A few folks asked about the content of press conference at NeurIPS, so here’s a thread...

The press conference mostly consisted of readouts from the chairs, the best paper award winners, and, of course, Terry. Terry always needs to make NeurIPS about him in one way or another.

These readouts could have and should be posted text on the NeurIPS website. Though for the paper awards, I’m not convinced they add much value beyond the papers themselves.

The readouts took about 40 minutes and then there were questions submitted by Zoom chat.

There were three questions.

@WillKnight asked for a comment on Timnit Gebru and if the conference had any reaction to that situation and the influence of corporate money and research on NeurIPS.

The chairs answered by focusing on ethics and inclusion efforts at the conference. Of which, now there are many and this deserves kudos to all who pushed so hard to change the culture of this conference.

However, the chairs didn’t address the issue of corporate influence on the conference.

Of course, the other two questions were about corporate language model GPT3, which won best paper.

I said it in 2017, and I’ll say it now, that this ridiculous conference has a press conference is a bad thing, not a good thing.

I said it in 2017, and I’ll say it now, that our field's main academic research conference has turned into a corporate job fair is a bad thing, not a good thing.

I said it in 2017, and I’ll say it now, that corporate research does not come with the same freedoms as academic research.

And I’ll say it now: I stand with Timnit.