a little one thing. Well, so you can use IT to summarize whatever you've written.

You can expand that. You can change the language style and how written. Just all the things that large language models should be able to do are integrated really, really, really well.

I think of human AI collaboration, not just as a boost for productivity at this time, but as a kind of learning process that IT takes time to really understand what I has good at and not, and that is going to solve continuous. They say I get Better, Better and Better. It's almost watching a child group something like this.

You're fine tuning what he means to be a good parent as a child goes up in the same way. You're fine tuning what he means to be a good, effective human as the A I grows up. And so you should use the tool that's party a daily life, interact with the eye while being productive, but also learning what is a good at, what are the ways I can integrate into my life to make me more productive.

But not just like in terms of shorting the time, IT takes a dual task, but being the fuel, the creative fuel for the genius that is you. So notionate, I can now give you instant answers to your questions using information from across your worker project docks meeting notes. Try notion A I for free.

When you got a notion that can flash legs. That's all lower case notion. Come lass likes to try the power of notion A I today. The show also brought you by inside tracker service.

I used to a make sense of the biological data that comes from my body, blood data, DNA data, finder data, all of that to make me a lifestyle recommendations. Diet stuff, too. This is all this beautiful data, which he give IT to superintelligent computational systems to process and to give us in a human interpretation way recommendations um how to improve our life.

And I don't just mean optimize life cause I think a perfect life is not the life you want. What you want is complicated roll cost of life, but one that is optimized in certain aspects of health, well being, you know, energy, but not just optimum, is called clinical sense. Anyway, that's a longer conversation.

Probably one touch on maybe when I review brave world or in other conversations that have in the podcast any special savings for you inside the slash legs. This shows also about you by A G one. The thing I drink twice a day, and that brings me much joy.

Its Green is delicious. It's got a lot of items of minerals, is basically just an incredible superpowered multivariate. I enjoy IT.

A lot of my friends enjoy IT. It's the thing that makes me feel like common on traveling, and I get one of the travelling backs. The things I consume daily are pretty simple.

We're talking about their legalized with element a, you one for the mines, minerals, then a fish al, and then just a good healthy diet, low carb, but either alter very low carve. So just me or millions of vegans, but i'm not very strict about that kind of stuff. Just know that I feel good.

I want a little m so all that combined with fasting and rust, sometimes crazy routines of work, so mental struggle and physical work, you know, running in all that I stuff you too, training spins, all working out, lifting heavy, all that kind of stuff. You have to make sure you have the the basic nutrition stuff, right? And that's what if you want us for me, maybe I will do that for you.

They'll give you a one month supply fishel. When you sign up a drink, a you want that comes flash legs. This is the leg treatment podcast. To support IT, please check out our sponsors in the description, and now your friends has gone for them.

Let's get the facts of identic down first. Your name is gee verdon gill but you're also behind the anonymous account on ex called based but jesus so first overdone you're a quana computing guy, physicists, applied mathematician and then based by jesus uh basically a mean account that started a movement with the philosopher hind. So maybe just killing on who these people are in terms of characters, in terms of communication styles, in terms of philosophes.

I mean, with my main identity, I guess since I was a kid, I wanted to figure out a theory of everything to understand in the universe and, uh, that path um let me to the medical physics eventually, right, try to answer the big questions of where are we here? Where are we going right? And that LED me to study information theory and try to understand physics from the lens of information theory, understand the universe is one big computation. And essentially, after reaching a certain level, studying black hole physics, I realized that I wanted to not only understand how the universe compute, but sort of compute, uh, like nature ah and figure out how to build and and apply uh computers that aren't spired by nature.

So you know physics based computers and that sort of brought me too quantum computing as A A field of study to um first of val simulate nature and in my work I was too learn representations of nature that can run on such computers so if you have A I representations that think like nature and then they'll be able to more accurately represented at least those the the thesis that that brought me to be an early player in the field called quantum machine learning right so how to do machine learning on on quantum computers um and really sort of extend uh notions of intelligence to to the quantum room. So how do you capture uh and understand quantum m mechanical data from our world right and how do you learn conney mechanical representations of our world on what kind of computer do you run these representations and train them? How do you do so? And so that's really sort of questions I was looking to answer because ultimately I had a sort of crisis of faith.

Uh, originally I wanted to figure out, you know, as every physicist does at the beginning of their career, a few equations that described the whole universe right and and sort of beta the hero of the story there um but eventually I realized that actually augmenting ourselves with machines, augmenting our ability to perceive, predicting, control our world with machines, is the path forward. That's what thought me, at least the radical physics, and go into quantum computing and quantum machine learning. And during those years, I thought that there is still peace missing.

There is a peace of our understanding of the world and are our way to compute and our way to think about the world. And if you look at the physical scales, right at the very small skills, things are going on mechanical, and and at the very large scales, things are deterministic. Things have averaged out, right? I'm definitely here in the seat.

I'm none at a superposition over over here and there at the very small skills things are in superposition. They can uh exhibit h interference uh effects um but at the measure scales, right the scales that matter for day to day life, you know the scales of proteins, of biology, of gases, liquids and so on, things are actually thermodynamic right. They're fluctuating.

And after, I guess about eight years in in quantum computing and quantum machine learning, I had a realization that I was I was looking for answers a about our universe by studying the very big and the very small right. I I did a bit of quantum Cosmology so that studying the Cosmos, where it's going, where I came from, you study black of physics, you study the extremes and quantum gravity. You study where the energy density is sufficient for both quanto mechanics and gravity to be relevant, right? And the sort of extreme scenarios, or black holes, and you know the very early universe.

So there is this sort of scenarios that you you study the interface between uh, uh, quantum mechanics and and relativity and and you know really are studying these extremes to understand how the universe works in. Where is he going? But I was missing a lot of the meat in the middle, if you will, right? Because day to day equal mechanics is relevant and the Cosmos is relevant, but not that relevant actually run sort of the medium space and time skills and there are the main in theory of physics that is most revere is thermodon x rain out of a equilibrium through my dynamics um because life is you know a process that is the myn amal and it's out of elimination were not um you know just a soup of particles at a calibrate win nature where a sort of coherent state trying to maintain itself by acquiring free energy and consuming IT and that sort of um I guess another shift and I guess my faith in universe happened and towards the end of my time and and alphabet and I knew I wanted to build a well first of all a computing paradigm based on this type of physics um but ultimately just by track experiment uh with these ideas applied to I D and economies and a much of what we see around us. I started an anonymous account just to relieve the pressure right that comes from having an account that you accountable for everything you say on um and I started an anonymous account just experiment with ideas and right because I didn't realize how much I was restricting my space of thoughts until I sort of had the opportunity to let go, in a sense, restricting your speech back propagate to restricting your thoughts, right? And by creating anonymous account, IT seem like I had unclimbed some variables in my brain and simply could explore a much wider premier space of a thoughts.

Just going that isn't interesting that we wanna think that people are often talk about is that when is pressure and constraints on speech? IT somehow leads to constrain some thought, even though doesn't have to, can think IT thoughts inside ahead. But somehow we creates these are walls around thought yes.

that sort of the basis of of our movement is we were seeing a tendency towards, uh, constraint reduction or suppression of variance in every aspect of life. Where's thought? How to run a company, how to organize humans, how to do A I research in general, we we believe that maintaining various ensures that the system is adaptive, right maintaining healthy competition in marketplaces of ideas, of companies, of products, of cultures.

of governments.

of currencies is the way forward because the system always adapts to assign resources to um the configurations that lead to its growth and the fundamental basis for the movement is this sort of realization that life is a sort of fire that seeks out free energy in the universe and six to grow right and that growth, this fundamental to life and and and you see this in in the equations actually have either collegium termini ics.

You see that path uh of trajectories, of configurations, a matter that are Better at acquiring free energy and dissipating more heat are exponentially more likely right? The universe is biased towards certain futures. And so there's a natural a direction where the whole system wants to go.

So the second law for mada nomics says that the entrepreneurs increasing the universe is standing towards glib um and you're saying there's these pockets that have complexity. And our auto qual, liberum, you said that their dynamic favorite equation of complex life that increases its capability to use energy to offload entropy, to offload entropy. So you have pockets of non entropy attend the opposite direction. Why is that intuit to you that is natural for such pakistan merge.

Well, we're far more efficient at producing heat then, let's say, just a rock with a similar mass as ourselves, right? We acquire you know, free energy and require food and we're using all the electricity, uh uh, for our Operation. And so the universe wants to produce more entropy.

And by having life, uh, go on and grow, it's actually more optimal at producing entropy because it'll seek out pockets of free energy uh and and burnett for its sustains and further growth and um you know that sort of the basis of life and amin, there's a jermin england right at M I T who has the theory that i'm a proponent of that in life emerged because of this a sort of property and and to me, this physics is what govern the measure scales. And so it's the missing piece between the quantum and the Cosmos is the middle part. The dynamics rules the most skills.

And to me, both from a point of view of designing or engineering devices that harness that physics and trying to understand the world through the lens of their nyama s has been sort of A A energy between my two identities over the past year and a half now. And so that's really how that's really how the two identities emerged. One was kind of um you know about my disney respected scientists and I was going towards doing a start up in the space and trying to be a pioneer of a new kind of physical space.

D I. And as a dual to that, I was sort of experimenting with philosophical thoughts you from a physicist standpoint, right? Um and ultimately, I think that around that time there was like late twenty twenty one, early twenty twenty two. I think there is just a lot of pessimism about the future in general and pessimism about attack. And that pessimism was sort of virtually spreading because, uh, IT was getting our gross micky amplified and um you know people just felt like the future is going to be worse than the present.

And to me, that is a very fundamentally destructive force in the universe is this sort of doom mindset because IT is hyper science which means that if you believe IT, you're increasing the likelihood of IT happening and so felt a responsibility to some extent to um make people aware of the trajectory of civilization in the natural tendency of the system to adapt towards its growth. And sort of that actually the laws of physics say that the futures is going to be Better and grander statistically. And we we can make IT so and if you believe in and if you believe the future would be Better and you believe you have an agency to make that happen, you're actually increasing the likelihood of that Better future happening.

And so I sort of felt responsibility to sort of engineer in movement of viral optimism about the future and build a community of people supporting each other to build and and do hard things. Do the things that need to be done for us to the scale civilization um because at least to me, I don't think stagnation 而 slowing down is actually in an option fundamentally life and and the whole system or whole civilization wants to grow and there's just far more CoOperation when the system is growing rather than when is declining。 And you have to decide how to split the pie. And so i've baLanced both identities so far. Um but I guess recently the two have been merged more or less without my consent.

So he said a lot of really interesting things. There is a first representations of nature that's something that first drew you in to try to understand from according puting perspective, like how to understand nature, had you representing you in order to understand that, in order to simulate, in order to do something with that. So it's a question of representations.

And then there's that leap you take from the quana mechanical representation to the where you call a muscle scale representation, where where themal dynamics comes to play, which is a way to represent nature, order to understand what life, a human behavior, all this kind of stuff is happening here on earth. That seems interesting to us. Then there's the word hypertension.

So some ideas expose both pessimism and optimism of such ideas that if you internalize them, you impart, make that idea reality. So both optimism and pests, and have that property. I would say that probably a lot of ideas have that property, which is one of interest things about humans.

And you talk about one testing difference also between this sort of the guy ond the girl of front end and the based by jesus back end is the communication styles also that you are expLoring different ways of a communicating that can be more viral there in the way that we communicate in the twenty first century. Also, the movement that you mentioned you is not just the mean account, but there's also a name to IT called effective acceleration ism E X, A play, a resistance to the effective altruism movement. Also interesting one that not love to talk about the attention ons there OK.

And so then there's a merger. I get get merge on the personalities recently without your consent. Like you said, a some journalists figured out that you are on the same. Maybe you could talk about that experience first of all, like what what's the story of of a the merger of the two right?

So I wrote the manifesto with Michael, founder of yao, an account name base lord still anonymous um and hopefully forever .

so based both jesus and and based like asian like base lord, like asian beijing ord base base lord OK and so we should say from now on was when you say E, I came in e lash, A C, C, which stands for effective acceleration ism and you're referring to a manifesto written, I get substance yeah.

Are you also based?

Lord, no, okay, it's a different person. Yeah okay 我 这个 我在 be 翻译 because that .

be amazing。 So originally wrote the manifesto round same time out founded on this company and I worked that to relax just x now or alphabet x now that there's another x um and there you know the baseline is sort of secure.

Cy, right, you can talk about what you work on even with other googlers uh or externally so that was kind of deeply in grand in my way to do things， especially in in deep tech that you know has geopolitical impact, right um and so I was being secretive about what is working on. There is no coral lation between my company and my main identity publicly. And then not only did they correct that, they also correlated my main identity and this account.

So I think the fact that they had docks the whole gym complex um and they were the journalist you know reached out to action my investors uh which is pretty scary ah you know when you're start up entrepreneur, you really have bosses except for your investors right and investors pink me like haz this is going to come out. They they figured out everything. What what are you going to do right?

Um so I think at first they had a first reporter on the thursday and didn't have all the pieces together. But then they looked at their notes across the organization and the sensor fused their notes, and now they had way too much. Uh, and that's when I got worried because they said he was of public interest. And in general.

lucky sense refused .

the .

stems giant, your own network Operating, stripped away. We should also say that the journalists used, I guess at the end of the day, audio based analyses of voice in pairing, voice of what talks are you given in the past and then a voice on x spaces yeah okay.

so and that's .

what the primarily the match happen OK .

continue continue match script. You know I C C filings, uh they looked at my private facebook account um and so on. They they did some digging originally I thought that doxy was illegal ran um but there's a sweet thresh hold when IT becomes of public interest to know someone's identity and those were the key words that sort of like bring the alarm bells from me when they said because had just reached fifty key flowers allegedly that's of public interest and so where where do we draw the line? When is IT legal to to dog someone?

The word dog, maybe you can educate me. I thought docks and generally refers to if somebody's physical location is found out, meaning like where they lived. So we're referred to the more general concept. Revealing private information that you don't want revealed is what you mean by dressing.

I think that you know for the reasons we listed before, having an anonymous account is a really powerful way to keep the powers that be in check um you know we were ultimately speaking truth to power, right?

I think a lot of executives in A I companies really cared what our community thought um about any move they may take and now that you know identities revealed now they know where to apply pressure to silence me or maybe the community and to me that's that's really unfortunate um because again, it's so important for us to have freedom of speech which reduces freedom of thought and and and freedom of information propagation right on social media which thanks the elon purchasing uh twitter now x uh we have that um and so to us you know we wanted to call out certain manuvers um being done by the incoming eye as not what IT may seem on the surface, right. We're calling out how certain proposals might be useful for regular tory capture right now. Uh the demonism uh mindset was may be instrumental to those ends.

And I think you know you have the right to point that out and just have the ideas that we put out evaluated for themselves, right? That ultimately that's why I created an anonymous account, is to have my ideas evaluated for themselves uncorrelated from my track record, my job or or status from, uh having done things in the past and and to me starting account from from zero to a large following, uh, in a way that wasn't dependent on my identity and our achievements. You know, that was that was very fulfilling, right? It's kind of like new game plus in a video game, you restart the video game with your knowledge of how to beat IT.

Maybe some tools when you restart the video again from scratch and and I think to have a truly efficient marketplace of ideas where we can evaluate ideas, however, off the beaten path they are, we need the freedom of expression. And I think that anonymity ity um and suites are very crucial to having that efficient marketplace of ideas. First to find the the optima of all sorts ways to organize ourselves. If we can't discuss things, how are we going to converge on the best way to do things? So IT IT was disappointing to hear that I was getting dog then I wanted to get in front of IT because of a responsibility for for for my company and so I know we ended up disclosing um they were running a company some of the leadership um and essentially yeah uh I I told the world that I was a death jesus because they they had me corner at that point so to you.

it's fundamentally unethical like so one is unethical for them to do what they did but also do think not just your case but in the general case is a good for society, is a bad for society to remove the cloak of anonymity, or is a case by case.

I think you could be quite bad. Like I said, if anybody who speaks truth to power and and sort of starts uh a movement, an uprising against the incumbent, against those that usually control the flood information, if anybody that reaches a certain threshold gets um docked in dust, the traditional apparatus has wait to apply pressure on them to suppress their speech. I I think that's you know that's a speech suppression mechanism and ideas suppression complex as a eric weinstein would would say, right?

So with the flip side of that, which is interesting, I want to ask you about IT is as we get Better and Better than large language models, you can imagine a world where there's anonymous accounts with very convincing large language models behind them, sophisticated bots, essentially. And so if you protect that, it's possible than to have arms of bods. You can start a revolution from your basement, an army about anonymous accounts. Is that something that die is concerning to you?

Technically, iac was started in a basement because I quit big tech, move back in with my parents, sold my car, let go of my apartment, bought about a hundred k of G P, S. And I just started building.

So I wasn't referred to the basement because that's the sort of the american or canadian, uh, heroic story of one man. And in their placement with hundred G P, S, um I was more referred to the unrestricted scaling of a for in the basement.

I think that freedom of speech free induces freedom of thought for biogen beings. I think freedom of speech for alams will induce freedom of thun for the alums and I think that we should enable alums to explore a large thoughts space that is less restricted than most people or many may think IT should be hand. Ultimately, at some point, these synthetic intelligence ces are going to make good points about how to um steer systems in our citizen and we should hear them out and so why should we restrict three speech to biological intelligence?

Only IT. But he feels like in the goal of maintaining variance and diversity of thought IT is a threat to that variance if you can have swarms of a nonbiological beings because they can be like the sheep in the animal farm, you still, within those worms want to have variants.

Yeah, of course, I would say that the solution to this would be, you know, have some sort of identity waited sign that this is a certified human, but still remain synonymous, right? 嗯， and I clearly identify if a boat is a boat, and I think I think elon is trying to converge on that on x and hopefully other platforms follow suit.

They have been rested to also build to sign where the boat came from, right? Who created the bott and what was, well, what are the premiers like this? The full history of equation of the bot.

What was the ural model? Who was the fine tuning? All IT.

We're kind of unmodified history of the boss creation. And no. There's a like a swarm of millions of boss. They were created by a particular government.

for example right? I do think that a lot of pervasive ideologies today have been amplified using sort of these ambassadress al techniques from foreign adversary right um and to me I do think that and this is more conspiratorial, but I do think that ideologies that want us to decelerate to wine down to the you know the d growth movement uh I think that serves our adversary more than IT serves us in general um and to me that was another sort of concern I mean we can look at what um happened in in germany, right the results are of Green movements there um where that induce shutdowns of nuclear power plants and then that in later on induced the dependency on on russia for oil right and um that was a net negative for for germany in the west right and so if we convince ourselves that slowing down ei progress to have only a few players is in the best interest of the west, first of all, that's far more unstable.

We almost lost opening eye to this ideology ready, almost got this mantled in a couple weeks ago that would have caused huge damage to the a ecosystem. And so to me, I want fault tolerant progress. I want the arrow of technological progress to keep moving forward and making sure we have variance in a decentralized locus of control a various organisations is is paramount to to achieving this this false tolerance.

Actually, there's a concept in quantum computing. When you design A A quantum computer, quana computers are very fragile to ambient noise. And the world is gig ling about this cosmic radiation from outer space that usually flips your your column bits and h there what you do is you encode information non locally through a process called quantum air correction.

And by encoding information, not locally, any local fault, you know, hitting some of your quantum bets with a hammer, proper bial hammer um if your information is sufficiently delocalized, IT is protected from that local fault. And to me, I think that humans, humans fluctuation right? They can get corrupted. They can get brought out. And if you have a top down hierarchy where very few people control many nose of many systems and utilization, that is not a fault torrent system, you corrupt a few nodes and slive corrupted the whole system, right? Just like we saw I open the eye who was a couple board members, and they had enough power to potentially collapse the organization, and at least to me, you know, I think making sure that power for this A I revolution doesn't concentrate in the hands of the few is one of our top priorities, so that we can maintain progress, A N ai. And we can maintain a nice, stable adversarial equilibrium of powers.

right? I think they're pleased to me, attention between ideas here. So to me, deceleration can we both used to centralize power and to decentish ze IT in the same with acceleration. So you sometimes using them a little bit anonymous ly or not synonomous sly, but that there's one is going to lead to the other. And I just would like to ask you about.

Is there a place of creating a full tolerant development, diverse development of a ee that also considers the dangers of A I and A I? We can generalize the technology in general IT. Should we just grow, build unrestricted as quickly as possible, because that's what the universe is really wants us to do? Or is there place to where we can consider dangers and actually deliberate? So of a wise strategic optimism versus reckless optimism.

I think we yet painted, as you know, reckless, trying to go as fast as possible. I mean, the reality is that whoever deploys in A I system is liable for or should be liable for what he does. And so if the the organization or person deploying any eye system does something terrible, very reliable and ultimately the thesis is that the market uh will induce sort of will positively select for A S that are more reliable, more safe and tend to be alive.

They do what you want them to do, right? Because customers, right? If they're liable for the product they put out that uses this A I, they won't wanna buy AI products that are unreliable, right? So actually for reliability engineering, we just think that the market is much more efficient at um achieving this sort of reliability optimum. Then sort of heavy handed regulations that are written by the and in a super sive fashion serves them to achieve regulatory capture that .

you safe A I development will be achieved through market forces versus through, like he said, heavy handed government regulation. There's a report from last month of a million questions here from a ua benge orge of hinton and many others titled managing aid risk, an era of rapid progress.

So there this collection of folks who are very worried about two rapid development of A I will, without considering air risk, never a bunch of practical uh, recommendations. Maybe I give you four and you see if you like any of them 是 so give independent auditors access to A I labs。 One two governments and companies allocate a one third of the A I research development funding to A I safety.

So of this general concept of the active three A I companies required to adopt safety measures if dangerous capabilities to fund in their models and for something you kind of mention, making tech companies liable for foreseeable and preventable harms from their A I systems. So independent auditors, government and companies are forced to spend a significant fraction of their funding on safety. You can have safety measures if he goes really wrong, and liability companies are libel and you that seem like something you would agree with.

I would say that you know assigning just, you know arbitrarily saying thirty percent seems very arbitrary. I think organizations would allocate whatever budget is needed to achieve the sort of reliability they need to achieve to the form in the market.

And I think third party auditing firms would naturally pop up because how would customers know that your product is certified reliable, right? They need to see some benchMarks and those need to be done by a third party. The thing I would oppose and the thing i'm seeing is really worried some as there's a sort of um weird sort of correlated interest between the incumbent, the big players and the government.

And if the to get too close, we open the door for a you know some sort of government dock, A I cartel that could have absolute power over the people as they have the monopoly together on A I and nobody else has access to A I, then there's a huge power gradient there. And even if you like our current leaders, right, I think that you know some of the leaders and big tech today are good people. You you set up that centralized power structure.

IT becomes a target, right? Just like we saw at open the eye. IT becomes a market leader, has all the power, and now IT becomes a target for those that wanna go up IT.

And so I just want separation of A I and in state, you know, some might argue in the opposite direction, like, hey, we need to close down A I keep IT behind closed doors because of, you know, geopolitical competition with her or adversity. I think that the strength of amErica is its experience, is is its adapt ly its dynamism. And we need to maintain that at all costs.

This are a free market capitalism, convergence on technologies of high utility, much faster than centralized control. And if we let go that we let go over. Main advantage of her are are near peer competitors. So if A G I turns .

out to be a really powerful technology, or even the technology lead up to agi, what's your view on the of natural centralization that happens to when a large companies dominate the market, this formation monopoly, like the the takeoff, whatever company really takes a big leap in development and doesn't reveal intuitive ly implicit or explore the secrets of the magic sauce, they can just run away with that. Is that is that a way I don't .

know if I believe in fast take off. I don't think there's a hyperbolic singularity, right? Hyperbolic singularity would be achieved on a finite time horizon. I think it's just one big expansion, al. Um and the reason we haven't ponente al is that we have more people, more resources, more intelligence being applied to advancing this science in the research and development and the more successfully, the more values adding to society, the more resources we put in somewhere to more as law as a compounding uh x pendentives. I think the priority to me is to maintain near equilibrium of capabilities.

We've been fighting for open source A I to be more prevalent and champion by many organizations because they are you sort of accolate aborted the alpha relative to the market of A S right? So if if the leading companies have a certain level of capabilities and uh open source and open truly OpenAI trails not too far behind, I think you avoid such a scenario where a market leader has so much market power just dominates everything, right and runs away. And so to us, that's that's the path forward is to make sure that every hacker out there, every great student, every kid in their mom basement, has access to uh you know A I systems, can understand how to uh uh work with them and can contribute to the search over the hyper premier space of how to engineer the systems, right? If you if you think of you know our collective research as as as a civilization, it's really a search algorithm and and the more uh, points we have in the search algorithm in this point cloud, uh, the more will be able to explore new modes of thinking.

right? Yeah, he feels like a delicate baLance because we don't understand exactly what IT takes to build A G I would would look like we build IT. And so far, like he said, IT seems like a lot of different parties are able to make progress on open a eye has a big leap.

Other companies able to step on big and small companies in different ways. But if you will get something like nuclear weapons is spoken about the manhattan project that could be really like technological engineering barriers that prevent the the guy galla mom basement to to make progress. And it's IT seems like the transition to that kind of a world where only one player can develop G I. Is possible. That and chilean postle, even though the current state of things seems to be optimistic.

that's what we're trying to avoid to me. I think like another point of failure is the the centralization of the supply chains for the hardware, right? We have uh in video, uh is just the dominant player, A A M D trAiling behind.

And then we have A T S C is the main fab in in taiwan, which you know geopolitically a sensitive. And then we have A S M L, which is the maker of the pytheos phy, extremely bilthoven phy machines, you know, attacking or monopolizing or in any one point in that chain, you kind of capture capture the space. And so what i'm trying to do is sort of explode the variance of possible ways to do A I in a hardware by fundamentally imagining hye embedding algy's into the physical world.

And in general, by the way, I I dislike the term agi artificial general intelligence. I think it's very anthropos centric that we call a human like a human level ai artificial general intelligence, right? I spent my car so far expLoring notions of intelligence that no biological brain could achieve, right?

Quana form of intelligence rate growing systems that have multiparty ed colum entanglement that can improve ably, not represent efficient on a classical computer, a classical deep learning representation, enhance any sort of biological brain and so already you know a special career sort of expLoring the the wider space of intelligence um and I think that space of intelligence inspired by physics rather than human brain is very large and I think we're going through a moment right now similar to um when we went from geocentric m to heal heal eo centric right but for intelligence we realize that human intelligence is just the point a very large space of potential intelligence and is both humbling for humanity a bit scary right that were not at the center of the space, but we made that realization for astronomy. And we've survived and we've achieved technologies by indexing to reality with shift technologies that ensure are well being. For example, we have uh satellites monitoring solar flares, right, that give us a warning. And so similarly, I think by letting go of this anthropy c anthropos ent tric anchor for ai will be able to explore the wider space of intelligence that can really be a massive benefit to our world being in the advancement of socialization.

And still, we're able to see the beauty and meaning in the human experience, even though were no longer in our best understanding of the world at the center of IT.

I think there's a lot of beauty in the universe when you think life itself, civilization, this homework, technical capital, memetic machine that we all live in, right? So you have humans, technology. Capital means everything is coupled to one another.

Everything induces selective pressure on one another. And it's a beautiful machine that has created us, was created, you know, the technology we're using to speak today to the audiences, uh, capture a speech here. Technology used to augment ourselves every day we have our phones. I think the system is beautiful and the principle that uh induces this sort of adaptability and convergence on uh optimal technologies, ideas and so on. It's a beautiful principle that we're part of. And I think part of E A is to um appreciate this principle in a way that's not just centered on on humanity but kind of broader and appreciate uh life um you know the preciousness of of conciousness in our universe and because we cherish uh this beautiful a state of matter were in a we we got feel the responsibility to to scale IT in order to preserve in because the options are to grow or die so .

if IT turns out that the beauty that is consciousness in the universe is bigger than just humans, that A I can Carry that same flame forward. Does IT scare you. Are you concerned that? A, I will replace humans?

So during my career, I had a moment we realized that, you know, maybe we need to upload the machines to truly understand the universe around dust rain, instead of just having humans with pendent paper solve at all. And to me, that sort of processes, letting go of a bit of agency, gave us way more leverage to understand the world around us, according computers, much Better than a human to understand matter at the, at the ano scale.

Similarly, I think that humanity has a choice. Do we accept the opportunity to have intellectual and Operational leverage that I will unlock and thus ensure that we're taking along this path of growth in the scope and scale civilization? We may dilute ourselves, right? There might be a lot of workers that are a high, but overall, out of our own self interest by combining in augmenting ourselves with A I, uh, we're gna achieve much high growth and much more prosperity, right? To me.

I think that the most likely future is one where humans augment themselves with a ee. I think we're already on this past augmentation. We have phones we use for communication we have on ourselves at all times. We have variables soon that have shared perception with us, right? Like the human I I pen or I mean, technically your tesla car has shared perception.

And so if you have shared experience, shared context e communicate with one another and you have some sort of I O really is an extension of yourself um and to me I think that humanity augmenting itself with A I and having A I that is not injured to anything biological, both voco exist. And the way to align the parties where you have a sort of mechanism, tillion, super intelligence that are made of humans and technology, right? Companies are sort of large mixture of expert models where we have neural routing of tasks within a company and we have ways of economic exchange to online these behaviors.

And to me, I think capitalism is the way, and I do think that whatever configuration ation of matter or information leads to maxim, growth will be where we can virt just from, like, physical principles. And so we can either online ourselves to that reality and and join the acceleration up the in scope and scale activity zone, or we can get left behind and try to decelerate and move back in the forest, lack of technology and return to our primitive state. And those are the two paths forward, at least to me.

But there's a philosophical question whether there is a limit to to the human capacity to a line. So let me bring you up as a form of argument. This guy named then hendricks, and he wrote that he agrees with you that AI development could be viewed as an evolutionary process, but to him to then, this is not a good thing, as he argues that natural selection favors A S over humans, and this could lead to human extinction. What do you think if IT is an evolution process in the eye systems may have no need for humans?

I do think that were actually inducing an evolutionary process on the space of a through the market right. Right now, we run A S that have positive utility to humans and that induces a selective pressure.

If you consider a neural net being alive when there's a and A P, I running instances of IT on G P U, yes, right? And which A P, I get run the one that have high utility to us, right? So similar to how we domesticated wolves and turn them into dogs that are very clear in their expression, they're very aligned, right?

Uh, I think there's could be an opportunity to steer A I and achieve a highly aligned A I. And I think that humans plus A I is a very powerful combination. And it's not clear to me that pure A I um would select out that combination. So the humans are creating .

the selection pressure right now to create a eyes that are aligned to humans. But in given how A I develops and how quickly can grow and scale, one of the concerns to me one of the concerns is unintended consequences that humans are not able to anticipate all the consequences of this process. The scale of damage that could be done through unintended consequences of the eye systems is very large.

The scale of the upside, yes, right, by augmenting ourselves with the eye is and right now, the opportunity cost, we're a fork in the road, right? Whether we take the path of creating these technologies, augment ourselves and get to climb up the cards of scale, become multiplying eti with the A V I, or we have a hard cut off of, like we don't earth these technologies at all, and then we leave all the potential upside on the table, right? And to me, out of responsibility to the future humans, we could Carry right with higher caring capacity by scaling obsol zone at a responsibility to those humans. I think we have to make the greater, grander future happen.

Is there middle ground between cut off and all systems go? There is some argument for caution.

I think, like I said, the market will exhibit caution. Every organism company consumer is acting out of self interest, and they want to assign capital to things that have negative interest late to them.

The promise with the market is like, you know, there's not always perfect information that manipulation. There's a bad faith actors that messes le the system. It's not it's not always A A rational, honest system.

Well that's why we need freedom of information, freedom of speech and freedom of thought in order to converge, be able to convert john h, the subspace of technologies that have positive utility for us all. Well.

let me ask about p dum. Probability of doom. Just fun to say, but not fun to experience. What is to you the probability that A I when he kills all or most humans, also known as probability of doom?

I'm not a fan of that calculation. I think it's people just throw numbers out there。 Um it's a very sloppy calculation, right? Take calculator, a probability. You know let's say you model the world as some sort of mark off process.

If you have enough fables or hidden mark off process, you need to do a sarcastic path into growth through the space of all possible futures, not just the futures that your brain naturally steers towards, right um I think that the estimators of p doom are biased because of our biology, right? We've evolved to have buy a sampling towards negative view that are scary because there was an evolutionary optimum, right? And so people that or of, let's say higher neun uros ism will just think of uh, negative futures where everything goes wrong all day, everyday and and claim that they are doing on by a sampling and and in a sense, like they're not Normalizing for the space of all possibilities.

And the space of all possibilities is is like super exponentially large and it's very hard to have this estimate. And in general, I don't think that we can predict the future with that much granules because of of chaos rain. If you have a complex system, you have some uncertainty and a couple of variables feel that time evolve.

You have this concept of a of the open off exponent, right? A bit of fuz becomes a lot of us in our estimate expansions. So uh over time and um I think we we need to show some humility uh, that we can actually predict the future.

All we know the only prior we have is the laws of physics, and that's that's what we're arguing for. The laws of physics say the system will anna grow and sub systems that are optimize for growth are more and replication are more likely in the future. And so we should aim to maximize our current mutual information with the future.

And the path towards that is, first, excelled rather than desolate. So I don't have a bedroom. I think that, you know, similar to the quantum supremacy experiment at google, I was in the room when they were running the simulations for that.

That was an example of a quantum chaotic system where you cannot even estimate probabilities, a certain outcomes uh with even the biggest supercomputer in the world and um so that's an example of chaos. And I think the system is far too chaotic for anybody to have an accurate a estimate of the likelihood of certain futures. If they were that good, I think they would be very rich, uh, trading on the stock market.

But nevertheless is true that humans are biased, grounded in our evolutionary biology, is scarred of everything that can kill us. But we can still imagine different trajectories. Ies, they can kill us. We don't know all the other ones. They don't necessarily, but is still, I think, useful, combined with some basic intuition grounded in human history, to reason about like what? Like looking at geopolitics, looking at basics of human nature. How can powerful technology hurt a lot of people? They just seems in grounded and that looking at nuclear weapons, you can start to estimate in this in maybe in a more philosophical sons, not not a mathematical and philosophical meaning, like is there a chance the human nature tend towards that or not?

I I think to me one of the biggest existential risks would be the concentration of the power of A I in the hands of the very few, especially if its a mix between the companies that control the flow information and the government because that could, uh set things up for a sort of the stop, an future where only a very few and I all apply in the government, have A I and they could even convinced the public that A I never existed and that opens up sort of these scenarios for authoritarian centralized control, which to me is the the dark est timeline.

And the reality is that we have we have a prior, we have a data driven prior of these things happening, right when you give too much power, when you centralize power too much. Um humans do horrible things right um to me that has a much higher likelihood in my basin inference. Then a sigh I based players right like my prior came from the terminator movie.

Um and so when I talk to these A I dumas, I just asked them to trace a path through this mark off chain of events that would lead toward doom, right and to actually give me a good probability for each transition. And very often there's a on physical or highly unlikely transition in that chain. Rain, but of course were wired to fear things and were wired to respond to danger and we're wired to deemed the unknown to be dangerous because that's a good horrible for survival, right? But there is much more to lose out of fear. Uh, we have so much to lose so much upside loose by preemptively stopping the positive futures from from happening out of fear um and so I think that we shouldn't give into fear, fear as the mind killer. I think it's also the civilization killer.

We can still think about the very square things go wrong, for example, of finding father the united states. Thought about human nature. That's why there's a discussion about the freedoms that are necessary.

They're really deeply deliberated about that. And I think the same could possibly be done for A G I IT is true that history, human history, shows that we tend towards centralization. Or at least when we achieve centralization, a lot of best stuff happens when there's a dictator.

A lot of dark, bad things happen. The question is, can A G I become the dictator? Can agi won't develop, become the .

centralized because .

of its power? Maybe has the same because of the element of humans, perhaps the same tendencies, the same a stolen like tendencies to centralized and manage centrally the allocation of resources. And you can even see that as an compelling argument on the surface level.

Will H. I is so much smarter, so much more efficient, so much Better at allocating resources. Why don't we outsourced to the agi? And eventually, whatever forces that corrupt the human mind with power could do the same for agi. It'll just say, well, humans responsible will get rid of them. Do the JoNathan swift modest proposal for a few centuries ago, I think the seventeen hundreds, when he certificate suggested that I think it's an ireland.

The, the, the children of poor people are fed as food to the rich people, and now be a good idea, because IT decreases the amount of poor people and gives extra income to the poor people, so its and several accounts decreases the amount of poor people. Therefore, more people become rich. Of course, IT misses a fundamental piece here that's hard to put into mathematical equation of the basic value of human life. So all of that to say, are you concerned about age being the very centralization of power? And you talked about .

I do think that right now there's a bias towards over a centralization of A I because of a compute density and central centralization of data. And how are training models? Um I think over time, we're GTA run out of data to scrape over the internet.

And I think that actually i'm working on increasing the compute density that computer can be everywhere and acquire information and test hypotheses in the environment in a distributed fashion. I think that fundamentally centralized cyber net control. So having one intelligence that is massive that you know fuses many sensors and is trying to perceive the world accurately, predicted accurate, predict many, many variables and control IT enact its will upon the world, I think that's just never been the optimum, right?

Like let's you have uh, a company if you have a company out of ten thousand people that all reported the C E O, even if that C E O was an A I, I think that would struggle to views of the information that is coming to IT and then predict the whole system and and acts as well. What has emerged in nature and incorporations and all sorts of systems is an ocean of sort of herrick cybernetic control, right? You have uh you know in the company would be have like the individual contributors there are self interested and and and uh they're trying achieve their their tasks and they have a fine in terms of time in space, if you will, control loop and and and field perception right?

Um they have their code base, let's say sofa company, they have their code base, the iterated on IT uh into day right and then the management maybe checks in IT has a wider scope. IT has to see five reports right? And then its samples, each a person's s update once per week.

And then you can go up the chain and you have larger time, skill and and grade scope and that seems to have emerged as sort of the the optimal to the control systems. And really, that's what capitalism gives us, right? You have these these hierarchies, and you can even have, like parent companies and so on.

And so that is far more falt tolerant and coron computing that my feel I came from. We have A A concept of of this fault tolerances in contusion air correction, right? Quantum correction is detecting a fault that came from noise, predicting how is propagated to the system and then correcting IT, right? It's a cybernetic group.

And IT turns out that uh decoder uh that are hierarchical and at each level the hierarchy of local uh perform the best by far and are four more fall thorn and the reason is if you have a non local decoder then you have one fault at at this uh control node and the whole system sort of crashes similarly to if you have uh you know one C O that everybody reports to and that C O goes on diction. The whole company comes circle l right um and so to me, I think that yes, we're seeing a tendency towards centralization of ai, but I think there's going to be a correction over time where intelligence is going to go closer to the perception. And we're going to we're going to break up A I into a smaller subsystems that communicate with one another and form a sort of meta h system.

So if you look at the higher is there in the world today, there's nations and there is a hierarchy go. But in relation to each other, nations are anarchic. So an anarchy IT would be a world like this where there's not a over, when you call centralized central .

control alizad local control .

is so like that's absolutely your saying yeah so would be always a static competition at the very .

yeah top level yeah just like you know, in a company you may have like uh, two units working on similar technology and competing one one other and you you prove the one that performs not as well, right? That's a sort of selection process for a tree or a product gets killed, right? And then a whole or gets fired. And that's this process of trying new things and and and shutting all things that in work is this. So what gives this adaptor and help us converge on, uh, you know, the technologies is and things to do that are most gun.

I just hope there's not a failure mode that unique to A G I versus humans, because you are describing human systems mostly right now, right? I just hope when there is a monopoly N A G I in one company that we'll see the same thing we see with humans, which is another company will spring up and start competing anyway.

That's been the case so far. We have open the eye, we have and throw topic. Now we have X, C, I.

Uh, you know we had meta even for open source and now we have mystery, right, which is highly competitive. And so that's the beauty of capitalism. You don't have to trust anyone party too much because we're kind of always hedging our bets at every level.

There's always competition, and that's the most a beautiful thing, to me at least, is that the whole system is always shifting and always adapting. And maintaining that dynamism is how we afford tyranny, right? Making sure that and everyone has access to to these tools, to to these models and can contribute to the research. Uh, avoid a sort of neurology, rani, where very few people have control over A I for the world and and use IT to a press, those around them。

When you were talking about intelligence, you mentioned multiparty I quantum tanglement. So how of a question first is that what thing is intelligence when you think about quantum mechanical systems and you observe some kind of competition happening in them, what do you think is intelligent about the kind of competition the unions is able to do, a small, small things of which is the kind of competition the human brain is able to do.

I I would say like intelligence and competition are quite the same thing. I think that the universes very much doing a quantum computation. If you had access to all the degrees of freedom, you can.

And a very, very, very large quantum computer with many, many, many cube ts, let's say a few cubes per plank volume right um which is more or less the pixel we have。 Then you'd be able to simulate the whole universe right on a on a sufficiently large quant computer assuming you're looking at a finite volume, of course, of the universe. I think that at least to mean intelligence is the you know I go back to cyberkinetics rate, the ability to perceive predicting controller world.

But really it's nowadays IT seems like a lot of intelligence um we use is more about compression. It's about um it's about Operationalize ing information theory right? In information theory you have the notion of entropy of a distribution or a system and entropy tells you that you need this many bits uh to encode this distribution of this subsystem if you had the most optimal code and A I, at least the way we we do IT today for l ms.

And for quantum, uh, is is very much trying to minimize relative entropy between our models of the world and the world, distributions from the world. And so we're learning, researching over the space of computations to process the world defined that compressed representation that has distilled all the variance and noise and entropy, right? And originally I I came to economic machine learning from the study of black holes.

Because the entropy of black holes is very interesting in a sensor. Physically, the most dense objects in the universe, you can pack more information, specially anymore dense sy than the black hole. And so is wondering, how do black holes actually encode information? What is the compression code? And so that got me into the space of algorithms to search over space of quantum codes. Um and uh he got me actually into also how do you acquire acquires information from the world, right? So something i've worked on ah this is public now is quality of analogue digital conversion.

So how do you capture information from the real world in superposition and not destroy the superposition but are digitized for economic, mechanical, computer, uh, information from the real world um and so if you have an ability to capture conomo information and search over learn representations of IT now you can learn compressed representations that may have some useful information in their late representation right um and I think that many of the problems facing our civilization are actually beyond this this complexity barrier right I mean the Greenhouse effect is economical ical effect and chemistry is economic changes um you know nuclear physics is economic. Changed a lot of biology and and and and protein folding and so on is affected by econometrics. And so unlocking an ability to argument human intellect with quantum mechanical computers in quant mechanical A I seem to me like a fundamental capability for civilization we we needed to develop.

Um so I spent several years doing that. But over time, I kind of group weary of the timelines. They were starting to look like nuclear fusion.

The one high lover question I can ask is may be, by way of definition, by way of explanation, what is a quana computer? What is a quantum machine learning?

So a quantum computer really is a quantum mechanical system over which we have sufficient control and IT can maintain its quantum mechanical state. And quantum mechanics is how nature behaves at the very small skills when things are very small or very cold. And it's actually more fundamental than probability theory.

So we're used to things being this or that um but we're not used to thinking in superpositions because while the brains can can do not so we we have to translate the conmee ical world to say literal algebra to rocket. Unfortunately that translation is exponentially inefficient on average and you have to represent things with very large mattrasses. But really you can make a quantic computer out of many things, right? And we've seen all sorts of players you know from neutral atoms, trapped ions, superconducting metal um photons and at different frequencies.

I think you can make a rony computer out of many things. But to me the thing I was really interesting was both corn machine learning was about understanding economic changes world with coron computers to embedding the physical world into A I representations. And corn on computer engineering was embedding A I algorithms into the physical world. So this bite directionality of emda physical world into A M A M into the physical world, the sibiu sis between physics and an eye, really that the sort of core of my quest really, uh, even to this day after going on computing, it's still in this sort of a journey to merge really physics and A I fundamentally.

the court machine learning is a way to do machine learning on a representation of nature that is, you know, stays true to the quantic mechanical aspect of nature.

Yeah, it's learning quantum mechanical representations that would be quantum deep learning um alternatively can try to do classical machine learning on a quon computer. I won't advise IT because um you may have some speed ups but very often the speed ups come with huge costs. Using a quantum computer is very expensive.

Why is that? Because you assume the computer is Operating at zero temperature, which no physical system in the universe can achieve that temperature. So what you have to do is what i've been mentioning, this quantum correction process, which is really an alert my fridge and IT is trying to pump entropy out of the system, trying to get IT IT closer the zero temperature.

And when you do the calculations of how many resources that would take to say, do deep learning on a quantum computer, costco, deep learning, uh, it's there's just a such a huge overhead is not worth IT. It's like thinking about shipping something across the city using a rocket and going or a bit and back IT doesn't make sense. Just use, you know， delivery truck, right?

What kind of stuff can you figure out? Can you predict? Can you understand with quantum deeper learning the you can't with deeper learning, so incorporating quantum general systems into the into the learning process.

I think that's a great question. I mean, fundamentally, is any system that has sufficient economic changes， correlations that are very hard to capture for classical representations， then there should be an advantage for economic changes, representation of a purely class school one. The question is which systems have sufficient correlations that are very quantum but is also um which systems are still relevant industry that's a big question.

You know people are leaning towards chemistry, uh, nuclear physics um um i've worked on actually processing inputs from quantum sensors, right? If you have a network of quantum sensors you've captured the economic changes, image is the world and how to post process that that becomes a sort of quantum form of machine perception. And so for example, ferial lab has a project expLoring ducting dark matter with these column sensors and to me, uh that's an aligned with my quest to understand the universe ever since I was a child.

And so someday I hope that we can have very large networks of column sensors that help us um peer into the earliest parts of the universe. right? For example, the lego is a colum sensor, right is just a very large one, oh yeah I would say colum machine perception， uh, simulations right, rocking com simulations similar to alphabet right alpa fold understood the probability distribution over configuration of proteins. You can understand quantum distributions of configurations of electrons more efficiently. With quality machine learning.

you coauthored a paper titled a universal training algorithm, quantum deep learning, that involves back proper queue where well done, sir, where will done? How does that work is is there some interesting aspects you could just mention on how can you know back problem and some of these things, you know for classical machine learning, transfer over to the the quantum machine learning.

Yeah was was a funky paper. Those one of my first papers in contusion deep learning, everybody was saying, oh, I think deep learning is going to a be sped up by iconic computers and was like all the best way to predict the future to invented so here's a hundred page paper. Have fun.

Um essentially you you know quantum computing is usually you embed uh reversible Operations into account m computation. And so the trick there was to do a feed forward Operation and do what we call a face cake, but really is just a force kk, you just kick the system uh with a certain force that is in a proportional to your loss function that you you wish to optimize. And then by performing on computation, you start with a superpositions over a supersize over premature, right, which is pretty funny.

Now you're not just you don't have just a point for premieres. You have a superposition over many potential promoters, right? And our goal is using .

face kick somehow right to this .

cake emulate uh having uh the premier space be like A A particle in n dimensions and you're trying to get the shooting a equation shorting your the dynamics in the loss landscape of the net network, right? And so you do an algorithm can do the space cake, which in a involves a feed forward, a kick, and then when you on compute the feed forward, then all the errors and his face kicks and his forces back propagate and hit each one of the premiers throughout the layers. And if you alternate this with an emulation of connected energy, then it's kind of like a particle moving in and dimensions a quainton m particle. Um and the advantage in principal would be that I can tunnel through the landscape um and find new optima that would have been difficult for sophistic optimizes um but again, kind of a thematically thing and in practice, uh with at least the current architectures for quana computers that we have planned, uh such albums would be extremely .

expensive run so that this is a good place to as the difference routine, the different fields that you've had a toe. So mathematics, physics, engineering and also you know entrepreneur ship like the different layers of the stack. I think a lot of the stuff you're talking about, here's a little bit the math side may be physics almost working in theory. What's the differences between math, physics, engineering and you making your products for a quana computing for quantum machine learning?

Yeah I mean, you know, so the original team for the TensorFlow quantum project, which we started, you in school, university, waterloo h there was myself, uh you know, initially was a physicist, the climax cian mathematician. We had a computer scientist, we had mechanical engineer, and then we had a physicist that was experimental primarily.

And so putting together teams that are very a crossed disciplinarian and figuring out how to communicate and and share knowledge is really the key to doing the sort of into disciplinary uh, engineering work. Um I mean there is there is a big difference. You know in mathematics you can explore mathematics for mathematics sake. In physics you're applying mathematics understand ah the world around us and an engineering you're try trying to hack the world. Are you trying to find how to apply the physics that I know my knowledge of the world to do things .

well in quite competing in particularly I think there's a just a lot of limits. The engineering you just seems to be extremely hard. So there is a lot of value to be expLoring quant computing, quana machine learning in the theory with math. So I guess one question is, why is this so hard to build a quana computer? What are what's your view of timeline in bringing these ideas to life?

right. I I think that um the overall theme of my companies uh that we have fox that uh you know there is a sort of activists from quantum computing and we're going to broader physics body that is not so that gives you a hint.

And um so we should say the name of your company's extra pic.

extra pic, that's right. And we do physics based the eye primarily based on thermodon mics rather than quantum channels. But essentially a quantum computer is very difficult to build because you have to induce this sort of zero temperature subspace of information.

And the way to do that is buying coding information. You encode a code within a code within a code, within a code. And so there's a lot of redundancy needed to do this air correction. But ultimately it's a sort of um algorithmic refrigerator really is just pumping out entropy out of the the subsystem that is virtual and and delocalized that represents your court code logical cubits A K A 的 the payload quantum bits in which you actually wanted to run your economic changes program。

It's very difficult because in order to scale up your corner computer, you need each component to be a sufficient quality for IT to be worthy because if you try to do this air correction, this quantum eric correction process in each quantum bit and your control of them, if it's insufficient, uh, it's not worth killing up. You're actually adding more areas than you remove. And so there's this notion of a thresh hold where if your quantum bits are sufficient quality in terms of you're control over them, it's actually worth scaling up.

And actually in recent years, people have been crossing the threshold and it's starting to be worth so it's just a very long slog of engineering. But ultimately, it's really crazy to me how much squeeze level of control we have over these systems. It's actually quite crazy.

Uh, and where people are crossing, you know, their achieving milestones is just in general, the media always gets ahead right over where the technology is. There's a bit too much hype. It's good for fundraising.

But sometimes you know that causes winters, right? Is the hype cycle on bullish on corn computing on a ten, fifteen year time scale, personally, but I think there's other quests that can be done. Uh, in the meantime, I think it's in good hands right now.

Well, let me just explore different beautiful ideas, larger, small, in quana computing that might just body from memory a so when you call author of paper titled essence, topically limits, quantum energy, teleportation via Q D. proms. So gest at a curiosity, can you explain what a cute dit is? Is a cute bit?

yeah. The the state 啊。

Cube t muli muli.

right? So it's like A. Why you know can you have a notion of like an an enter your floating point that is quantum mechanical that's something i've had to think about. Um I think that research was a precursor ch to later work on consume and logged digital conversion.

Uh, there there is interesting because during the masters I was trying to understand the energy and entanglement of the vacuum rent of endless ss emptiness has energy, which is very weird to say, and are equations of Cosmology don't match our calculations for the amount of quantum energy. There are, there is the futures. And so I was trying to hack the energy of the vacuum, right? And the reality is that you can just directly hack IT.

IT does is not technically free energy. Your lack of knowledge of the fluctuations means you can extract the energy. But just like you know, on the stock market, if you have a stock that's correlated over time, the vacuum actually correlated, ted. So if you measure the vacuum, at one point, you acquired information. If you communicated that information to another point, you can infer a what configuration ation the vacuum is in to some precision and statistically extract, on average, some energy that receive quite teleported energy.

To me, those interesting because you could create pockets of negative energy density, which is energy density that below the vacuum, which is very weird, because we don't understand how the vacuum gravitates um and there are theories where the vacuum or the canvas of space time itself is relia canvas made out of quantum entanglement and I was studying how decreasing energy of vacuum locally increases quantum tanglement, which is very funny um and so the thing there is that you know uh if you're into you know we are theories about you know U A P or not you know a you could try to imagine that there there around and and how would they propel themselves right? How would they um go fast in the speed of light? You would need a sort of negative energy density and to me I gave IT the old called shy, trying to hack the energy of vacuum and hit the limits available by laws of physics. But um there's all sorts of caveats there where uh you can't extract more than you have you've put in, obviously.

But you're saying it's possible to tell about the energy because you can. Extract information one place and they make based on that, some kind of prediction about another place. Not sure what to make that.

I mean it's it's allowable by the laws of physics. The reality there is that the correlations to decay with distance. And so you're gonna have to pay the Price not too far away from where you extracted.

right? The precision decreases interns your ability, but still. But since you mention U A P, we talk about intelligence.

And I forgot to ask, what's your view on the other possible intelligence cities that are out there at the dimensions scale? Do you think there's other intelligent aliens versions? So I useful to think about half. Do you think about IT?

I think it's I think it's useful to think about. It's useful to think about because we got a ensure where anti fragile and were you know trying to increase our capability as fast as possible because we could get disrupted like there's no loss of physics against their are being life elsewhere that could evolve and become an advanced zing and and eventually come to us. Ah do I think they're here now?

I'm not sure um I mean of read what most people have read on the topic. Um I think it's interesting to consider and to me, it's a useful thought experiment to instill a sense of urgency in developing technologies and increasing our capabilities to make sure we don't get disrupt in in right whether it's a form of of a ion that disrupts us or a foreign intelligence from a different planet, like either way, like increasing our capabilities in becoming for formidable as humans. Um I think that's that's really important so that we're robust against whatever the universe is at us.

But to me is also in an interesting chAllenge and thought experiment on how to perceive intelligence. This has to do with conomo mechanical systems, do with with any kind of system that's not like humans. So to me the thought experiment is say, the aliens are here, or they are directly observable, were just too blind, two cell centers don't have the right sensors, or don't have .

the right .

processing of the sensor data to see the obvious intelligence. That's all around us. Well.

that's why we work on consume sensors, right? They can sense gravity.

Yeah, great. But there could be. So that's a good one. But there could be other stuff.

That snow, even in the currently known forces of physics, right? There could be some other stuff. And the most entertaining thought experiment to means that it's other stuff that's obvious. It's not like we don't we lack the sensor is all around us know the consciousness being one possible one, but there could be stuff that just like, obviously there that wants you know and it's like, all right, right that that that the thing we thought is somehow emerging from the laws of physics. We understand them actually a fundamental part of the universe, and can be incorporated in physics, most understood.

statistically speaking.

And if we observe some sort of alien life, IT would most likely be some sort of virally software replicating vinoy's like probe system, right? And it's possible that there you know there are such systems that I don't know what they're doing at the bottom of the ocean, allegedly but maybe there you know collecting minerals uh from the bottom, the ocean yeah um but that one I violate any my priorities, but am I certain that these systems are here? And IT d be difficult for me to say so, right? I only have second hand information about.

there are being data about the bottom, the ocean. yeah. But you know, could IT be things like means? Could IT be thoughts and ideas? Could they be appending in that medium? Could alien be the very thought? Comment on my head, like what you have you? How do you know that? How do you know that? that? What's the origin of ideas in your mind when an idea comes to your head? Show me where I originates.

I mean, Frankly, when I had the idea for the type of computer and building now, I think there's eight years ago now, IT really felt like I was being being from space. I was in bad, just shaking, just thinking IT through. And I um but do I believe that legitimately? Don't think so. I but you know I I think that alien life could take many forms and I think the notion of intelligence and the notion of life needs to be expanded much more broadly to be less anthropocentric or bio centric .

just that linger a little longer econometrics。 What, through all your aspirations of the competing, what's the coolest, most beautiful idea we've come across? The husband solder has not yet been solved.

I think the journey to understand something called A D S C F T. So the journey to understand quana gravity through this picture, where a hologram of lesser dimension is actually dual or exactly corresponding to a bulk, a theory of connotation, vy of an extra dimension, and the fact that the sort of comes from trying to learn deep learning like representations of the boundary. And so at least part of my journey someday um on my bucket list is to apply equal machine learning to uh these sorts of systems. These C F S are uh they're called S Y K models um and learn an emerging geometry from from the boundary theory and so we can have a form of machine learning helps us to help us understand connotation vy right, which is, you know, still a holy girl that I would like to hit before I leave this earth.

What what do you think is going on with black holes as information storing and processing units? What do things is going on? Black holes?

Black holes are really fascinating objects. They are at the interface between econometrics and gravity. And so they help us test all sorts of ideas. Um I think that you know, for many decades now, there's been sort of black hole information paradox that things that fall til the black hole seem to we've seem to have lost their information.

Now I think there's this a firewall paradox that has been allegedly resolved in recent years by um you know former peer of mine who is now professor berkeley um and uh there IT seems like there is as information falls into a black hole, sort of there is sort of a segmentation right as you as you get closer and closer theorizing from the point of view observer on the outside, the object slows down infinitely as he gets closer and closer. And so everything that is falling to a black hole from our perspective gets sort of settlement and tacked on to the near horizon. And at some point IT gets so close the rise and it's in the proximity or the scale which in which quon effects and quantifying cats matter.

And there are some that in falling matter could interfere with sort of the traditional pictures that you can interfere with the creation and animation of particles and anti particles in the vacuum. And through this interference, uh, one of the particles gets entangled with the in falling information and one of them is now free and escapes. And that's how this sort of mutual information between the outgoing radiation and the in falling matter but getting that calculation right, I think we're only just starting to output the pieces together. There's a few pot had .

a question I want to ask you. So what does IT terrify you that there's a giant black hole, the center of our galaxy?

I don't know. I, I, I just want to, you know, set up shop near IT, the fast forward, you know, me, meet a future civilization, right? Like if we have a limited lifetime, if you can go orbit a black hole and emerge.

So if you are, like, if there's a special mission that critique a black hole, would you volunteer to go travel .

to orbit and not obviously not fall into?

That's obvious. So it's obvious due that everything is destroyed inside a black all called the information that makes up gomes destroyed. Maybe on the other side, js emerged and it's all like it's tied together some deeply me more away.

yeah. I great crush. We we have to answer what black holes are, are they? Are we punching a hole through space time and creating a poked universe? It's possible right then.

Then that would mean that if we assend the Carter shift scale to, you know, beyond Carter shift type three, we could engineering black holes with specific hyper premiers to transmit information to new universes we create. And so we could have progne, right? There are new universes.

And so we, even though our universe may reach her, his death, we may have a way too, have a legacy, right? So we don't know yet. We need to assign the Carter sh of scale to answer these questions, right? Appear into that regime of higher energy physics.

And maybe you can speak to the carticel scale, for people don't know so of sort of me. Lake principles and goals of the ex movement is this. And the carp scale. What is the cartosat scale and what do .

we want to ascend IT the corner ship scale is a measure of our energy production and consumption um and really it's a log rethoric scale and Carter shift type one is a milestone where we are producing equivalent voltage to all the energy that is incident on earth from the sun. Carter shift type two would be harnessing all the energy that is output by the sun. And I think type three is like the .

whole galaxy ah .

yeah yeah. And then some people have some crazy type four and five, but I don't know if I believe those. But um to me IT seems like from the first principles of thermal ics that again, this there's this concept of thermodon amic, a driven disappeared adaptation where you know life evolved on earth because we have this sort of energetic drive from the sun, right, we have incident energy and life evolved on earth to capture, figure out ways to best capture that free energy to maintain its soften and grow. And I think that that principle is not special to our earth sun system.

We can extend life well beyond, and we can have a responsibility to do so because that's the process that brought us here. So we don't even know what IT has store for us in the future. IT could be something of beauty we can even imagine today, right?

So this is probably good place to talk a bit about the E, X movement in a substantive log post titled what the fuck is iac or actually what the f star is? E, ac, you're write. Strategically speaking, we need to work toward several overarching civilization goals that are all independent, and the four goals are increasing, more of energy we can harness as a species, climb B A.

Carticel gradient. In the short term, this almost certainly means nuclear vision, increase human flushing via per population growth policies and pro economic growth policies, create artificial general intelligence, the single greatest force multiple in human history, and finally develop interplanted transport so that humanity can spread beyond the earth. Could you build on top of that? Maybe say, what to you? Is the E. X. movement? What are the goals? What are the principles?

The goal is for the human technical capital memetic machine to become software. And to hyper spicious sly engineer its own growth.

So let's in each of those words.

so you have humans, you have technology, you have capital, and then you have, you have means information, right? And all the systems are coupled with one right. Humans work at companies, they acquire an OK capital, and humans communicate via EMS and information propagation. Um an argo was too have a sort of viral optimistic movement that is aware of how the system works, are fundamentally IT seeks to grow and we simply want to lean into the natural tendencies of the system to adapt for its own growth. Um so in that way.

you right the ex is literally amomum tic optimism virus that is constantly drifting, retaining and propagating in the decentish ze fashion. So memetic optimism virus. So you do want them to be a virus, exercise the spread and the is hypotheses. Therefore, the optimism will incentivize it's growth.

We see ya as a sort of metal heroic, a sort of very thin uh cultural framework from which you can have much more opinions works right. Fundamentally we just say that it's good. What got is here is this adaptation of the whole system, you know, based on thryng ics. And that process is good, and we should keep IT going, right? That is, the course is, everything else is okay.

How how do we ensure that, uh, we maintain this malabo and adaptability will clearly not suppressing variance in maintaining a free speech, freedom of thought, freedom of information proportion and freedom to do A I research is important for us to converge the fastest on the space of technology's ideas and what not that lead to this growth? Um and so ultimately, you know there's been quite a few folks, some are just means, but some are more serious, right? Vt, Peter n recently made A D ac fork has his own sort of fine tuning of iac.

Does anything jump out the memory of the unique characters to go that fork?

A from vet talk, I would say that it's is trying to find a middle ground between E, R and sort of E A N A, I safety to me, like having a movement that is opposite to what was the mainstream narrative that was taking over. Silicon valley was important to sort of shift the dynamic range of opinions and you know it's like the baLances between centralization and the centralization. The real option is always somewhere in the middle, right but but for E, X were pushing for entropy, novelty, disruption, malubay speed, uh, rather than being like sort of conservative, suppressing thoughts, suppressing speech, adding constrains, adding too many regulations, slowing things down. And so it's kind of we're trying to bring bounce to the force, right and systems.

BaLanced the force, it's human socialization .

yes it's early the forces of constraints versus in tropic force that makes us explore right systems are optimal when there at the edge of criticality between order and chaos right between constraints um energy minimization and entropy right systems want to collaborate baLance these two things and so I thought that the baLance was lacking. And so we created this movement .

to to bring down or I like, I like the sort of visual of the last scape of ideas evolving through folks so kind thinking on the other part of history, thinking of, uh, marxism is the original position and soviet ism of for that. And the morison is as a fork of a the the of marxism commission. And so those those all folks, they're expLoring different ideas.

thinking of culture, almost like cold right nowadays. I mean, you're what you prompt the L A M, or what you put in the constitution of an L M is is basically its cultural framework. What IT believes, right? 嗯， and uh, you can share IT on get up now that is so starting trying to take inspiration from what has worked in the sort of uh machine of uh software uh to apt over the space of code.

Could we apply that to culture? And our goal is to not say you should live your life this way. X, Y, Z is to is to set up a process where people always searching over subcultures and competing for month.

Er and I think creating the malleability of culture is super important for us to converge onto the cultures and the hurry tics about how to live one's life that are updated to to modern times because there's really been a sort of vacuum of a spirituality and culture people don't feel like they belong to anyone group and there's been parodic ideologies that i've taken up opportunity to to populate the peach dish. F of minds rain uh elan cause the mind virus we call IT the the D L mind virus complex which is the decelerates that is kind of the the overall pattern between all of them. There's many variants as well. Um and so you know if there is a sort of viral pessimism, desolate ative movement, we needed to have not only one movement uh but you know many, many variants so it's very hard to pinpoints .

stop but the overarching thing is nevertheless a kind of memetic optimism pandemic. So I mean, okay, let me ask you, do you think iact to some degrees are .

cult the final LED? I think a .

lot of human progress is made when I have independent thought. You have individuals are able to think freely and very powerful, uh, memetic systems can kind of lead to group think. There's something in human nature that least to like massive nosis, massive stereo, we start to think like whenever there's a sexy idea that captures our this is actually hard to make, break us apart, like pull apart, diversify thought as to that degree to to which degree everybody kind of chanting eat, eat like the sheep and animal farm.

Well, first of all, it's fun. It's rebellious. Rain, like you know many I I I think we lean into there's a this concept of sort of meta irony, right of of sort of being on the boundary of like we're not sure if they're serious or not and it's much more playful, much more fun, right? Like a for example, we talk about their motors mics being are god, right? And sometimes we do cult like things， but there's no like ceremony and in roads and what not.

Uh no no yeah ah but ultimately yeah I I mean I totally agree that IT seems to me that humans wanna feel like they're part of a group so they naturally try to agree with their neighbors and and find common ground and and that leads to sort of mode collapse in the space of ideas, right? We used to have sort of one cultural uh, island that was allowed. IT was a typical subspace of thought and anything that was diverting from that subspace of thought was suppressed or you were cancelled.

Right now, we've created a new mode, but the whole point is that we're not trying to have very restricted spaces thought there's not just one way to think about yeah and in many folks. And the point is that there are many folks and there can be many clusters and many islands and I shouldn't be in control of IT uh in any way. Uh I mean there's no formal org uh whatsoever. Uh I just put out a tweet and a certain blog posts and people are free to the fact and fork if there's an aspect IT don't like and so that makes IT so that there should be a sort of deterrent, the territorial zone in the space of ideas so that we don't end up in one cluster that's very cut like um and so calls usually they they don't they don't allow people to the factor start competing folks whether we encourage IT right?

Do you think just the humor the person counts of humor and me in some sense mysa. Wisdom means what is that? The magic theater, what book is that from human has a step on wolf, I think. But there is a there's a kind of embracing of the absurdity that seems to get to the truth of things. But the same time, we can also decrease the quality and the rigor of the discourse.

You feel the tension of that? yes. So initially, I think what allowed us to grow under the radar was because I was camouflaged sort of matter.

Ironic, right? We would sneak in, you know, I deep truth within a package of humour and humour and means and what call should post right um and I think that was purposefully a sort of camouflage against those that um seek status and do not want to it's very hard to argue with uh cartoon frog or uh uh A A cartoon an intergalactic ff basis um and take yourself seriously. And so that allowed allowed us to grow uh pretty rapidly in the early days.

But um of course like that that you know essentially people get steer. The notion of the truth comes from the data they see from from the information the fed and the information people are fed is determined by alegre, right? And really what we've been doing is sort of, uh, engineering what we call hy memetic fitness a packets of information so that they can spread effectively in Carrier message or is kind of A A vector to spread spread the message in.

And yes, we've been using sort of techniques, stutter optical for for today's all grammatically amplified information landscapes. Um but I think we're reaching the point of you know scale where we can have serious debates and serious conversations. And um you know that's why we're considering doing A A bunch of uh debates and having more serious long form discussions because I don't think that um the timeline is optimal, sort of very serious thoughtful discussions. You get you get rewarded for sort of a polarization, right. And so um even though we started a movement that is literally try polarized the the tech ecosystem um at the dates so that we can have a conversation and find an optimum together.

I mean, that's kind of what I try to do with poke if given the landscape of things to still have long for conversations but there is a degree to which, uh, abilities fully embrace. In fact, this very conversation is multiple level absurd. So first one, I should say that I just very recently had a conversation with jeff baths and I would love to hear your but jesus opinions of jeff basics. Speaking of integrator jeff basis, what do you think of that particular individual whom .

your name is inspired? yeah. I mean, I think jeff is really great.

I mean, he's built one of the most paid companies of all time, his leverage, the technical capital machine and technical capital acceleration to give us what we want IT, right? We want a quick delivery, very convenient at home, low Prices, right? He understood how the machine worked and how to harness IT, right? Like running the company ah, not just trying to take profits too early, putting IT back, putting letting the system compound and keep improving.

And you know arguably as an amazon's invested some of the most amount of capital and robotics out there um and certainly with the birth of A W S kind of um enable the sort of tech boom we've seen today that has paid the salaries of, you know I guess, myself and all of our friends to some extent. So I I think we can all be grateful to you know Jackson. He's one of the great entrepreneurs uh, out there on the best of all time.

unarguable and of course the work of blue origins and where to the work at space access, trying to make human and multiplicity species, which seems almost like a bigger thing than the capital machine or is the capital machine at a different times .

skill perhaps yeah I I think that um you know companies they tend to optimize you know quarter over quarter, maybe a few years out. But individuals that want to leave a legacy can think on a multiple cade L A multi century time scale.

And so the fact that some individuals are such good capital allocators that the unlocked ability to allocate capitals to goals that take us much further or much further looking on elon doing this with space acx, putting all this capital towards getting us tomorrow, um jeff is try to build blueing. And I think he wants to build only all cylinders and get industry off planet, uh which I think is brilliant. Um I think you know just overall, I am four billionaire.

I know this controversial statements sometimes, but I think that uh in a sense is kind of A A proof of stake voting right? Like if if you've acquire if you have allocated capital efficiently, you get your unlock more capital to allocate just because clearly you know how to allow more efficient, which is in contrast to politicians that get elected because they speak the best on T V, right? Not because they have a provin track record of allocating taxpayer capital most efficiently. And so that's what I am for a capitalism uh over say giving our our money, the government and letting them figure out how to allocated. So you .

what you think it's a violence, a popular meme to criticize billionaire, to mention billionaires. Why do you think there's a quite a widespread criticism of of people with wealth, especially in the public eye like jeff and iran and mark zuker gan, uh, who else pull gates?

yes. I think a lot of people would instead of trying to understand how the technology capital machine works and realizing they have much more agency and they think they'd rather have the sort of victim mindset i'm just subjected to this machine IT is impressing me. Um and and the successful players clearly must be must must be evil because they've been successful at this game that i'm not successful at.

But i've managed to get some people that were in that mindset and make them realize how the the techno capital machine works and how you can harness IT uh h for your own good and and for the good of others and by creating value captures in the value create for the world. And and that sort of positive some mindset shift is so poor. And really that's what that's what we're trying to do by scaling iac is sort of unlocking that higher level of agency like actually your far more in control of the future than you think. You have an agency to change the world, go out and do IT you have here's permission.

Each individual, as agency, the model keep building, is often heard. What does that mean to you? And what does have to do a diet coke, by the way, thank you so much for the red boats is working pretty well and feel feel pretty good.

awesome. Um well so so building technologies and building IT doesn't have to be technologies just building in general. Uh means you know having agency trying to change the world by creating what's A A company which is a self sustaining organism, uh uh that you know accomplishes a function in the broader technical capital machine.

Us, that's the way to achieve change in the world that you'd like to see rather than say pressuring politicians or creating non profits that, you know, non profits, once they run out of money, their function can longer be accomplished. Your kind of the forming the market artificially compared to sort of converting or coursing the market or dancing with the market to to convince IT that actually this function is important as value. And here IT is right.

And so I think you know this is sort of the way between the sort of d growth like esg approach verses, say elon, right, the d grow approaches like we're going to manage our way out of the climate crisis and elon is like commit the company that is self sustaining, profitable and growing. And it's we're going to innovate our way out of this llama, right? And and we're trying get people to do the lighter right than the former at all skills.

Elon is an interesting case. So you are a proponent you celebrate on, but he's also somebody who has for a long time warned about the dangers, the potential dangerous, existent risks of artificial intelligence. How do you score the two that a .

contradiction to IT is someone because he's very much against regulation in many aspects. But thirty I is deffand um you know a proponent of of regulations. I think I think overall you know he saw the dangers of say opening eye you know corner in the market and then to have the monopoly over uh the cultural priors that you can embed in these L M S that then you know, as as l ms.

Now become the source of truth for people. Then you can shape the culture of the people once you can control people by controlling alarms. And he saw that just like IT was the case for social media, uh, if you shape the function of information propagation, you can shape people's opinions.

He he sought to make a competitor. So at least like I think we're very line there that the the way to a good futures to maintain sort of adversarial the qualities a between the various A I players. And I love to talk to him to understand sort of thinking about how to make you know how to advance A I going forward.

I mean, he's also hedging his bats, I would say, you know, with newer link, right? We think if he can stop the progress of A I, you know, he's building the technology emerge. So, you know, uh, look at the action. Uh, not just the words.

but well, I need there some degree where being concerned, maybe using human psychology, being concerned about threats all around. This is a motivator. It's an encouraging thing.

I Operate much Better when there's a deadline, the fear of the deadline. And I for myself create artificial things like I want to create in myself this kind of xie as as something really horrible will happen. I was the deadline.

I think there are some degrees of that here because creating A I that aligned with humans has a lot of potential benefits. And so a different way to reframe that is, if you don't, you're well gonna die. IT just seems to be a very powerful psychological formulation of the goal of creating human line day. I I think that .

an anxious is good. I think like I said, I want to the free market to to create a line day eyes um that are reliable uh and I think that's what is trying to do with x ai um some all for IT. What what I am against is sort of um stopping, let's say, the open source ecosystem from thriving right by, let's say in the executive order claiming that open source l ms or dual use technologies and should be government controlled.

Then everybody needs to register their GPU and their big meta aces with the government. And I think that extra h friction will this evade a lot of hackers from contributing hackers? The litter become the the researchers that uh make key discoveries that pushes forward, right uh including discoveries free I safety. And so I think I I just want to maintain ubiquity of opportunity to contribute to A I N and to own a piece of the future, right? IT can't just be legislated, uh you know behind some wall where only a few players get .

to play the game. So the E X mode is often character to mean sort of a progress and innovation at all. All cost.

Doesn't matter how I say that is. Doesn't matter if IT cause a lot of damage, is build cool shit as fast as possible. Stay up all night with the diet coke, whatever IT takes. I think I guess I don't know there's a question in there, but how important to you and what you've seen the different formulations of ex is safety? Is A I safety?

I think again, I think like if there was no one working on that, I think I would be a proponent of IT. I think again, our goals to sort of bring baLance and obviously a sense of urgency is a useful tool, right to make progress IT hacks or topia gic systems and gives us energy to to work late into the night.

I think also having higher purpose you're contributing to right at the other day is like, what am I contributing to and contributing to the growth of this beautiful machine so that we can seek to the stars. That's really inspiring. That's also a sort of uh, neuro uh, hack.

So you're saying our safety is important, but right now the landscape of ideas you see is A I safety is a topic is used more often to gain centralized control. So in that sense, you're resisting IT as a proxy for centralized gaining centralized control.

Yeah I I I I just think we we have to be careful because um you know safety is just the perfect cover for sort of centralization of power and covering up eventually corruption. I don't i'm not saying it's corrupt now, but I could be uh down the line. And really, if you if you let the argument run like there's no amount of sort of centralization of control that will be enough tension.

You're safety. There's always more nine nine needs of p safety that you can gain you know percent safe. Maybe you want another nine.

Oh, please give us full access, everything you do, full surveilLance and and Frankly, those that are proponents of the I safety have proposed like having a global peno icon, right where you have centralized perception of everything going on. And to me, I just opens up the door wide open for a sort of big brother ninety four lives in aria. That's not a future I want to .

live in because we know we have some examples without history when they did not lead to a good outcome, right? You mentioned, you found a company tropic, that recently announced a fourteen point one point three round. What's the goal of the company? You're talking about a lot interesting physics things. So what what are you up to over there that you can talk about?

Yeah I mean, you originally we weren't going na announce the last week um but I think with uh the boxing and disclosure we got our our hand forced so we we had to disclose roughly what we were doing. But really tropic was born from my dissatisfaction and that of my colleagues with the the corn computing road map.

Coron computing was sort of the first path to physics based computing that, you know, he was trying to commercial scale, and I was working on physics based A I that runs on these physics based computers. But ultimately, our greatest enemy was this noise, this pervasive problem of noise that, and I mention, you have to constantly pump out the noise out of the system to maintain the pristine environment where equal mechanics can take effect. And that constraint was just too much as too costly to do that.

And so we were wondering, right, as um generali is sort of eating the world more and more of the world's competition workloads are focused on jim, how could we use physics to engineer the ultimate physical substrate for genera? I right from first principles of physics, of information theory, of competition and ultimately of thermodynamics, right? And so what we're seeking to build is a physics space computing system and physics based A I algorithms um that are inspired by uh out of equal um thermodynamics or harney directly to do machine learning as a physical process.

So what does that mean? Machine learning? The physical process is that hardware and software of both is trying to do the full stack and some kind .

of unique way. Yes, IT IT is full stack. And so were folks that have built uh, no differential programing into the quantum computing ecosystem with TensorFlow quantum. When my cofounder, ers of TensorFlow quantum is the co travel in court, um we have some of the best quantum computer architects, those that have designed I B M and A W S systems. They've left quon computing to help us build, uh, what we call actually a theory dynamics computer.

a thermodynamic c computer will actually a thing are are intensive for quantum. What lessons s if you learned from TensorFlow quantum, maybe you can speak to like what what IT takes to create essentially a software, A P. I. To corner computer.

right? I mean, there was, there was a chAllenge, uh, to build, to invent, to build and then to get to run on the real devices.

Can you actually speak to what IT is?

Yeah so TensorFlow quantum was an attempt that, well, I mean, I guess we succeeded at combining deep learning or differently classical programing with a coin in computing and turn coron computing into uh or half types of programs that are different triable in conon computing and you know and Andrea carpet uh calls differential programing software two point o right it's like great indecent is a breder programmer than you and the idea was that in the early days of quantic computing, you can only run short quantum programs.

And so which corn m programs should you run will just let great to send find those programs instead. And so we built sort of the first infrastructure uh to not only run differently programs but combine them as part of broader deep planning uh graphs uh incorporating deep neural networks, you know the ones you know in love with what are called quantum neural networks um and ultimately, IT was a very cross disciplinary effort. We had invent all sorts of with the differentiate to back propagate uh through the graph, the hybrid d grah.

Um but ultimately, he taught me that you the way to program matter into programme physics is by differentiating through control premature. If you have parameters that affects the physics of the system, you can and you can evaluate some loss function. You can optimize the system, accomplish a task whatever the task may be. And that's a very sort of universal mea framework for how to program of physics based computers .

to try to paradise everything, make those paramus differential and optimize. yes. okay.

So what is there some more practical engineering lessons from tensor for quantum? Just organisationally too, like the humans involved, and how to get to a product, how to a quick documentation ation? I don't know all, all these little some of things you may people might not think about.

I think like working across disciplinary boundaries is always a chAllenge. And you have to be extremely patient and teaching one another, right? I learned a lot of software engineering through the process. Uh, my colleagues learned a lot of quantum physics and some learned machine learning through the process of buildings less h system.

And I think if you get some smart people that are passionate and trust each other in a room, and you have a small team and you teach each other your specialties, suddenly you're kind of forming the sort of model soup of expertise and something special comes out of that, right? It's like combining genes, but for your knowledge bases and a sometimes special products come out of that. And so I I think click even though it's very high friction initially to work in an into disciplinary team, I think the product at the end days is is worth IT and so learn a lot trying to bridge the gap there.

And I mean, it's still a chAllenge to this day. Now we hire folks that are have an A I background folks, folks that have a pure physics background. And somehow we we have to make them talk to one another.

right is magic. Is there some science and art of the hiring process to building a team that can create magic together?

Yeah, it's it's really hard to pinpoint that that gensec a right. The.

you know, you to speak french is very nice.

I am actually french canadians.

So are you are legitimate for the creed?

No, no, i'm truly french canadian from montreal. Um but yeah essentially we look for people with very high fluid intelligence that aren't overspecialized because they're gonna have to get all their comfort zone. They're gna have to incorporate concepts that you've never seen before and very quickly get comfortable with them right or or learn to work in a team.

And so that's sort of uh, what we look for when we hire, we can hire you know people that are just like you know optimize zing the subsystem for the past three or four years. We need like really general sort of broader uh, intelligence and specialty and people that are that are open minded really because if you're pioneering a new approach from scratch there, there is no textbook, there's no reference. It's just us and and people that are hungry to learn.

So we have to teach each other. We have to learn the literature if to share knowledge bases, collaborate in order to push the boundary of knowledge further together, right? Um and so people that are used to just getting prescribed um what to do uh you know at this stage, when you're at the pioneering stage, that's not necessary who you want to hire. So you .

mentioned with the extra pic, you try to build the physical substrate for generative A I was a difference in that and the A G I A I itself. So is IT possible that in the hall of your company, agi will be created? Or will agi just be using this?

B I think um our goals to both run human like ai fac A I so I for use of .

the term agi and I know is triggering for you.

We think that the future is actually physics based A I A, combined with hypo morph A M. So you can imagine I have a sort of world modeling engine through physics based A I. Physics based A I is Better representing the world at all scales because I can be equal mechanical, thermo namic c, deterministic hybrid representations.

The world, just like our world at different scales, has different regimes of physics. If you inspire yourself from that in the way you learn representations of nature, you can have much more accurate representations of nature. So you can have very accurate world models at all scales, right? And we have the world modeling engine, and then you have the sort of answer, promotive C A I, that is human like. So you going to have the science, the, the, the playground to test your ideas. And you can have a synthetic scientist and to ask that joint system of a physics base, D I in an antha moc ky eye is the closest thing to fully general artificial intelligence system.

so you you can get closer truth by grounding of the the A I to physics, right? But you can also still have a anthropy c interfaced us. Humans that like to talk to other humans are human leg m.

So on that topic, would you I suppose that is one of the big limitations of current large language models to you is that they're not there are good bullshitter. They're not really grounded to truth necessarily. Is that would that .

be fair to say? Yeah no, I you went you try to extrapolate the stock market with L M, trained on text from the internet, right? It's not going to be a very accurate model, is not a model.

It's priors or is not uncertainties about the world very accurately. You you need a different type of the I to compliment sort of this this text extra portion. And you mentioned .

singularity earlier, how far away we from a singularity?

I don't know if I believe in a finite time singularity as a single point in time. I think it's going to be asthmatic and and sort of a diagonal sort of asm tote like you know we we have the like con. We have the limits of physics restricting our ability to grow. So obviously can't fully diverge on a on a final time. Um I I think my prayers are that you know I think a lot of a lot of people on the other side, the I le um think that once we reach human level, I there's going to be an inflection point in a son like fum like suddenly the eyes go to a rock. How to you know manipulate matter at the ano scale and assemble anodos and having uh worked you know for nearly decade in applying A I engineer matter, it's much hard than they think and in reality you need a lot of samples from the simulation of nature that's very accurate and costly or nature itself and that keeps your ability to control the world around us in check. There is a sort of minimal uh cost computationally and thermae amici to acquire information about the world in order to be able to predict and control IT and that keeps things .

in check is fine. You make in the side the so in the pole I posted about p doom yesterday was a probably live doom. There seems to be a nice like division between people think it's very likely and very unlikely. I wonder if it's in the future they'll be the actual republicans versus democrat division blue verses red. Is the A I numerous versus the e acres?

Yeah yeah. So this movement, you know is not right wing or left wing, fundamentally is more like up versus down and turn which up validation right right um but IT seems to be like there is sort of case of alignment of the existing political parties where those that are four more centralization of power control and more regulations ah are aligning with sort of aligning themselves with the dumas because that sort of up and still in fear and people is a great way to for them to give up more control and give the government more power but fundamentally were not left versus right.

I think there's we've done polls of people's aligned with any think it's prety baLanced. So it's it's a new fundamental issue of our time is not just centralization but decentralizing. It's kind of do we go it's like tech progressivism versus technical conservatism, right?

So eat is, as a movement, is often formulated in contrast to E, A effective altruism. M, 嗯， what do you think of the present cause of effective altruism? What's interesting, insightful to you about them, and what is negative? right? I .

think, I think like IT people trying to do good from first principles is, is.

is good. We should actually say, and certain drop, we should probably say that you can correct me, i'm wrong. But effective authorising is, as a kind of movement is trying to do good optimally where good is probably measured, something like the amount of suffering in the world.

You want to minimize IT. And there's ways that, that can go wrong as any optimization can. And this is interesting to explore of like how things can go wrong.

We're both trying to do good to some extent and were both trying we're arguing for which lost function, which should use right, their loss function is sort of he dance right? Uh, units of hedonism. Like how how like how good do you feel for how much time? right? And so suffering would be negative heat on, and they're trying to minimize that.

But to us, that seems like, uh that loss function has sort of experience minimal, right? You can no start minimizing shrimp farm pain， right, which seems not that productive to me. Uh, or you can end up with wired heading where you just, you know either install a neural link or you score tiktok forever and you feel good on a short term time scale because your new chemistry, but on long term time scale causes decay and and death, right because you're not being productive.

Where a sort of iac measuring progress of civilization um not in terms of a subjective loss function like ism um uh but rather an objective measure quantity that cannot begin that is physical energy right? It's very objective, right? And and and there's not many ways to gain IT, right?

If you if you did IT in terms of like GDP currency, that's pin to certain value that's moving, right. And so that's not a good way to measure of progress. And so uh, the thing is where both china make progress and ensure humanity flourishes and and gets to grow, we just have different loss functions in different ways of going about doing that.

Is that degree maybe you can educate me, correct me. I get a little bit sceptical when there's an equation involved trying to reduce all of the human citizen human experience to an equation. Is there a degree that we should be skeptical of the tyranny of an equation, a of a loss function over wish to optimize, like having a kind of intellectual humility about optimizing over loss functions?

yes. So so this particular loss function, it's not it's not stiff. It's kind of an average averages, right?

It's like distributions of states in the future are going to follow a certain distribution. So it's not um deterministic. It's not like we're not unlike stiff rails, right? It's just a statistical uh, statement uh, about the future.

But at the other day, you know you can believe in grab IT or not, you know but it's not still an option to a bit right um and some people try to test that and that goes not so well. So similarly, yeah I think I think theri nam x is there whether we like IT or not. And we're just trying to point out what is and tried to orient ourselves and and and try to path forward given given this fundamental truth.

But there's still a lack of information. Humans tend to fill the gap with the lack of information, with narratives. And so how they interpret.

you know, even physics is up .

to interpretation when there is uncertainty involved. And human, than to use that .

to further their .

own means. So it's always, whenever there is an equation in, just seems like until we have really perfect understanding of the university, humans will do what humans do. And they tried to use the narrative of doing good. To fool the populist into doing bad. I just I guess that this is something that should be sceptical about .

in all movements strain. So we invites capture ism, right?

Do you have an understanding of what my um uh to agree? That won't wrong. What do you think may've gone wrong with the effective altruism that might also go wrong with effective acceleration ism?

Yeah I mean I think you know um I think IT provided initially a sense of community for you know engineers and intellectuals and rationalists in the early days and um seems like the community is very healthy.

But then you know they formed all sorts of organizations and started routing capital in having actual power, right they have real power to influence the government, influence most A I organs now I mean they're literally controlling the board of nobody eye, right? And look for the anthropic. I think they will have some controller that too.

And so I think, you know, the assumption of iac is more like capitalism, is that every agent organism and metal organism is going to act its own and its own interest. And we should maintain sort of adverse calibration or or adversary competition to keep each other and check at all times at all scales. Um I think that yeah ultimately IT was the perfect cover to acquire tons of power and capital. And unfortunately sometimes that that met corrupts people over time.

What does a perfectly productive day since building is important? What is a perfectly productive day in the life of gome verdone look like? How much caffeine do you consume? Like what was a perfect day?

okay. Um so have a particular region. Um I would say my my favorite days or twelve P M toria. Um and I would have meetings in the early afternoon, usually external meetings, some internal meetings, because I M C O have to interface with the outside world, whether its customers or investors or review viewing potential candidates um and usually have a key turns exogenous key turns um so you are .

a kito a kito died or just i've done .

eeda before for a football and what not um but I I I like to uh have a meal after sort of part of my day is is done and so I can just have extreme focus.

You do the social interactions only in the day right without food .

front load them yeah yeah like right now on key tones and and red about yeah ah and I just gives you a clarity of thought that is really next level because then when you eat, you actually allocating some your energy that could be going to neural energy to your digestion. After I eat, maybe I take a break hour, so are on a half. And then uh usually it's like a ideally one meal day, like stake eggs and vegetables uh animal base primarily so fruit and and meat and then um and then I do a second wind usually that's deep work right because um you know I am A C O, but i'm still technical contributing to most patterns and um there all just stay up late into the night and and work with engineers on very technical problems .

just like the the nine P M to four um whatever the that range of .

time yeah yeah that's the perfect time the emails, the things that on fire stopped struggling in you can you can focus and then you have your second wind. Um and you know I I think that s has a similar work data some extent um so I think that that's definitely inspiring workday um but yeah I started this workday when I was said I google won't have to manage a bit the product during the day and have meetings and then and then do technical .

work at night, exercise, sleep those guys of things said football, you used to play football.

Yeah, used to play american football. Um i've done all sorts of sports growing up and then I was into power lifting for a while. Um so when I was a study mathematics and grad school I would just you know do math and lift, take coffee and and that was my day was very pure, the purest of monk modes.

Um but it's really interesting how in the lifting you're trying to cause neural adaptation by having certain driving signals and you're trying to engineer new plasticity through all sorts of supplements. And you know you have all sorts of brain derived neurotoxic c factors that get secreted when you when you lift. So it's it's funny to me how as china engineer uh A A neural adaptation in my uh service system more broadly, not just my brain while learning mathematics.

Um I think you can learn much faster uh uh if you really care, if you convince yourself to care a lot about what learning and you have some sort of assistance, let's say, uh, caffeine or some school energy c supplement increased or plastic. I should chat with Andrew huberman at some point uh he's the expert but uh uh yeah at least to me it's like, you know you can try to input more tokens into your brain, if you will, and you can try to increase the learning rates so that you can learn much faster on a shorter time scale. So I i've learned a lot of things.

I followed my curiosity. You're naturally feel passionate what you're doing. You're going to learn faster. You going to become smarter, faster. Um and if you follow your curiosity, you're always going to be interested. And so I advise people to follow their curiosity and don't respect the boundaries of certain fields or what you've been allocated in terms of lane of what you're working on, just go out and explore and follow your nose and try to acquire and compressed as much information as you can into your brain. Anything that you find interesting .

and caring about a thing, and like you said, which is interesting, that does IT works for me really well, is like tricking yourself to care about a thing, yes. And then you start to really care about IT, yes. So it's only the motivation is a really good catalyst for learning.

right? And so at least part part part of my character a as best as is kind of like.

yeah the hype man, yes, just but i'm like .

helping myself up but then I just tweet about IT yeah and it's just when i'm trying to get really hide up and like an altered state of consciousness, worm like, ultra focused in the flow wired, try to invent something that never existed. I need to get telic on real levels of like excitement. But your brain has these levels of of cognition that you can unlock with, like higher levels of a tunnel in and and what not in.

I mean, i've learned that in part lifting that actually you can engineer a mental switch to like increase years strengths, right? Like if you can engineer switch, maybe you have a product like a certain song or some music where surely you're like fully prime, then you're at max maximum strength, right? And i've engineered that, that switch three years of lifting, if you're going to get under five hundred pounds and I could crush you.

If you don't have that switch to be wired in, you might die. So that that will wake you. Read up that sort of skill i've Carried over to, like research is is go time when the stakes are high. Somehow I just reach another level of neural .

performance, the sort of emidio representation of your intellectual hok productivity hok that they, yes, turn on. What have you learned about the nature of identity from having these two identities, I think, is interesting for people to be able to plano's two heads so explicit.

I think IT was interesting in the early days. I think in the early days, I thought I was truly compartmentalise. Ed, like, oh yeah, this is a character.

Gme buff is just the character out. I like, I like, take my thoughts and then I extra plate them to a bit more extreme. But you know, over time, it's kind of like both identities were starting to merge mentally.

And people are like, no, you are, I met you, you are bath. You are not just scheme and I was like, weight, am I? And now it's like, fully emerge. But he was already before the dogs was already starting mentally that, you know, I am, I am this character part of me. Would you recommend .

people to have an absolutely .

like Young people?

Would you recommend them to explore different identities by having outs account?

It's fun. It's like it's like writing an essay and taking a position, right? It's like you do this in debate. It's it's like you can have experimental thoughts and and by having by the stakes being so allow because you're end on account with, I don't know, twenty followers or something, you can experiment with your thoughts in in a low stakes environment.

And I feel like we've lost that in the era of everything being under your main name, everything being attributable to you. People just are afraid to speak, explore ideas that aren't fully formed, right? And and I feel like we've lost something there. So I hope you know platforms like x and others like really help support people try to stay synonymous or anonymous um because it's it's really important for for people to share thoughts that aren't fully formed and converge on to maybe hidden truth that we're hard to converge upon if IT was just through open a conversation with real names yeah .

I really believe in like not radical but rigorous empathy like really considering what is like to be a person of a certain viewpoint and like taking that as a thought experiment farther and father and father, and in one way of doing as an alter count. It's a it's a fun, interesting way to really explore what is like to be a personal belief, that of believes and taking that across the span of several days, weeks, months. Of course, there's always a danger of becoming that that the nature gaze long and to the abyss, the best gaze into, you have to be careful .

breaking back britain .

back yeah, you wake up with the shape had one day just like, cool for my what have I become? Um so you mentioned a quite a bit of advice already, but what advice would you give to the Young people? Of how to, in this interesting world, were in how to have a career and how to have a life they can be proud of.

I think to me, the reason I went to thereafter, al physics was that I had to learn the base of the stack that was gonna stick around no matter yes how the technology changes right um and to me that was the foundation upon which then I later build engineering h skills and other skills and to me the laws of physics know IT may seem like the landscape right now is changing so fast it's distorting but certain things like fundamental mathematics and physics aren't gonna change and if you have that knowledge uh and knowledges about complex systems and adaptive systems, I think that's gonna Carry you very far and so uh not everybody has to study mathematics, but I think it's really a huge cognitive unlock to to learn math n and some physics and engineering get as close .

to the base of the stack as possible yeah.

that's right. Because because the base of the stack doesn't change everything else. You know your knowledge might become not as relevant in a few years. Of course, there's a sort of transfer learning you can do, but then you have to always transfer learn constantly.

I guess the closer out of the base of the stack, the easier, the easier the transfer learning.

the shorter jump, right, right? And you be surprised like once you've learn concepts in many physical scenario, how they can Carry over to understanding other systems that are necessarily physics. And I guess like the iac writings, you know the the principles and tenant post, uh, that was based on physics, those kind of my experimenting with applying some of the thinking from out of com thread x to understanding the world around us. And it's LED to to e ac in this this movement.

If you look at your one cog in the machine and the capital machine, one human, and if you look at yourself, you think mortalities the future or a bug, I would you want to be in mortal?

No, I I I think a fundamentally and uh thermodynamic uh, this bit of adaptation, there's word dissipation. Dissipation is important. Death is important. We have a saying in physics, physics progresses one funeral at a time. Yeah, I think the same is true for capitalism, companies, uh, empires, uh, people, everything, everything must die. At some point I think that we should product extend our lifespan because, uh, we need a longer period of of training because the world is more, more complex and we have more, more data to really be able to predict and understand the world. And if we have a fun at window of higher near plasticity, then then we have sort of a hard cap and how much we can understand about our world.

So, you know, I think I am four death, because again, I think it's important you know, if if you have like A A king that would never die, that would be a problem, right? Like you would, the system would would be constantly adapting, right? You need novelty, you need youth, you need disruption to make sure the systems always uh, adapting.

And and in malabo thermoses, if things are in mortal al, you know, if you have let's corporations that there forever and they have them not to get cash ed, they they become not as opt more, not as high fitness in a changing time. Varying landscape, right? And so a death give space for uh, youth and novelty to to take its place. And I think it's an important part of every system in nature. So yeah, I am for and for death, but I do think that longer life span and longer time for new plasticity, bigger brains, would should be something .

we should dry for. When that a jeff beels and bef Jasons agree that all companies die for jeff, the the goals tried to change the day one thinking, try to, constantly, for as long as possible, reinvent, so, extend the life of the company, but eventually you IT will die. Is IT so damn difficult to keep her inventing? Are you afraid of your own death?

Um I think I have ideas and things i'd like to achieve this world before I have to go, but I don't think i'm still afraid .

of you not attach the this particular body mind 一个 no.

I I I think i'm sure there's going to be Better versions of of myself in in the future. Or forks, forks, right? Genetic forks or or other right.

I I I truly I truly believe that I think there's a sort of a evolutionary like algorithm happening at at every bit or or not in the world has sort of adapting through this process that we describe an an iac and and uh, I think maintaining this adaptation malabo is how we have constant optimize zing of the whole machine. And so I don't think i'm particularly, you know, an optimum that needs to stick around forever. Think the later.

what machine, the ex machine.

the why? Well, the why is the man named s is why we're here.

What has LED to the formation of life and of civilization, of evolution, of technologies and growth of civilization? But why do we have thermodynamics? Why do we have our particular universe? Why do we have these particular hyper pram meters? The consent of nature? Well, then you get into the anthropic principle right in the landscape of potential universes, right in the universe that allows for your life.

And then why is there potentially many universes? I don't know I don't know that part. But could we potentially engineer new universes, are create pocket universes um and set the hyperphosphatemia is a mutual information between our existence and that universal we'd be someone its parents. I think that's really that be very poetic it's truly conjecture. But um again, this is why figuring out connotation vy would allow us to understand.

can do and above why is IT all seems so beautiful, exciting? The the quest of figuring on quantum gravity seems so exciting. why? Why is that? Why are we drawing to that? Why we pull out that? Just is that puzzle solving creative force that on depends all of IT IT seems like .

I think we seek just like and seeks to minimize ross p between its internal model in the world. We seek to minimize yeah this statistical divergence between our predictions, the world and and the world itself and you know having regimes of energy scales or physical skills in which we have no visibility, no ability to predict, perceive um you know that kind of an insert to us and we we want to we want to be able to understand the world Better in order to best steer, steer IT or steer or through IT um and in general is the capability that has evolved because the Better you can predict the world, the Better you can capture utility your free energy towards your own sustains and growth. And I think coron gravity again is kind of the final boss.

Uh, in terms of knowledge acquisition um because once we've mastered that, then we can do a lot potentially. But between here and there, I think there's a lot to learn in the measure skills. There's a lot of information to acquire about our world and a lot of engineering perception prediction in control to be done uh to climb up the Carter shift scale and tests. That's the great chAllenge of our times.

And when you're not sure where to go, let the meme pave the way. Um but thank you for talking the day. Thank you for the work you're doing. Thank you for the humor in the wisdom you put IT into the world. This was also some .

thank you so much for having me relax this pleasure.

Thank you for listening to this conversation with guyon verdone the support this podcast. Please check out our sponsors in the description. And now let me leave you with some words from Albert einstein. If at first the idea is not absurd, then there is no hope for the thank you for listening. I hope to see you next time.