@david_chisnall

sympathy for a pathetic, lazy, rich asshole?

he should have his license to practice law revoked.

also, he committed perjury. he should be in prison.

@david_chisnall This is why I always refuse consent (while I can) to AI transcription and summarization. While transcription may be kinda-sorta OK and it probably uses older forms of speech-to-text ML, summarization is a big no-no because the probabilistic word extruder is liable to put words in my mouth that I never said.

https://sfba.social/@drahardja/117262408172666801

@david_chisnall I remain intensely frustrated every time someone says it’s me than just likely token generation, because clearly that’s still the enduring core of what happens. Everything can get better trained and given more APIs, but it will always be likely token generation

@david_chisnall i am not a lawyer but even i am aware of several cases like this where ChatGPT hallucinated facts, laws or legal precedents during court proceedings that got high-profile press coverage, so i have very little sympathy for a lawyer who isn't aware of those things.

@david_chisnall

i have no sympathy for this lawyer at all, he was playing with people's lives because he responded to AI advertising? he should go to jail, i wonder how many of his clients have because of his incompetence

@david_chisnall On one hand, it's the responsibility of a licensed attorney to be more careful than that; and anyone who has actually been keeping up with the profession should also be aware by now that AI has been a problem in courts.

On the other hand, if your "AI summary" product fundamentally doesn't do what you market it as, but sometimes basically succeeds anyway, while failing horribly other times... shouldn't there be consequences for that?

"You'd have to be stupid to believe our product actually works" isn't actually a defense anyway.