Follow

So, Google fired their star AI ethics researcher Dr. Timnit Gebru after she tried to publish a paper demonstrating that Google’s BERT natural language model is unethically biased along gender and racial lines and not fit for use. It sounds like Google squashed the paper because it found the contents “objectionable”, and is claiming that Gebru chose to resign instead of being fired.

wired.com/story/prominent-ai-e

@thegibson Google’s going all in on the cyberpunk corpo memes.

I'm willing to bet she did resign rather than "be fired", technically. High-visibility people are often offered a resignation parachute where they sign an NDA and perhaps some other terms (like not pursuing a particular topic for some time or agreeing not to sue over a specific thing) in exchange for a chunk of money; it's take that deal or be fired with nothing but unemployment to fall back on

I would 1000% not blame her for taking that option

@calcifer boosting this since it's an important perspective to remember in a hellcountry without social safety nets

@calcifer Yeah... the more I’ve looked into it, the more conflicting information I’ve found between her statements, those of Google’s head of AI Jeff Dean, various other google researchers, etc.

It’s very difficult to know what is true, what is PR, etc etc.

That means we are left with nothing to go on but for how we feel about the incomplete or incorrect info we do have access to; which isn’t the basis for any kind of rational analysis.

@vortex_egg the emails about her departure that were posted in another article were gross as hell. I guess I don’t expect much from goofle anymore, but fuuuuuck.

Sign in to participate in the conversation
hackers.town

A bunch of technomancers in the fediverse. Keep it fairly clean please. This arcology is for all who wash up upon it's digital shore.