janus comments on Transcript of Sam Altman’s interview touching on AI safety

[janus]

You can’t finetune GPT-3.5 through the API, just GPT-3