Max Sky

Keith

Faris Habib

PARSA

Amanda

Petri Kajander

Norgard

Kumar

Zbynek Drab

Ksenia

T Timothy Tran

Daniel

Nicolas

Naval

Kamal

Such that nothing that happens on one 2d patch is changes the geometry of the next 2d patch over, and no matter how you deform the total 3d quilt, the 2d patches are the same. So locally invariant to global changes and vice versa.. 2/3

3/3 by which i don’t mean to be anal, since you capture the functionally absolutely correct- you can work on the patches and total quilt separately as you wish without foundationally altering laws of physics/dealing with cross talk. Bravo.

Finally this is also why ToE is already functionally here. GR is the quilt, QCD/Standard model are the patches. Main thing about math is to know when far enough is far enough.

 So, I'm curious, was it like this before 2020?

 Was it really, were people seriously, was it in general a trend for people to expect

 to raise significant seeds with like kind of bupkis and powerpoint deck and this is

 our team and we're going to build it and 

 we have like zero product or market feedback

 or experience or clue? Is that a newish phenomena or is that?

 Long time ago, my scientific advisor, who was top of the top of Soviet Union

 theoretical physics superstring theory effort, he liked to repeat to his

 students that there is only one thing that equation likes the most. And the

 most things that equation likes is for somebody to keep looking at them and

 examining them. And it took me years to understand what he actually means that

 equation likes to be examined. So if you take even simplest equation like x

 squared plus y squared, obviously it is extremely simple. But even in such simple

 equation, if you start thinking about it, with sufficient amount of effort

 from first principles, you will start discovering stuff. You will discover

 trigonometry, you will discover trigonometry functions. From functions

 you can actually discover complex numbers. So there is lots of stuff hidden

 even in very simple equations. But of course the equation that we use in

 industry and science these days are nothing but simple. So kind of another

 canonical example would be Einstein general relativity equation.

 And everybody knows canonical story that Gödel, as part of birthday present to

 Einstein, brought to him yet another solution to this equation. And that was

 like, what, 30 years after Einstein published his equation. And Gödel's

 solution to the general relativity equation was like crazy rotating

 universe with closed time loops. So basically time loops that allowed time

 travel. Well admittedly that time travel would take a few billion years, so you

 cannot go into yesterday. You can only go kind of into those huge loops,

 billion years in the past. But nonetheless, just because Gödel looked and analyzed

 and thought about Einstein equations, 30 years later he came up with solutions

 that even Einstein did not suspect existed. So where was that equation, that

 solution, hiding? And of course it was hiding in structure of these

 equations. But here's a key point, that equations are incredibly complex. Every

 equation is effectively, think about it, its own universe, its own reality. When

 you write down equation on paper, this does not mean at all that modern

 scientists, kind of modern papers, that if you write down equation, you

 know 100% everything about it. This is not what it is at all. Think about

 scientific equation, they are more like doorways, they are more like, here is a

 door and what you find behind the door, well it could be anything. It could be

 something as simple as x squared plus y squared, or it could be its own whole

 universe. It could be actually infinite universe. For some equations it would be

 true. So here we go, that message that you're looking at, it basically

 snippet of investor newsletter. And what this investor newsletter is kind of

 about, it's about our usual suspect, Apple, Qualcomm, trying to build a radio

 that goes into every phone. But what's the root of all these problems? Well the

 root is, is of course EM radiation, Maxwell equation. Maxwell equation is

 kind of over 200 years old. But the thing is, Maxwell equation basically kind of

 tensor field equation of electromagnetic field. And I cannot overstate enough, even

 something as simple as classical electrodynamics, it is incredibly

 complicated. We actually, well I would say, we do not 100% completely understand

 even how this equation works. And that's to a degree what makes it so exciting,

 because progress is possible even in Maxwell equations. Even in designing

 analog radios, because the complexity of solution of this propagating

 electromagnetic wave are incredible. And that's why companies keep kind of trying

 to push the limit, keep trying to find out with new solutions, some new

 optimization, because even electromagnetic equations have almost

 kind of unlimited potential, despite, well yeah, it's not relative, it's not

 quantum mechanics, it's just kind of plain old Maxwell stuff. So one big

 lesson to take from all this, remember every complex equation wants only one

 to be examined. And by examining it, you

 kind of discover layers and layers of source code of our reality, but even one

 simple equation can have near-infinite complexity hidden in its solutions.

 Max, I absolutely loved this chit. Thank you very much.

 As backdrop, I've chosen the Cauchy-Riemann equations for Riemann surface,

 which electromagnetic fields or Maxwell's equations absolutely obey,

 and it is so true. I would say, not only the 

 equations, there's something, even with just the numbers,

 the concept of zero and the numbers, frankly, only up to seven,

 kind of eight and nine are optional because they're not prime.

 Those numbers, from those, you can create the whole number line.

 You can, frankly, you can explore the entire body and possibility

 of all of mathematics is included in those numbers.

 And this may be man with a hammer stuff, but in any base system,

 it tends to be true. Like, if you're in base 10, 

 you just need, you know, you need those six numbers.

 And that is suspiciously close to the Nyquist sampling relationship

 where you just have to sample at least, you know,

 at least at twice the highest frequency, and you'll 

 be able to detect any changes. So, anyway, cool stuff.

 To rub in the salt on Apple wounds, the 

 content of that newsletter is effectively that Apple

 spent multiple billions of dollars trying to replicate Qualcomm Radius, absolutely failed.

 Furthermore, the current report from, I think, Bloomberg stating that they are even further

 behind than people estimated. So they are not catching 

 up to Qualcomm, they are actually falling further behind.

 And kind of, yeah, basically it will cost Apple even billions more when they have to

 extend the licensing agreement with Qualcomm in 2026.

 So that's all business. But what this tells me, as a 

 scientist, is that here we go, you have 200-year-old equations

 and Apple, with all the resources, with all the engineering might, but that's engineering

 might in hardware design, is they cannot fully understand these equations.

 It's not enough. Qualcomm, because it has a lifetime 

 of experience working exactly with this AM Radius, they

 understand these equations better. Not completely, as 

 this is by far not a complete knowledge of Maxwell equation.

 But why it's called black art? Because they have lots of 

 these partial solutions, like, hey, what happens if antenna is too

 close? What happens if it's this condition, that condition?

 Their space of solution to Maxwell equation is so huge, so infinite, that Qualcomm is

 basically making tens of billions just understanding some little spot solutions, kind of what's

 happening in specific circumstances, and apparently they're extremely good designing these tiny

 antennas we have in our phones. So that's kind of giving you 

 the scale of solutions, of business opportunities, effectively

 solving one set of equations that takes four lines of text, and that had been invented

 200 years ago. Just that. Tens of billions, probably hundred billions dollars in value.

 Just look at and understand these equations, that's all there is to it.

 And that's Apple and Qualcomm are locked into 

 battle of giants just for these partial solutions.

 And then going a step further, ideally you want to have proof of concept, maybe

 show either sales or some sort of progress made with the product. However, I

 am wondering if people still do that, if you have a prototype and you can get

 investments. I know approaching funding can be tricky because you usually just

 get one opportunity, maybe you get a second one down the road, but your first

 impression should always be the more prepared you can be the better. But I'm

 wondering if that's still a popular option to 

 invest in a startup while they still have the prototype.

 Thank you, Max. This is amazing. I actually needed just help in this matter.

 I've been tinkering for a long time. Not a long time, maybe like about a year.

 But I guess now the next step for me is just proof of concept.

 But this helped a lot as far as reassurance that I'm 

 at least moving in the right way. So thanks for that.

 Alex, a really simple point at the core is that even if you disagree with somebody

 politically and even don't like them personally, no matter how much you

 dislike them politically or personally, if they have been treated kind of so

 unfairly, so kind of absolutely obviously kind of using false pretext for false

 persecution, you still have moral obligation to defend them even if you

 do not, extremely do not like them. And 

 as I learned more things about Sam, kind of

 I realized that I don't disagree with him 80%, I disagree with him probably 99%,

 so it's like every single thing he done kind of in the past, I probably would

 have picked exactly the opposite thing, but it doesn't matter, still have to defend him.

 I see. Yeah, thanks for clarifying. This makes sense from like your own perspective.

 What are some of the points of disagreement, especially 

 the 99% given that, you know, even as you were talking here,

 probably some of the technology that they're enabled have been used to help us.

 Actually, I don't know if AirChat folks use some OpenAI 

 stuff, but I assume at least some of it is used. Maybe not.

 So I will start answering Alex's question 

 about what is the 99% I disagree with Sam Altman.

 But before we start, I would say some of those disagreements are existential.

 I do not claim any moral superiority of my opinion.

 It's an internal class conflict of Silicon Valley, builders versus hustlers,

 because builders will never ever see eye to eye with hustlers,

 and they will always approach how we solve 

 problems, how we build companies in very different ways.

 So I would probably say there is no moral high ground whatsoever.

 It's just anybody who listens, they can make their own conclusions,

 which skill set, which periodic table of actions is more applicable to your personal skill set.

 Or maybe most importantly is what world you 

 want to live in, what trends you want to support.

 Do you want to support a world where hustlers 

 all the shots and builders work at their pleasure?

 Or you want to live in the world where builders call the shot,

 and hustlers just market and promote what the builders had built.

 So first sort of cosmetic difference but still in my view pretty important one is

 than kind of many of us. So kind of all that

 experience we had starting companies from scratch, getting first hundred

 thousand seed investment, working with team of five, team of ten etc. That's

 basically typical story of kind of ground start builder. That is not Sam

 experience. So Sam I think he probably had that like only once in his life.

 Sam experience he's always appointed by some figure of authority to run

 something built by others and so he always come in with already kind of high

 momentum, relatively high funding companies be that kind of Y

 Combinator being assigned by PG or OpenAI assigned by Elon Musk. So that's

 his life story. He's schmoozing in right circles, going to the right dinners

 and getting assigned by this kind of high-level characters to do the bidding.

 It is extremely different life path and mindset that typical zero-to-one Silicon Valley founder.

 Now, if you remember OpenAI early years, remember that the company started not like any usual

 technology startup. It started with an incredible amount 

 of sanctimous hollers and marketing bombast that like, no,

 we are on a mission to save humanity from 

 AI, we are on a mission to give humanity all

 technological builder, that sounds kind of fairly fake.

 It sounds like that reality distortion field cranked out to absolutely unrealistic parameters,

 given what they actually had at the time, and which, remember, well, next to nothing

 because what they needed was invented by Google a couple years later.

 So yet again, it's a kind of philosophical difference, kind of what kind of hustler leadership

 would put up front and what builder leadership kind of would put up front.

 But of course, after that start, I would say things got significantly worse on that hustling

 reality distortion field product axis, and that's where we're going next.

 So now imagine in a kind of abstract theoretical world all you have is a

 text box with amazing autocomplete function and you have image box which

 can mix in artistic works made by other human artists in some interesting

 plausible combinations. That's it. You have basically two autocomplete