ProjectArcturis

Uhh... Powell's term as chairman ends in May 2026. His term as a board member ends Jan 2028. Senate confirmation is irrelevant because Trump will not nominate him again for anything.

With shorting, you run the additional risks that you could lose the borrow and be forced to buy back at any time. Or get margin called if the price moves against you. With puts, you have to get the timing right, but no external factors can force you out of your position.

Read the article. He used options, which can't be margin called because they're already paid in full.

When Trump replaces Powell next year, the odds of the Fed doing something other than bathing a crisis in liquidity goes way up.

Michael Burry is a stopped clock. He's predicted 20 of the last 1 crashes.

It seems to me that if a puzzle has gone unsolved for 35 years despite many very skilled people trying hard to solve it... It is not actually a good puzzle?

Like, here - here's a code that no one will ever solve: ITIWKSMNDIWKD WJSIKWMWMSONQ

Turn that into a sculpture and put it outside the CIA.

At this point the British royals are basically tourist attractions. They probably generate more revenue than they consume.

This is probably an uninformed question, but why are you comparing layers' output to ground truth? Isn't the point that ground truth is unknown?

OpenAI is not going bankrupt tomorrow. But in the universes where it does go bankrupt in a few years, it would mean the AI bubble had popped, and the demand for top-end chips would have collapsed.

>These finance types (family offices, small secret investment funds) eat clueless VCs throwing cash on the fire for lunch… and they’re salivating at what’s ahead. It’s a “Big Short” once in 20-30 years type opportunity.

No - it's very hard to successfully bet against anything in finance, and VCs and non-public investments are particularly hard. When you go long, you simply buy something and hold it until you decide to sell. If you short, you have to worry about borrowing shares, paying short fees, and having unlimited risk.

How would you even begin to bet against OpenAI specifically? The closest proxy I can think of is shorting NVDA.

There's also nobody whose job it is to make big one-time shorts. Like you said, it's a once in 20-30 years opportunity, so no one builds a hedge fund dedicated to sitting around for decades waiting for that opportunity. There will certainly be exceptions, and maybe they'll make a Big Short 2 about the scrappy underdogs who saw the opening and timed it perfectly. But the vast majority of Wall Street desperately wants the party to continue.

In the world in which OpenAI goes bankrupt, their used GPUs would be worth perhaps 10% of their value today.

The situation certainly sucks for the actual PhD holders. Go through 5-8 years of intense work and minimal pay, and the outcome is maybe a 10% chance of getting the job you want?

It's unsustainable. I got my PhD 20 years ago, but I would certainly advise my children to avoid it.

What if I think all three of those industries are bad?

The 2A guys have only ever been LARPers.

You think he's calling the gambling industry deplorable because of information asymmetry?

Might have more to do with exploiting addiction and ruining lives for profit.

Maybe 1/100th of that?

I'm confused why this is addressed to Azure instead of OpenAI. Isn't Azure just offering a wrapper around chatGPT?

That said, I would also love to see some examples or data, instead of just "it's getting worse".

House prices actually have nothing to do with CPI. "Shelter" is calculated using "Owners Equivalent Rent": https://www.bls.gov/cpi/factsheets/owners-equivalent-rent-an...

It's pretty clear that for-profit entities will optimize to maximize profit at the expense of all other variables, including safety and the environment. We certainly can't trust the CEO class to properly prioritize safety concerns. So we will always need government regulations to keep these tradeoffs balanced.

You seem to be calling for a magical new set of regulations that are just as effective but don't cost as much. That's not going to happen.

You should reexamine your media diet, because it has you believing absolutely insane things.

Well, c'mon, let's be fair. Your source says overdose-related liver failure. And reading further, most of the overdoses are either intentional (suicide attempts) or a result of addiction and opiate/acetaminophen combination drugs.

Stock is down 3% after hours.

The powers that be don't care at all about thousands of insurance adjusters losing their jobs. But they care very deeply about vaporizing hundreds of billions in shareholder equity.

Are you suggesting that Jeff Bezos somehow convinced all his PE buddies to tank Sears (and their own loans to it) in order for him to build Amazon with less competition? Because, well, no offense, but that seems like a remarkably naive understanding of capital markets and individual motivations. Especially when it's well documented how Eddie Lampert's libertarian beliefs caused him to run it into the ground.

>few people have any idea of who conquered whom in 1620 AD and what were the consequences for their distant ancestors

I bet most people in the US could tell you in broad strokes who used to live in North America and who conquered them.

The bar is "can it write as well as these accomplished professional writers?", not "Can it imitate their style better than the average person?"

Who is the intended user for this? Pretty much anyone can calculate a mean and standard deviation without a bespoke website.

Yes, I think that's a reasonable guess. Far from proven, of course.

One of the reasons Alzheimer's research has been so ineffective is that we don't know what physiological role tau and amyloid play in healthy brain function. Evolution didn't invent them so we would get Alzheimer's.

Seeing that tau is elevated during development suggests what that role might be. It also suggests that blocking tau entirely in adults might not have any important side effects.

It's more expensive for a worse consumer experience.