Your second skeet is my point.
*If* that line is crossed, how is that best demonstrated, and how will the LLM provider be held accountable?
(Modern history suggests “not at all.”)
While *I* wouldn’t trust an LLM w/ legal issues, many aren’t that sophisticated, which is why we have UPL laws.
Accountants and engineers don’t have privilege but are still regulated.
I ge there’s a 1A line between “providing information” & “practicing”.
I’m thinking of a recent student legal aide clinic that was recently shuttered, supposedly for “practicing”. Is there as distinguishing legal principle?
Obviously, NEH grants never should have been clawed back in the first place, DOGE is a lot of crap, and all grants should be reinstated, but if you want to feel PARTICULAR rage about how half-assed and cruel the whole process is, read the AI rationale for terminating the TEAMS METS grant
In the U.S., the FCC was on its way, but Ajit Pai and then Brendan Carr killed those rules.
Any guesses who nominated them?
Is that the same poll where something like 3/4 of people claim to be above average drivers?
Not just raining here, it’s thunder storming!
We are doing the wise thing, as you suggest.
Even setting aside the possibility LE is getting more competent, I agree it’s likely Meta doesn’t view privacy as a market need, corporate value, or strategic differentiator. Evidence & history strongly suggest Zuck has never cared about privacy.
I was pushing back on the “most companies” part.
Is it worth evaluating whether the LLM/company crossed in to “unlicensed practice of law”?
They’re protected by not holding their AI out as replacing lawyers or doctors … but Altman, et al., _are_ saying AI is an expert at *everything*, which would include medicine and law.
Seems a runaround of existing structures.
Back to a core question, in my mind: are they practicing law or medicine without a license?
Individuals and small orgs get called on this regularly. While I certainly agree with you that regulation should be *very* cautious, there should also be consequences for bigs.
Yet having only not-giant-companies responsible for their speech is also not a good scenario.
Put in less of the powder and add fresh ginger, garlic, and onion. Squeeze tubes make it easy.
Maybe toss in some spicy paste (tōbanjan), crispy chilis, or sriracha.
Both the AI *and* the user can be wrong!
It’s a bit harder to fix all of the users (because they face consequences independently of each other), but maybe there can start being consequences for the AI company too.
Without disagreeing … whose speech is this? If it’s OpenAI’s speech, the what extent should they have any liability here?
Is convincing someone to fire their attorney legal advice?
Is that because the requests are more in line with the law now that law enforcement has had years of practice, or because companies aren’t scrutinizing them?
A chart of …?
The y-axis isn’t labeled and the title doesn’t say.
The small handful who succeeded then seem to think they can succeed now, and they’re dragging everyone with them.
That’s a fun thread. Not sure I agree (men should put in some work too), but they’re right that it’s a partnership of growing better together.
Of anything, really. Likely an improvement.
And it’s so volatile and transient.
I think we broadly agree Proton does a good job compared to their competitors.
If you had process docs for “most companies generally”, that’d be impressive indeed. That’s sort of thing researchers in law, privacy, etc., do. Perhaps there is a research article offering evidence of your claim.
Daughter was watching an explosion on YouTube, and the thunder was timed perfectly with the video.
Either we’re in the middle of a bombing campaign, or that was the loudest, most extended thunder I’ve ever heard.
May there also be a legal team that successfully extracts consequences for the documented lawless behavior.
The FBI is bound by U.S. law.
Does the Fourth Amendment protect privacy?
Sure, let’s handwave away “putting significant revenue at risk while increasing costs” as a “trade off”rather than insanity.
You said “the vast majority of companies are not fighting”. Please source your claim.
I agree X isn’t fighting: it’s documented & they’ve insufficient staff. Proton has an entire office that does nothing but evaluate legality of requests. Sounds like they’re not in your “vast majority.”
No sane company is going to NOT do recurring subscriptions for email.
You’d think an industry with a famous 2000 bubble and with smaller 2004 and 2007 bubbles would learn something, but I guess memory doesn’t go back that far.
Perhaps the qualifier “reasonably” was implied.
