arder

This is only a partial solution to a problem though. If x ai fails to build a good model they rent to Google. But that means all these companies are incentivised to build as much compute as possible. If they win their margins look great, if they lose they still make money. BUT at some point aggregate supply will outstrip aggregate demand and the bottom will fall out of the market.

This was covered in some of the coverage much earlier about the bankruptcy proceedings. Essentially it's within the court's power to not only pay the creditors but to pay them interest on their losses if that's possible. That's obviously balanced against the rights of the equity holders to get back any value if there is any equity left. In this specific case the people with equity were: the fraudsters who ran the scam, venture capitalists. No one wants to pay off the fraudsters and the VCs would rather pretend this whole thing didn't happen (because it makes them look like drunken coked up degenerate gamblers on a weekend at Vegas) so everyone agreed the court should be generous with how they calculate the return to customers. Government entitires also have a claim for any fines or taxation or whatever, but they're relatively happy for the retail customers to be protected first too - but they care less about the equity getting wiped out.

My guess would be anything above what the customers get may well go out to government fines rather than returning to the equity holders.

While that's true, if you have a good understanding of bankruptcy law, and a reasonable understanding of what assets FTX owned you could make a reasonable guess of how much money the eventual bankruptcy would pay out. Since the administrators publicly published what assets there were quite early on I think it's fair to say $270k always looked like a good deal.

In our libertarian utopia every employee at every company would receive a complex set of financial derivatives to incentivize them to manipulate the stock in exactly the way we want. Will Bob from Accounts push out recognising some revenue from this Quarter so that we miss projections and he can profit off some short dated Puts? No, because we had our HR team string together a series of quarterly short strangles to incentivize him to stabilize our share price.

The difference between a betting company the stock market is that in the stock market you have regulation to separate the exchange from the market maker. So in the stock market you go to the exchange, you buy or sell at the market price (or place an offer in the book) the exchange takes a commission but you're trading with a 3rd party. In the gambling world that generally doesn't happen, the exchange and the market maker generally are the same person - and there's good reason for that, if you can do it it's a better business model.

But either way, the same effect occurs. The sharks in the stock market profit from making good trades and in order to account for that the market makers have to quote a wider spread in the book, which effectively means the retail trader pays a larger spread. The net effect is money transfer from the retail trader to the smart trader. Would it be better for the market maker to just refuse to trade with the smart trader and then give the retail investor a better spread? Welcome to payment for order flow. Could a smart investor pretend to be a retail trader and get some good trades through Robinhood? Maybe, that's pretty analagous to what these guys are doing.

These are different mechanisms for the same thing but I'm not certain one is clearly morally superior.

The problem is the other way around.

A: Can you hear me? B: Yes B: What time is it? A: ...

At the point that B has replied Yes, B knows that it can hear A and that it can send to A but it doesn't know that A can hear B. As long as A makes the first move in the rest of the conversation that's fine - the next message from A confirms that B's "Yes" was received, but if A has nothing to say then B has to send it's next query and hope that A received the Yes successfully. If it didn't then B thinks the connection is established but it actually hasn't been.

Well this just goes to the core of your view on the role of luck in life. Are there 1,000 startups coming out of YC every year and 5 of them are run by geniuses who single handedly disrupt loads of markets. Or are there 1,000 startups coming out of YC every year full of roughly equally good people 5 of which get extremely lucky and make boatloads of money.

Airbnb just forked hotels, Stripe just forked Visa.

I think the most acheivable way of having some verification of AI images is simply for the AI generators to store finger prints of every image they generate. That way if you ever want to know you can go back to Meta or whoever and say "Hey, here's this image, do you think it came from you". There's already technology for that sort of thing in the world (content ID from youtube, CSAM detection etc.).

It's obviously not perfect, but could help and doesn't have the enormous side effects of trying to lock down all image generation.

Well for a start to run an even passable market making operation and to get into the jobs Sam did you have to be atleast fairly good at the mathematics behind probability. He's definitely more able to reason about probabilities than the average person. But sure, "I'm gonna go steal all these guys deposits to go gambling" wasn't an aggressive but understandable bet, it was the act of a degenerate gambler.

Nate Silver interviewed as SBF as part of research for his book and I think the big take away from it was basically that Sam's attitude to risk was pathological - he was willing to take any sized bet that he thought was positive expected value. The obvious problem with that is that you if you continually take higher and higher risk bets it's certain that you'll eventually lose one of them.

It's difficult to see when the 80% will ever pay off. They're a year in and he's got 94% health - but with that limit in place that's an effective battery capacity of 73% (0.94*0.8). At the rate that the others' batteries are degrading it's going to be close to 2 years before they even hit 80% capacity let alone the 60-70% that would make the 80% limited battery have better total real battery time. And at that point what are we saying? You're trading a good chunk of your battery in the first two years of its life for a few percent more several years into owning the phone?

It seems difficult from these numbers to see when, if ever, this choice is going to pay off for the people who opt in.

Hold up, they didn't do nothing. They tried and failed. There were several attempts at packaging an FPGA with a CPU and none of them worked.

It's really sad to see how far Intel has sunk, back in the good old days they'd lay off 10,000 at the drop of a hat. These days they're so rubbish it's taking them months to deliver those juicy juicy firings. Nice to see they're still doing well with their arbitrary movement of thousands of employees between TLAs though. Man that CCG is looking juicy right now.

Mr Beast videos do single digit millions in revenue per video, and he operates on razor thin margins re-investing everything. Youtube does $8.5Bn a quarter in revenue. For startups the target is the Youtube exit, not the Mr Beast exit. In fact, whilst Mr Beast is obviously doing a great job and making tonnes of money it's not clear if he even ever could exit. What Mr Beast is doing is incredibly successful, but it's not the silicon valley start up model.

This, like a lot of the advice is "Things that worked for me that likely won't work for you". A lot of people are going to talk to Mr Beast that won't talk to you, Mr Beast is doing a variety of one off projects that he'll never need to revisit. Mr Beast has a shit tonne of money and a shit tonne of resources. For all those reasons, it's something that he can do that you probably can't.

You know this guy gets paid by the word because he spends about a thousand words saying "The earth is more hospitable than Mars and there's still places on earth where life basically doesn't survive". Which is one argument, made in an extremely long way. But the counter is surely "Yes, everything that could live at the top of mount everest prefers to live at the foot of mount everest, that option is not available to life on mars". Or "Everest is uninhabitable because we've never tried to make it habitable".

I was hoping there would be a hard scientific fact the author would state like "It doesn't have a magnetic field and solar flares do X" but we didn't quite get there.

A lot of doomsterism in this article is incredibly one sided. Yes, it's true the sun will be around for another 4 billion years, but that doesn't mean we shouldn't go to mars, it means you get to ask the quesiton "Do you think at some point over the course of the next 4 billion years we'll successfully create a martian colony?". 4 billion years! We're talking about 4 billion years and you're worried about the carbon foot print of coffee beans?

I fundamentally agree that Musk's particular effort is pie in the sky. But (a) the steps along the way aren't (starlink, cheap satellites) and (b) it's his money he can fuck all the way off with it - I'd rather he spent the money shooting for Mars than shooting to land little orange men into the white house.

The argument he makes about buybacks just doesn't work. If the money didn't go to buybacks it'd go out in dividends. If there were no way to extract profits from a company and return them to shareholders then the whole investment model doesn't work. At the end of the day the owners of the company have the first claim on the profits. If the profits don't get paid out in dividends or used in buybacks they don't suddenly become the property of the employees. Having said that, yes stock based compensation works to align CEO incentives with share holder incentives and yes share holder short term incentives are to drive down employee wages.

Fundamentally I think the bigger issue is the taxation system not only preferencing unearned income, but then also being full of loopholes to mask income entirely. I think you'd be much better off proposing new rules around what it means to "realize" a capital gain (for example, if your asset is used as collateral, you are realizing it's value and it should be taxable). Rather than creating more byzantine rules around who can earn what.

Yeah I think this encapsulates really well what I don't like about "founder mode". Is it really "founder mode" or is it "Here are the reasons why I'm amazing and you all suck". A huge amount of these posts about what it means to be a founder is just very transparent bragging combined with a determination to bitch about other people.

I get it, you don't like typical corporate style management. Like Scott Galloway points out, part of the reason he's a founder is that he can't work for someone else's company. That's fine. But if your big take-aways are "You should respect me more and also you're less competent than you think" well that's not really anything to do with founding a company is it. That's just the exact old status games.

I'm not sure this is really good advice. I'm a hardware engineer and my industry is ~decades behind software engineering. Our compilers are slow, closed source, poorly documented. Every build system is some level of custom and broken. There are tonnes of people in our industry who still don't really get tools like Jenkins or git etc. And those people are going to ignore AI and they're going to be less productive because of it in the same way they're still doing bi-annual releases and churning through junior employees by tasking them with "verification".

It's very easy to look at those guys on twitter who a few years ago were claiming block chain was the future, or 5 years ago were claiming Tesla would make driving obsolete, and think "Yeah they're just bullshit artists". And a lot of them are. Weirdos writing egotistical dross about how you just don't understand how to be the ubermensch. You've got to do flounder mode or you're a nobody. I'm a naturally very skeptical person about all that - it's egotists on the internet trying either trying to be loved or genuinely trying to have a "brand" for business purposes (mainly the former).

But what you need to know about AI today is simple: that it probably is actually helpful for you today. Don't go off and learn it. Continue doing what you're doing, but maybe try out AI tools to see if they can help. For me at the very least ChatGPT is as helpful as Google for the first step in sanity checking something I want to do when coding. I think it's decent productivity boost and in my industry, you make money by being more productive than the guy at the company you're competing with so it's a simple choice.

There's no rush, you definitely can learn it later, but also - it's good today. It's not a waste of time - like learning about blockchain would be, or refusing to get a driving license in order to hold out for self-driving.

The Altera merger was a perfect storm. I could write for pages about what went wrong. But here's a few key things: First FPGA companies have been massively hit and miss with their product - they were perfectly capable of screwing up their next chip all on their own (it's a mutually beneficial duolopy of crapness with Xilinx). Second, when Altera agreed to fab with Intel the deal was written such that Intel could never buy any FPGA company other than Altera, so from day 1 this wasn't "Let's produce this together" it was "Let's line up Intel to buy Altera" and so Altera never had to deliver anything, they just had to wait for the acquisition. Third, Altera were one of Intel's first customers and it turns out that Intel's fab process couldn't make some of the gates that Altera needed, which is why in the end they partitioned off the entirety of the transcievers to a separate tile fab'd by TSMC and sellotaped to the edge of the Stratix 10.

Also just to check your facts on the 10nm fab problems. Intel's first FPGA for Altera was always planned to be 14nm. It was totally trash for reasons entirely within Altera's control (don't rock the boat, just wait for the acquisition to close). And the synergistic products were a dead end. I'm sure the 10nm catastrophe didn't help but really that time would've been well spent unpicking the disaster of Stratix 10.

It's the old Ex Machina problem though. If the machine is more intelligent than you, any protections you design are likely to be insufficient to contain it. If it's completely incapable of communicating with the outside world then it's of no use. In Ex Machina that was simple - the AI didn't need to connect to the internet or anything like that, it just had to trick the humans into releasing it.

This is just capitalism though. Like, what else are you suggesting? Opt for companies whose products are more expensive in the hope that they are less likely to hike prices later on? The real issue is when these companies become monopolies and then can raise prices without paying a competitive cost for doing so. But that's not Canva.

This really sets the government up for failure. The next time there's a tragedy on our rail system the question I'm going to ask is "Would this have happened if the person in charge took safety concerns seriously". This just makes Hendy's position totally untenable.

Well they're specifying the AMD EPYC and one of the things that the server line of AMD CPUs do that the consumer grade ones don't, is they have lots of connectivity. So for example an AMD EPYC 8324P is a 32 core CPU with 96 lanes of PCI Gen 5. Given that the 4090 GPU is PCI Gen 4, I think that's where you get the discrepancy. The 6 GPUS are connected in parallel to the CPU with 6 x16 connections (96 total lanes), the CPU could do this at Gen 5 (64GBs for each GPU) but the 4090 GPU is Gen4 only, so you'll only actually get 32GBps per connection.

I think the plan with this all along was George went off and built exactly what he thinks he needs for his specific work and then just makes it available. So is the wildly underpowered CPU bad? I don't know, I don't know his use case.

It also seems just weird from a business point of view. He's not going to sell many, he's not going to offer support, he's not at a scale where vendors are going to offer much particular support, and despite being absolutely tiny in scale he's still offering two totally different SKUs.

To be honest, I don't really see how it affects employees either, they've had multiple rounds of layoffs, their peak valuation was $11Bn and they got to the point where they were offering a share buyback program at a valuation of $300m. I don't see how any employee at this point could be expecting any value to the equity they got by working there. Having said that, after this funding I would expect the CEO to probably issue new incentive plans to employees because otherwise what are they even trying to do.

The real people getting screwed are previous investors who are basically being told they either have to put a tonne of new money in at an incredibly rich valuation or their existing holding is going to be diluted to 0. I don't see how you could possibly value the company like that and the fact that they're using this lever to force it likely means they think that actually what they can do is either get a very rich valuation from their hostages (prior investors) or kill the hostages (wipe out their equity and hand their share of the company over to the new investors). The result would either be a very generous funding round or a lower valuation in total because the new equity in would be getting a larger share of the company for their investment. In fact the rich valuation may be explicitly designed to do this - prevent prior investors from re-upping so that they can wipe out their stake.

It's clear that the returning CEO left with bad blood so I could imagine the structure of this was deliberate to screw the old investors who turfed him out in the first place.

If I were youtube I'd be concerned about the opposite. The whole point of youtube is you have all these different creators doing different things in different ways and youtube has stepped and said "No, don't do it your way, say our exact phrase and we'll help you". The downstream impact of that is Youtube's thumb on the scale is going to drive everyone to one particularly unconvincing phrase.

Sure, but the "PC Market" isn't anything like what it was. In the data centre their market share probably looks more like 20% and they have almost no presence on mobile. So yes, they continue to dominate 1 sector (and even that they aren't as dominant as they were) but that sector is now just 1 part of a market that has several significant sectors they are not dominant in.

It'll likely be the same for Google -they'll continue to be the dominant search engine, but search engines will be significantly less structurally important going forward.

One thing that I think this article misses is a way in which Intel explains it's business internally. The author kind of presents it as "You have CPU design companies, CPU design companies use companies like TSMC to fab the chips, but Intel does it in house and this is a competitive advantage". This is not how Intel explains it's business internally historically. Historically it's the other way around, Intel is a manufacturing company that happens to design CPUs to drive demand.

Also, I think that the thing about their chip design not being a relevant strategic advantage is wrong. The reason they lost the mobile market is to a large extent because their chip design was bad.

The other thing to mention of course is that strategically breaking the foundry off from the rest of the Intel business lines Intel up to basically dump their foundry the way AMD did. Strategically, Intel's foundry pays a massive cost (a) by canabilising the profit of the product group - you now have an internal allocation of costs/profits that don't functionally exist, and (b) competitors don't like to use your foundry, because you're likely to be their biggest competitor and have in the past used foundry customers as terrible acquisition targets.

Well to push the point a little - Imagine that Microsofts anti-trust case was in 2008, not 1997 and the iPhone was out and clearly gaining traction. Suddenly Microsoft doesn't look like some perpetual juggernaut. That's where we are with Google today - we can see the product that's likely to displace it. Windows is still big today, but it's not anywhere near the strategic force it was in the early 2000s. In the same way, I don't think Google is likely to be in the same place in a few years time.

I kind of am asking to have it both ways - they should be able to move against these companies earlier by levying smaller sanctions. That way they don't need 1 big killer case to get a judge to order the entire company broken up or spun off.