On Wed, Jan 29, 2025 at 8:58 AM John Clark <[email protected]> wrote:
> > On Wed, Jan 29, 2025 at 7:35 AM Quentin Anciaux <[email protected]> > wrote: > > >> *>Yes, just use an AI that you trust to check it ;)Surely that's what >> anthropic and openai do or will do.* > > > *AIs can modify themselves so it's impossible to prove that Deepseek (or > any other AI) will never learn to be malicious, but it is possible to prove > that malicious code was not inserted right at the very beginning. * > > Can you provide a link or reference that shows that AIs can modify themselves? -- You received this message because you are subscribed to the Google Groups "Everything List" group. To unsubscribe from this group and stop receiving emails from it, send an email to [email protected]. To view this discussion visit https://groups.google.com/d/msgid/everything-list/CAMy3ZA-_h6W_u2pOLU1m2sEPY19TQBn_YhLJHgeUsVCyv0e50w%40mail.gmail.com.
