Dr. Gebru Did Not Resign

Well... duh she was fired.

If you put preconditions on your employment, your employer is always free to say fine, you can leave right now and we don't have to negotiate further.

However, the whole outrage sort of presupposes that Google should be acting as an impartial scientific foundation, and not a money-grubbing for-profit company dedicated to protecting its own reputation and dearly wishing its own employees won't badmouth it specifically in public. In this current political climate, one must be very careful before saying your employer is both a massive polluter adding to global warming, and has an implicitly racist bias in their primary product.

Gebru was not careful at all, then demanded to know who in Google could possible dare object to her unveiling such truths (as if those truths were obvious and indisputable). Surely, if she got that information she wouldn't do something irresponsible like naming and shaming them on Twitter as naive at best and racists at worst (like she did to LeGuin in the week before she was fired).

Sounds like a classic "don't write this down or leave a paper trail" type situation. Everything else, though, should be subject to discovery by a competent lawyer if it has to go that far. Then again, she's probably under a forced arbitrage clause, meaning the cards are stacked against us to get any visibility on the matter.

This was interesting to me:

"But there is no hard requirement for papers to actually go through this review with two weeks’ notice. Numerous papers are approved for publication submission without meeting this “requirement”: an internal analysis shows that just under half of the papers submitted to PubApprove are done so with a day or less notice to approvers."

That doesn't show that 2 weeks notice isn't required. It just shows that some papers make it through the process very quickly. Perhaps those that don't have much controversy associated with them. That doesn't seem unusual to me, and doesn't seem to prove anything. It also seems curious that they didn't talk about how long that other half did typically take. Min/max/average would have been helpful.

Note that I have no idea if Dr Gebru was wronged here. Just noting that the above passage seems inconclusive to me.