3 ms·
New AI chatbot is scary good [Axios on ChatGPT]
- mdp2021 4y agoThe article insists on the point of being «"confidently wrong"», of narrating fiction without signals of statements being invented - and remains oblivious of the critical point that this engine has "unreliable pseudo-reasoning". They kind of re-invented in worrisome forgiving terms the notion of "artificial bullshitter" shifting it in the direction of "liar", instead of "does not really know what it is talking about".