I'm curious as to how it's different than what they've been doing in general.

Didn't they have big ol machine learning competitions for recommendations a decade ago?

Don't misunderstand what was found. What they found was fossilized blood cells and veins. I doubt they'll even be able to detect the ribosomes, which were almost certainly there.

Yeah, is an extremely well preserved fossil, but don't overread it. It's really no more surprising than the detection of the colors of dinosaur feathers.

Not until January 20, 2027.

It could well be that he doesn't trust either party. This is quite reasonable, and neither deserves trust. (Personally I tend to be left of most Democrats...so I think of myself as independent, but if the two parties are on the ballot, I generally hold my nose and vote for the Democrat, even though they hardly represent my position.)

Those are showing up. The robots *are* coming. But not just to the US.

The US surged to dominance in the aftermath of WWII, when most of the other major countries had sustains severe damage. Since then, they have recovered, so the US no longer has that much advantage. Several of the other countries are smaller, or have other disadvantages, but the US has some disadvantages of its own. And we've essentially abdicated out position. China seems to be moving in to fill the vacancy. (Don't take this too seriously, I don't spend much time analyzing this stuff, but that's my take.)

AI *IS* going to be extremely important and shape the world over the next decade. It's an extremely important thing to get right, and getting it wrong could be deadly. But it's already clear that it's going to dominate repetitive jobs. And China is well aware of this. So, apparently, is Amazon, and a few other places. What's not clear is just how cheap it will be. (But China is pushing it hard...so probably cheaper than humans.)

The US is not at war with Russia. I think that has been decided as the definition of an enemy country under the treason definition.

Odd, I thought he was acting more as a Russian puppet, even though China was the main beneficiary.

Both the Democrats and the Republicans are fascists by the central part of Mussolini's definition. It's mainly about centralization of powerful entities under the control of the state. If you believe in a powerful state, fascism is practically a necessary part of your beliefs. And both the Democrats and the Republicans believe in that, whatever else they disagree about.

For that matter, it's not clear how society should be structured without a powerful state. I've yet to hear proposed a believable anarchist system that was both scalable above Dunbar's number, and stable. Some MinArchists have a point, but the US started out as a MinArchist system, and didn't stay that way.

Weakest of the 5 games.

They messed up the level scaling something fierce.

Try DuckDuckGo. I rarely need to use anything else.

The number is at least credible and not far off from the optimistic studies of actu productivity gain.

This isn't "we're laying off 40% and we'll do more". An 11% productivity gain is in the realm of reasonable.

You're probably right, but this headline does look a lot diffy than other AI layoff headlines.

Would you believe, so they can be used as an energy storage system?

It's actually a pretty good idea, but only if you can keep the command-and-control system secure.

Alas, it's historically been easy for people to slip into a mindset of "no hacker would ever know or care about OUR little system, because [reasons]", and use that as a justification for ignoring security concerns altogether.

Of course that's never really been the case, but now that any bozo with $200 in their pocket can just pay an AI to find every possible security hole and report back, it's REALLY not the case.

Thank you for this sensible comment.

model weights already more or less "contain" several partial OS kernels,

No, they don't. They were trained on that data, but they are not that data. You can't extract it from weights, and I highly doubt you'd be able to reproduce it from weights, even on an exceedingly large model.

At best, it's a work inspired by the original, not even a derivative. There's no "substantial or material" replication.