That’s the thing from the customer IT perspective. When your Microsoft-administered systems go down at the same time as the rest of the world, nobody blames the CTO or any of the sysadmins. The global outage becomes an acknowledgement that it’s not their fault, so nobody in the company actually gets blamed, and nobody gives them a hard time.

It might not result in more uptime, but it does mean that any downtime doesn’t get pinned on them.

Nvidia designs the chips that TSMC manufactures and sells to Supermicro to build servers to install in data centers owned by Coreweave and rented out to OpenAI.

You can mix and match different companies in each level, as very few are exclusive to one vendor/supplier/customer.

It’s a trash source that gets a lot of traffic because its editorial lean is anti-AI and it summarizes other paywalled sources. I’m not a fan, even if I’m also anti-AI.

Exactly. Cooling a maritime vessel is easy, because the ocean is a giant heatsink. The heat capacity of water is so far above almost any other common material, and water is plentiful in the ocean.

Put the fresh water in a closed loop and let salt water carry the heat away from some kind of heat exchanger.

It’s true that programs today are several orders of magnitude more capable, but it is also true that code now is hideously RAM inefficient and bloated.

The entire codebase of software in production, yes. But one thing that people might not realize is that some of our software algorithms themselves, at least for certain algorithm families, are also much more efficient than before.

I thought this was an interesting overview:

https://futuretech.mit.edu/publication/how-fast-do-algorithms-improve

The focus isn’t on memory usage, but the same general trend applies there: when someone discovered a new algorithm that improves on old ones, it can change the amount of memory used to accomplish a specific result.

I don’t think they need to make these announcements public if the target audience is their own developers. I know some Anthropic engineers, and some former OpenAI engineers, and from our conversations it sounds like both companies do rely on internal messaging to manage workplace morale and address concerns. And warn them about honeypot spies looking for secrets, apparently.

So maybe this message is mainly for regulators and investors, not necessarily their current employees.

Repaying debt is cash cost. But this is convertible debt that could be converted to stock, in which case the holders of those newly converted shares don’t get paid cash from the company, but instead own a chunk of the company worth more than that cash. So the conversion itself would alleviate the cash demands of the company, while also diluting the ownership stakes of any preexisting owner. That’s what that paragraph means.

Companies issue convertible bonds usually because they can get better interest rates on it, in the event that they need to repay it. Or they’re happy to deal with diluting their shares if the company becomes so valuable, because that means they’ve gotten a lot richer and just have to share their newfound riches with the lender.

That’s the tradeoff, of course. If a company liquidates, it pays its debts first, then its preferred stockholders, then its common stockholders. But each step down the line gives a greater upside in exchange for less certainty in getting paid at all. Bondholders get paid at higher priority, but have the exact amount they get paid defined by a specific mathematical formula. So they won’t get any additional benefit from the company becoming super valuable. Or they can get in the back of the line to see what unlimited upside they have from the company’s value going up a lot.

Also, Anthropic’s bonds aren’t exactly low interest. They’re right there at the edge between junk bond and investment grade.

Usually companies will publish an “earnings before interest, taxes, depreciation, and amortization” or EBITDA as a way to show whether their core business is actually profitable (even if shareholders aren’t the ones to get the benefit of that business’s operating profit). But that’s not a standardized metric, and Generally Accepted Accounting Practices (GAAP) won’t define earnings in that way.

The big problem is that the more a company has to deal with high cost of borrowing and huge capital expenditures, the more disconnected EBITDA is from actual GAAP earnings. I can build a hotel for a billion dollars and rent out a bunch of rooms, while each room transaction brings in way more money than it costs for me to have a guy at the front desk give them a key and housekeeping to clean their room, I need to actually make a lot more in order to cover the cost of all the furniture I put in the room, the construction costs on the building, and the interest payments to the bank that financed the whole thing. And if the furniture and the building won’t actually last long enough to where I made all my money back before I have to reinvest a bunch of money into renovations and new furniture, then I don’t have a profitable business. Or if it can be profitable on those numbers but I can’t cover the interest payments to the bank, then the bank will eventually take it over and I’ll lose the business entirely, so owning stock in that company isn’t worth all that much.

So if they’re saying they’re barely profitable when you take out the very expensive costs of training new models, entering big expensive deals with data center companies to build out infrastructure, the stock they keep issuing to employees and partners, etc., that might keep things going a bit longer, but it’s still not actually a path to true GAAP profitability.

This right here is how all of this will come crashing down

No, this is pretty normal, and more importantly, doesn’t actually cost the company cash. They gave lenders bonds that can be converted to stock, so this loss on their books is pure paper loss, the same way wealth in owning stock in this money losing company is pure paper wealth.

These numbers are just illustration for simplicity, but imagine a company borrowing $20 billion and telling the lender that you’ll either pay back the $20 billion plus interest, or they can opt to convert it into stock worth 10% of the company. When the company is worth $100 billion, the bondholders won’t convert to stock. But if the value of the company shoots up on paper to where that same stock would be worth $100 billion, the lenders would rather take the stock than the cash repayment. And because it’s worth $80 billion more, the company will have to record that on the books as replacing $20 billion in bonds for $100 billion in stock, and an $80 billion loss compared to their last statement.

But look at the underlying financial cash flow. When that happens, the company actually takes $20 billion in debt off their books. They don’t have to repay that loan anymore. So it improves their cash flow position at exactly the same time that they’re booking a loss.

The whole thing will come crashing down for totally different reasons, relating to operational cash flow, not the actual accounting shenanigans where one kind of paper number gets converted to another kind of paper number.

“we lose money on every sale but we make it up in volume!”

Does it?

Reuters is reporting that Anthropic’s Investor Day event a few days ago, where they showed their audited Q3 financials to investors, caused institutional investors to shift their sentiment and that they are being pushed to lower their IPO valuation target down from $2 trillion to $1.5 trillion.

Coreweave is down almost 20% this week. Nebius is down more than 15%.

Investor sentiment is shakier than it was a month ago. And for hype-fueled valuations, investor sentiment is a two-way street.

Once there’s a public expectation for profit from the two biggest AI labs

I don’t think that it’s necessarily the order things will go.

Anthropic and OpenAI have bills to pay, including obligations to pay a ton of money to partners/vendors in the coming years. They need cash to do that. And the cash can come from 3 main places: revenue from customers, loans from lenders (usually in the form of issuing bonds), and issuing new stock/equity to raise funds.

The IPOs are basically their last chance to raise funds by issuing new equity. Even if each of them raise $100 billion, that’s probably not enough to hit self sustaining profitability. As lenders start to get more skeptical, in a macroeconomic environment with high interest rates and high energy prices, they start to run into a very real possibility that they hit illiquidity issues, even if their core business model actually works by then (and I don’t think it will). There’s just not enough runway for them, even in the most optimistic long term projections.

I want a convertible, a laptop that folds all the way back and gives a tablet/stylus interface, but has stiff enough hinges to hold the screen like a laptop at any angle with a keyboard, and that closes and covers the screen like a laptop. They played around with some form factors like that 10 years ago but they never really took off. Plus they tended to be Windows only, with poor Linux support on the hardware.

Whatever the reason, it’s basically become true.

Annualized Revenue rate is a made up number

It is fundamentally a projection about the future using recent results, and transparent in its methodology. It’s shit methodology, but it’s not hiding anything. So I don’t know that I’d classify it as fraud.

Playing games with “revenue” is another problem, but because that’s misleading about the past, and making an active decision on whether to classify money that was never gonna end up in their pockets as revenue share to partners or cost of revenue. Companies sometimes engage in shenanigans where they give discounts but classify the discounts as marketing/sales expenses rather than foregone revenue. That kind of stuff strikes me as more squarely in the category of accounting shenanigans.

Financial Times smells blood in the water. They’ve been reporting on leaked OpenAI documents for a bit the last few weeks.

If the major financial outlets start reporting on the negative, there may be a sentiment shift that does actually destroy the companies propped up by hype rather than fundamentals.

The number they used for annualized revenue for these rapidly growing companies is taking the most recent month’s revenue and multiplying it by 12. It’s not a guarantee about the future, but it is misleading.

The extreme version is taking a good week’s revenue and multiplying it by 52, or even worse, multiplying a day by 365. Those are outright dishonest. But not accounting fraud, because it’s not accounting at all.

The tablet one eventually came true. Desktops and laptops aren’t anywhere near as popular for personal use by non-technical people.

It is believed that pi is “absolutely normal” in the strict mathematical definition: that in any base numbering system, any finite sequence of digits appears as frequently as any other sequence of that length. If that’s true, then yes, a particular binary string, including one that corresponds to a particular digital file (like a particular mp4 file), can be found in pi.

That said, it hasn’t actually been proven that pi is normal. If it’s not normal, then the number can be infinitely non-repeating and still never hit that particular sequence of digits.