Comment Re:Let me guess (Score 1) 60
Over 8 billion people have an IQ under 100%.
Over 8 billion people have an IQ under 100%.
98% of traffic problems on highway 17 are caused by dipshits who should never be allowed to use the left lane. 1% are caused by that one dip in the fast lane on the EB side that they don't seem to be able to fix ever. The remaining 1% are caused at the first exit on the inland side when it's busy.
Before Teslas even existed I was passing the clowns who don't know what a passing lane is for on the right on the 17. The only meaningful difference between then and now is that there's more drivers, including more of those clowns. Well, that and that the speed limit used to be 65, so more competent drivers could make it over the kill. Of course, it was killing incompetents, so they reduced the limit and we all have to suffer. This is what happens when you design your nation around the car.
Ostriches aren't really domesticated, just in captivity. They are not your friend. Even people who keep them have to take precautions because sometimes they flip their shit and attack, and they can kill you.
This "ambitious partnership" sounds like pandering bollocks. The entire press release is written like this.
You don't seem to understand the purpose of a press release.
Nah, there's some for-pay ones. I want all that AND I want it to be free too
The paper releases dust and fibers, which is why the correct tool is a foam swab.
That too has to be correct, though, or it will release dust. This is especially true if it is aged, so the foam swabs have to be fresh.
IPA is the correct solvent.
How much Intel stock do you have? Is it actually legal for you to mod me down?
That's not how LLMs work. They can only extrude what's in their training data
There are a whole bunch of examples of LLMs which have access to actual resources, so that's completely false. Why don't you go somewhere else and talk about things you know nothing about?
XML is far more readable than JSON. It does also suck, though.
I haven't had to do much with YAML, but I cannot imagine what you mean by "easier to transmit", given their similarity.
Why are we so predisposed to falling for this hype cycle?
The people without hope gave up and died.
The people with too much hope are suckers.
The people with just enough hope to succeed but not enough to be suckers took control of the system and manipulated it in ways that make people into suckers, and make suckers into bigger suckers.
If Musk cared, it was because he was worried about his speech being censored. His own actions prove that he's not anti-censorship in the least, as long as he's the one doing the censoring.
Here is where the honesty test kicks-in
At the point where you provide an amicus curiae brief as if it were a court decision?
Powell and Rice are not "said to have done so," they are perfectly well known to have done so. There was no legal problem with it then. Hillary operated her own mail server at their suggestion. And there was no legal problem with her doing so at the time she started to either.
I read an article on this a little while ago, I believe it quoted Powell as saying that he didn't advise her to do this until she had already done it. Who knows whether that's bullshit or not, or for that matter, if I'm even remembering correctly. I'm not currently on the machine with the relevant browser history, or I'd try to find it.
BUT, there absolutely was a legal problem with it then, especially for Powell, because he explicitly advised Clinton to run her own server for the purpose of avoiding discovery by being able to delete communications. And that is illegal when you are required to retain all official communications, which was the case.
I think most of us who have been here more than a hot second also remember discussing here that one of Clinton's staffers was reported to be seeking advice on email deletion.
The first two did not exist but the last two do exist but had nothing to do with the topic cited. It would still take a human to determine that these cases where bogus citations.
If you gave an LLM access to a legal database, not training it on it but allowing it to pull things from it, then you should be able to use it to check citations. It should be able to handle both looking things up and reporting null results, and also summarizing those cases it does find and doing a better-than-nothing job of reporting whether they were cited correctly.
If I were a lawyer I wouldn't trust such a thing to check my citations, but I would absolutely use it to look for low-hanging fruit in my opponent's work.
The AI should be smart enough to know what a citation is
Why would you think that? I thought you knew something about this technology.
Nobody's gonna believe that computers are intelligent until they start coming in late and lying about it.