GorbachevyChase

Like Stargate? We’re investing $0.5T! Well, over 4 years, but $100B now for sure. Well, maybe just $50B and we will borrow the rest, probably. Well… not today. it’s more of a “commitment” than an investment. OK, OK, we haven’t actually raised any money for this.

I can understand wanting to be perceived as being on “the right team” but that comment is so silly that it undermines credibility. To put it otherwise, could you imagine a scenario where I had a labor, arbitrage opportunity that involved a higher paying job in Shanghai, China and that I had lived there for a few years to do that. Let’s also say that I was found guilty of some similar crime. Would you call me a good old fashioned red-blooded Chinese crook?

It’s OK to acknowledge that economic migrants are a thing, and that they likely have only transactional interest in where they live, such as a Bengali construction worker in Dubai, for example. That’s just part and parcel of labor mobility. For better or worse, shareholders, or middleman representing shareholders, have decided this sort of thing is a really good idea in the US, and now around half the population falls in that bucket. It’s a free country, and freedom means being free to choose short term interests. That also means you’re free to support such policies because they are good for Blue-team redistricting so we can provide free healthcare to all 8 billion people in the world somehow.

But please, nobody becomes a Yankee by the mere fact of standing on the ground. If you want that pejorative title, then you need to earn it.

Policy and practice might not be the same thing. The company and the entire management staff should be on somebody’s blacklist for future procurement.

To be fair, reusable rockets and electric cars people wanted to drive were not in much supply prior to those companies. Then again, the boring company is probably only at a parity at best with other miners.

I’m still not quite sure why we want humanoid robots and not something more useful or purpose-built.

What really bothers me is how an independent investigator made a compelling case that identified a member of the DC metropolitan police as the suspect who placed a pipe bomb on Capitol grounds. Then after years of inactivity, the FBI suddenly broke the case and arrested a mentally unwell black kid.

The whole apparatus is shameful.

They basically are. It really isn’t anything resembling an open market. They are effectively extensions of the state that happen to be funded by user fees rather than general funds.

Maybe, but if they spend more than $100/mo on alcohol, nails, or weed, then it’s a perspective problem.

Yes, Cuba definitely doesn’t have such wild delusions to the benefit of its residents.

Please stop. Critical theory is easy. Something about “X” sucks. Got it. What is the alternative? It’s the completely unserious philosophy of the peanut gallery.

I am wondering if everyone is moving to an IPO and striking these bizarre circular deals because they’ve hit the ceiling on what can be done with more compute until a major architectural innovation happens.

Still amazing, but 5.5 does feel like incremental progress with a massive up charge.

Adults aren’t the target market. I pay a premium because I know there are content standards and at least the big titles aren’t going to bombard my kids with inappropriate content.

So assuming you drive, when a police car gets really inappropriately close behind you for a couple minutes and then backs off, then they are probably using their eyes to look at your license plate and having someone run that or texting while driving to do that on the computer in the car. I don’t think there is a fundamental difference between this process and using a camera other than a camera doesn’t expect you to give it a pension.

Even given that I think solving the problem would require a certain amount of personal agency and volition to drive useful experimentation, and then you still have an inescapable problem that a design process is never verifiably done; it just a sense of taste when a product is good enough and it’s time to stop working on it.

I’m not sure this benchmark is even very interesting because it requires a language model do something that it really cannot do. Maybe it would be possible with a novel harness in an ensemble system, but I would never expect a pure language model that is run in a minimal harness to ever be able to do this.

Seems like a big deal. Surprised there isn’t much engagement here.

There are some comments which are identical to comments on X as well. That is not the say the frontier labs do not engage in highly unethical marketing, but this is a little bit too obvious.