How on earth can you "exhaustively" deidentify millions of chat logs that could contain literally any personal details, and presumably all without OpenAI's own employees also sifting through personal information in exactly the way they're claiming would be bad if others did it?

Nothing would make it “help get a little closer to making it a reality” if it’s not physically possible, and there’s a very strong argument that that’s the case. If nothing else, the maximum specific tensile strength allowed by covalent bonding - which is fundamental physics that we can’t change - combined with the reality of defects in a 36,000 km cable - is far below what’s needed to build a space elevator in Earth gravity. It might be possible to build a space elevator on the Moon or even (in the far future) on Mars, because their gravity is such that real materials could potentially do the job. But doing that involves bootstrapping an entire offworld industry, which is far beyond anything even the most advanced nations are capable of currently, let alone a technologically stunted oil state.

There’s no chance they’d actually succeed in building a space elevator. The Line is at least theoretically physically possible with current technology, even if it was always an obviously spectacularly bad idea. A space elevator is not.

Of course any revenue, including the shadiest one would be used to fund the expenses, but why not say that it paid dividends or bonuses for the execs or whatever lobby they're doing (I assume there's a fund for that (I mean a publicly known one).

Is the idea here that high frequency trading and self-dealing can be used to pump-and-dump a given proposition?

So, I find some low-traffic topic suggesting that Pigs Will Fly by the end of 2025 which has "yes" shares trading at $0.01. I buy a bunch of "yes" shares and then buy/sell a small chunk of them back and forth with myself, driving the price up to $0.50. Now I sit back and sell off my "yes" shares for something between $0.50 and $0.40 to anyone who shows up looking to get in on the rapidly-rising "Pigs Will Fly" proposition until a whole bunch of people have bought up the $0.01 shares for 40 times their actual value.

Or is there some other scam at play here?

Not sure if it's 1:1 with this new matter thing (I think it's their third push in this direction, at least), I can only recognize the door/window open sensor (even that is SLIGHTLY different), anyway I have a bunch of door sensors, leak/water sensors, and movement sensors. They all work over zigbee, went into Home Assistant with no fuss, I think they can do even firmware OTA updates. Best design ever for each specific use, batteries last forever (and work, even better, and report well the capacity for NiMh), super-happy. In some threads I follow people reported repeatedly that they had problems with the batteries draining quickly, but mine will go over a year for sure (I recharge them just in case once before the winter together with the ones for some thermostats and such) but it might be that there are some quality control issues, OR some environmental (radio) factors that might play a role here.

No comment.

Remember when Windows 95 wasn't installing the IP stack by default? Better yet you had to buy it as Trumpet Winsock for Windows 3.x ?

That skill proved useful in his later career.

Russia is testing nuclear delivery systems like their new "Skyfall" missile. But they're not testing warheads. Now, in fairness, Trump is very old, quite possible senile, and not terribly bright so it's entirely possible that he doesn't understand the difference between Russia testing a missile and Russia testing a bomb. But his order is making news because, as written, it's calling on the United States to resume the live-fire testing of nuclear weapons and we stopped doing that in (off the top of my head) 1992.

If Trump is trying to go tit-for-tat with a rival over nuclear weapons testing it's basically North Korea. China hasn't shot one off since 1996, Russia stopped before us in 1990.

Both China and Russia are suspected of having run clandestine tests in the 2000s but if US intelligence has more than a suspicion they're playing it close to the vest.

Testing of nuclear weapons among the major nuclear powers tapered off with the end of the Cold War and the international norm against testing creates a real disincentive to test, even in well contained, underground scenarios.

Back when testing wasn't so taboo the United States had a HUGE advantage in terms of the measurement and recording of test data. That advantage stemmed from computing advantages which have since ebbed. Normalizing live testing gives Russia and China an opportunity to catch up with that data and modeling advantage consequence free. "The US is testing, so we should too."

Trump isn't leaning into testing because Russia or China told him too -- he's just a vainglorious blowhard who likes the idea of setting off nuclear weapons -- but this nevertheless benefits American adversaries a great deal more than it benefits the United States.

Per mentioned 18 months that would cost otherwise $399.

THIS. I have some Raritan lights out management boards that need some Java run in a way that isn't at all supported in any modern browser, even with any tweaks, whitelisting or anything. They have an app now but such older boards aren't supported. The only simple way to connect is just to use a VM with Windows 7 and Internet Exploder. Then I realized even Windows 98 would do, faster and the VM is tiny. Faster CPUs need a small patch but work just fine (as long as it's still on an x86 CPU, I just couldn't figure out how to run it reliably with qemu or dosbox on a different CPU).

For sure. And that's why I think it's important to distinguish harm caused despite good intentions and reasonable practices being followed from harm caused because someone did not follow reasonable practices or actively chose to cut corners.

Just my personal opinion, but given the track record in this particular industry, I think there should be demonstrable intent by decision-makers to follow good practices, not merely a lack of evidence of intent to circumvent or cut corners. This is expected in other regulated industries, compliance failures are a big deal, and for good reason. I see no reason why similar standards could not be imposed on those developing and operating autonomous vehicles, and every reason they should be given the inherent risks involved.