On 2026-09-15 11:13:39, Kevin Chadwick via Dng wrote:

> > You expect the AI has the project you ask it to, but you never know if it hasn't
> > another, hidden, agenda.
>
> True, but not because AI have any consciousness as that is just one of the lies
> aimed at investors and adopters. The real danger is how easily even large
> commercial models can be poisoned. Potentially an xz systemd ssh CVE situation.

Hidden corpocate agendas, poisoned, or just plain hallucinating. Pick three.