darby_nine

> It’s still a bug

Only if it's actually possible to remove this behavior. I don't think there's any evidence of that at this point.

Cannot agree enough. Hallucination is a terrible name for this behavior.

> I interpret supernatural to mean religious in this context.

TBH I have no clue what the word "religious" means outside of abrahamic faiths. Might as well be a meaningless word. Anyway, the supernatural certainly encompasses many aspects of religion without implying anything about aliens.

JUST when you thought the chatbot was dead

Idk about this specific thing, but this is much simpler than it sounds if it operates like other anthropology. Ritualistic means intentional, and supernatural means symbolically meaningful. In other words, this is less of a specific claim and more of an "idk but they were clearly up to something".

i don't write understand how that would lead to anything but a slightly different response. How can token prediction have this capability without explicitly enabling some heretofore unenabled mechanism? People have been asking this for years.

Do either CCL or SBCL have any kind of partial evaluation or tracing compilation?

I feel like crowdstrike is perfectly capable of mounting its own defense

I don't get the distinction. Are you not just searching through chunks?

> the odds are very strong that there's a good reason for that

The easiest way to show this would be to give the responsibility of managing the bug bounty to a third party who isn't involved in the business.

"chain of thought" is just search, right? Wouldn't it make sense to tailor the search with heuristics relevant to the problem at hand?

Sure, but companies and shareholders are the most destructive forces on earth. I have a hard time seeing investor tears as anything but a glimmer of hope.

Dockerfiles compose and aren't restricted to running on linux. Those two reasons alone basically mean I never need to care about systemd again

Cool, i don't care about either of these things. Jobs are a poor index of economic health for americans and we should really be guaranteed one as a fundamental human right if we also refuse to implement modern welfare.

Good, let's learn to get along with people without trying to dominate them for the first time in living memory

Seems like a terribly irrational and inefficient way to run an economy but I'm sure russia will invade any day now. Or is it china these days?

Ah, my bad.

Ah yes you're correct, I misread the above comment.

Is this not the legally obligated line of thinking for american companies? You have some leeway to argue with only providing value to shareholders in some specific manner, but not much.

> The auto companies are strategically significant.

Not really, they're more politically significant than economically critical. I've certainly never owned an american car and never plan to. We'll be fine continuing to rely on foreign companies to produce our vehicles. Foreigners certainly don't want our cars (outside of maybe tesla in the nordics, i guess?). None of this is true for the plane market, or at least not until boeing acquired its current popular reputation of not being very good at making planes (deserved or not)

The fact we bailed out the companies but refused to take ownership should be considered treason. Same for the banks, the car companies, the airline industry we bailed out for more than its entire value, etc etc. completely corrupt and incompetent governance

It seems like all their other IP is derived from the game IPs, so I'm a little confused how this leaves Annapurna in a better position.

This is hilarious and rather sad. How did Annapurna fumble such famous talent so bad?

We all know boeing won't be "wrecked"; they're too critical for state interests to be allowed to fail. If they're smart they'll either nationalize the company to remove the profit inefficiency or force the company to split to remove the profit inefficiency. Either way, the profit motive and lack of any competition is clearly a national security risk.

> Are you saying chatbots don't offer anything useful over search engines? That's clearly not the case or we wouldn't be having this conversation.

No, but that is the value that's clear as of today—RAGs. Everything else is just assuming someone figures out a way to make them useful one day in a more general sense.

Anyway, even on the search engine front they still need to figure out how to get these chatbots to cite their sources outside of RAGs or it's still just a precursor to a search to actually verify what it spits out. Perplexity is the only one I know that's capable of this and I haven't looked closely; it could just be a glorified search engine.

> The formula for success in any field is simply to make a product that other people want to consume

Well, the formula for success in selling products is this. Most people don't define success in terms of business acumen.

Except, of course, businessmen. If you perceive our society as centered around successful people, of course you'll see it as merit-based. If you perceive our society as poorly run and catering to the rich, of course you'll see success as primarily a product of circumstance outside of your control. Is it so hard to see that "merit" is necessarily defined in subjective terms?

You could just as easily argue that model creators don't own the model either—it's like charging admission to someone else's library.

Is intel even in a position to be competitive on the fab level? I was under the impression intel's fabs haven't been "cutting edge" for a long time now.

> Is it the role of a for-profit public company to "support" another for-profit public company?

It is (or it can be) if you want to avoid antitrust lawsuits. Same reason Microsoft bailed out Apple in the 90s.

Very fast collapse typically comes with positive evidence of this as people leave their lives out in the open, expecting to return later. Eg see the Roanoke colony in Virginia.

Then it shouldn't be used for secure input.