LLM is replacing therapists, advisors, lawyers, and even doctors in some cases, according to news articles. But guess what? None of the usual protections apply. If your private chat can be subpoenaed, there is no NDA. There is no legal or ethical obligation. Governments love this. Some human professionals will not hand over your data and will protect you, but how are you supposed to get protection from greedy AI companies ? This is another danger. So be careful what you share with these chatbots
Even OpenAI's CEO Says Be Careful What You Share With ChatGPT

The AI model is not your therapist and the company behind it can't necessarily keep it secret. Sam Altman thinks that's 'very screwed up.'

CNET
@nixCraft Good idea to say to chat gpt that you want to kill or killed someone, you can trust it for sure đŸ€Ł

@nixCraft

Thus the fix is simple; don't use large language models LLM that you do not control 100% from server to client

#LLM #InfoSec #Source

"SegĂșn artĂ­culos periodĂ­sticos, la maestrĂ­a en derecho (LLM) estĂĄ reemplazando a terapeutas, asesores, abogados e incluso mĂ©dicos en algunos casos. Pero, Âżsabes quĂ©? No se aplican las protecciones habituales. Si tu chat privado puede ser citado, no hay acuerdo de confidencialidad. No hay obligaciĂłn legal ni Ă©tica. A los gobiernos les encanta esto. Algunos profesionales humanos no cederĂĄn tus datos y te protegerĂĄn, pero ÂżcĂłmo se supone que vas a protegerte de las codiciosas empresas de IA? Este es otro peligro. AsĂ­ que ten cuidado con lo que compartes con estos chatbots."

@nixCraft

@nixCraft "can't necessarily"? He means OpenAI will not keep a secret, unless it's their own.

@nixCraft

prompts are still cheaper than healthcare in the US

@nixCraft perfectly timed with Proton's new AI assistant
@nixCraft it’s replacing husbands and wives too. It’s replacing human relationships.

@nixCraft

AI seems to be intended to grow a two-tier society.

The 1% get real doctors, lawyers, therapists, teachers and other expertise.

The 99% get AI bots, with a built-in 20% error rate, as doctors, lawyers, therapists, teachers instead.

The fossil fuel industry's obsession with investing in AI, when seen in the context of defunding public education, higher education, libraries, literacy & numeracy programs, research institutions, newspapers, & broadband can only be ...

1/

@nixCraft There'll be whole entire suburbs, cities of Wrong.
@nixCraft AI seems to be evolving as a "weapon of mass destruction"
@nixCraft continuing the proud trend from Uber and crypto etc etc, removing consumer (and worker) protections.
11 Things You Shouldn't Be Using ChatGPT For

This AI Chatbot has a lot of functionality for everyday tasks, but it's still unsuitable for some things.

CNET
@nixCraft IMHO "Ai" ain't replacing shit, instead it's displacing a whole bunch of stuff of which there was too few to begin with. @_elena