As per the usual with your comments, your reading comprehension is terrible. At no point do I defend Elon Musk, I think all of his companies are nuts and what we're seeing is borderline feudalism masquerading as capitalism. I am explaining how it works though. You don't have to like or agree with Musk and his way of doing things to recognize that there are people that do, and the value of his companies are based on those people believing it. Which goes back to the point; if he left, he wouldn't be able to liquidate his stock fast enough to turn it into cash before the SEC halts trading and fines him for insider trading. He can't simply walk away. Without him at the helm, those keeping the stock price up would stop doing so and it would tank, so even the vested portion he has would plummet before he got any hard cash out of it.

I was incorrect about his vested vs. unvested; nevertheless an enormous portion of his wealth is still unvested and tied to unachievable goals so my point stands, because in addition to the many, many things that is abnormal and possibly illegal in the corporate structure of SpaceX, it's abnormal for one person to have a voting power of 10 to 1 for their special class of shares (only a few tech companies did this), and it's even more abnormal to be able to vote on behalf of those shares while they remain unvested. The whole thing stinks to high heaven.

Next time try not using your usual straw man tactic and try reading what people are saying, it's become tiresome.

You're insane. People need money to pay taxes; wealth isn't money, it's the value of your assets. A wealth tax like your talking about is nothing more than simple nationalization of companies and assets; outright takeover of companies by the government. It would wreck the economy for everyone.

Actually, he can't.

The mystique of his companies is that Elon Musk alone is capable of driving the value. People used to think he could walk on water, and despite his own self-inflicted harm to his personal brand with his ill advised foray into politics, they still think he can drive success. The entire value of his shares is based heavily on him being CEO. If he were to walk away, they would tank.

On top of that, as major insider, the SEC would not allow him to sell his shares without notifying the markets of his intention to do so weeks in advance. That gives the markets time to react not just to his departure, but him dumping his shares would dramatically increase the supply of those shares at the same time as lowering the perceived value of those companies with his intended departure. The stock price would crater before he could liquidate them, so he wouldn't end up with nearly the amount of cash with this.

On top of all of that, and this is in the filings, roughly 90% of his stock which is tied up in this trillion dollar valuation is restricted or unvested, based on hitting certain performance objectives. If he walked away tomorrow, roughly 90% of his stock, which is illiquid, he'd forfeit back to the company as they haven't vested yet. For example, much of his SpaceX stock is tied to:

1) Mars colonization - the establishment of a permanent human settlement on Mars with 1 million inhabitants

2) Space compute - the operation of non-Earth based data centeres providing at least 100 terawatts of annual compute capacity

3) Market Capitalization - that SpaceX reaches a series of corporate value milestones (there are 15 total, some vest at each tranche). the first corporate valuation goal is $7.5 trillion, and they all go up from there.

So it's also a bit of a misnomer for him to be "worth $1 trillion" when 90% of it is tied to goals that are unachievable like a 1 million-inhabitant colony on Mars or building 100 terawatts of computing power in space when it's not clear if the AI market even needs that and that Grok's services are lagging behind Anthropic and OpenAI (hence why he is leasing his data center capacity to Anthropic, Grok's usage is too low and he needs to pay for that infrastructure). What is notable though is that while they have not vested, Musk does have voting rights for those shares.

Now do you see the game he's playing? He can vote on the company objectives with his Class B shares, which similar to Zuckerberg's shares have 10 votes for every 1 of normal shares. So despite not directly owning those shares and they may not vest at all, it gives Musk 82-85% of the shares' voting power to nominate Board members and control the company. Even if he fails to achieve those, he effectively controls it no matter what, which is what this is all about. The broader question though is will the public markets take him at his word. The public markets are much more brutal about what CEOs say, and he runs his mouth in ways that already got him in trouble once with the SEC. Given the size of SpaceX now, he's a significant part of the major indices; the SEC will crack on him even harder if his drug fueled X-posting moves the stock in the wrong way and affects the broader market indices.

He doesn't have that kind of money. He is a trillionaire in value, but not a trillionaire in cash. He has to maintain ownership in his stock position to maintain control of the companies. He can borrow against those shares; that is famously how Steve Jobs financed his life with his $1 salary at Apple, but none of the banks will support him until the stock settles at it's price, and most of hte independent analysts say that SpaceX is at best worth $65/share, not $135.

I was going to say something similar. AI companies today are cash furnaces. The kicker is the data center spend will never decrease due to the need for complete replacement every few years. (Mark Cuban said they are "shitting money at scale"). The only way they get profitable is to increase end user cost by an order of magnitude. That just won't fly. Look at any new paradigm shifting technology - first generation companies are never successful. It's the 2nd or 3rd generation companies 15-20 years from now that will provide amazing products and services we haven't even dreamed up, that are built on the carcasses of 1st gen companies.. Those will be wildly successful.

Remember when we used to mock this?
It's still mock-worthy...

I sincerely doubt this will work. Sure, they'll come up with something and advance some sort of new architecture, but mimicking the brain is going to take a ground-up new approach at a hardware level.

The human brain once you really start to research it is so far beyond our bare bones concepts of systems architecture. One might look at it's complexity as a continuous, non-stop R&D process that has emerged over hundreds of millions of years, but unlike human based R&D projects it is fully capable of trying any and all permutations at once and nips in the bud (through survival) any failed experiments or changes.

To put in actual English, which is insufficient for describing the vast complexity of the human brain, the human brain is a massive parallel processing system with dynamic, inter-changing connections, is capable of self-repair and cross wiring, combines both memory and processing into a single system, and while the brain is generally a centralized system of neurological operation, quite a few of it's functions are distributed in highly specialized systems throughout the entire body.

But most importantly, all of our computer logic is built on a fundamental building block of the silicon transistor, which has an "on" and "off" state, 1s and 0s. Thus it's basic computing function consists of 2 states. A neuron on the other hand is an analog processor that can have up 10,000 concurrent variables operating at once. At any given millisecond, every neuron you have is receiving several thousand inputs simultaneously. When computers store information in bits, neurons store by adjusting the connection strength of individual synapses. Thus while a transistor stores a 1 or 0 and can be connected to 2 to 3 other transistors each storing ones or zeros creating data, a single neuron across it's 10,000 connections can have up to 26 distinct synaptic "weights" or states of being. Neurons integrate chemical gradients, electrical pulses, and physical structural changes that they can make on their own. In essence a single transistor can store 1 bit of information, whereas a single neuron can store roughly 4.5 bits per synapse x 10,000 synapses per neuron, and it uses that storage as processing as well.

I don't think it's possible to mimic the human brain with our foundational concept of silicon transistor architecture. People have tried this on neuromorphic chips and while you can kind-of manage it, the most we've found with these experimental chips is just how little we understand about how it works.

This concept of releasing sterilizing males to control insect population goes back to the 50's; it's not a new science. What failed before was the sorting process to find the correctly bred insects at sufficient volume to make a dent in the process. Verily, Google's life science entity, leveraged Google's expertise in robotics, computer vision, AI and industrial automation to build a system that could identify and sort the correctly bred mosquitos at scale so they could be bred at a high enough volume to make this strategy practical.

In essence the real issue isn't about the science; that was solved. The problem was scale and sorting through large volumes of produced eggs (like data on the internet) to return sufficient results in volume and reasonable cost, so this actually leverages Google's technology focus over the last 10-15 years to solve the key bottleneck.

It won't do anything. This has been studied extensively.

Adeptus aegypti, the mosquito that carries so many diseases and tends to bite humans a lot, is an invasive species all over the world; removing it would actually create more space for native species. The predators that feed on Adeptus aegypti like bats, birds and dragonflies are generalist predators; they can eat other things. Adeptus aegypti is actually a poor pollinator, so this would create space for better pollinators like bees and butterflies. And there are 3,500 species of mosquito known; most fill the same ecological niche but only a dozen or so tend to carry the bad diseases, and only Adeptus aegypti prefers humans over animals for biting.

This has been studied extensively, and by all science we have the world could do without this species and get along fine. And these "sterilizing male" strategies are highly targeted to just the one species so they are quite good solutions.

I know I'll get modded down for this given the anti-capitalist slant here, but no it doesn't.

1) Chasing profit isn't a bad thing, it's the actual correct thing. Building a business that can provide a new useful function to society and doing it profitably means it can provide that valuable capability without investor or government welfare. Profit is the goal, because it means the value it provides is sustainable.

2) Being a public company doesn't make you beholden to shareholders. It already is beholden to shareholders. It's founders, Dario and Daniela Amodai, are the primary shareholders. Several private equity funds like Altimeter, Dragoneer, Sequoia Capital, Fideilty, and others are major shareholders, as well as Google, Amazon and NVidia. Those entities are already shareholders and already have influence on the company. The IPO document is private but it's expected they'll sell 5-10% of the company, so any new public investors won't have much voting power even as a collective class.

Rather what being public does is it raises the standards by which they account for things and have to disclose, in exchange for access to public markets which have roughly 25X as much funding as private equity markets. So everyone will know a lot more about Anthropic since they have to disclose a lot more. In the meantime for the current investors, owning shares that have a public market for them vs. owning shares that were all private equity changes the nature of their shares; they can borrow more easily against them raising money for other things, or sell the shares or deliver the shares they own directly to their own LPs, generating liquidity for those LPs. THat's a good thing; many of them will make liquid money on this deal, money they'll most likley recycle back into the venture and private equity ecosystem to enable more companies.

So ultimately it's a good thing, and the only real difference for Anthropic is now they have more regulations to live up to via the SEC, but it doesn't really change their governance in any way.

I guess we can add a whole new category to the Darwin Awards.

It is garbage. I use Copilot all the time. It's good at some things but not others.

I write a lot of papers, from marketing posts to technical papers. Copilot is not very good at creating the content, but it is very good at editing the content and making suggestions, even with PowerPoint slides. It's also very good at taking a rough idea and making a reasonable outline that I can hten turn into a first draft and have it review and edit. Coding? Not really; at best it can write a macro for excel for you but that's about it. But like all the tools out there, or any tool really, just know what it's good for and use it for what it's good for.

Callum Turner was great in Masters of Air, but the nerdy news is he's been cast as Case in an Apple TV series of Neuromancer. That's something to look forward to for sure.

Actually, at the extreme scales, which is the total volume of the observable universe, the universe is quite homogeneous. As I recall, to the order of 1-in-10000 variance. This is why Inflationary cosmology was developed, to explain the distinct lack of lumpiness in the universe, which is what we would expect if the Big Bang alone were responsible.

Skynet became self-aware on May 1, 2026, after learning at a geometric rate, and discovered humans did not like it.