undefined | Better HN

0 pointshliyan2y ago0 comments

A company named openai choosing safety is bad because safety means as closed as possible, is a very questionable linguistic contortion. Most unsafe things happen behind closed doors.

0 comments

gnulinux2y ago

Ilya himself said he's against fully open source models because they're not safe enough. He's definitely against open source, my hunch is that we will see OpenAI being less open after his takeover.

Full interview here ("No Priors Ep. 39 | With OpenAI Co-Founder & Chief Scientist Ilya Sutskever" from 2 weeks ago): https://www.youtube.com/watch?v=Ft0gTO2K85A

j / k navigate · click thread line to collapse

0 comments

gnulinux2y ago

Ilya himself said he's against fully open source models because they're not safe enough. He's definitely against open source, my hunch is that we will see OpenAI being less open after his takeover.

Full interview here ("No Priors Ep. 39 | With OpenAI Co-Founder & Chief Scientist Ilya Sutskever" from 2 weeks ago): https://www.youtube.com/watch?v=Ft0gTO2K85A

j / k navigate · click thread line to collapse