hey chapo friends! Today Google fired Timnit Gebru, a highly regarded researcher studying ethical problems/biases with AI systems ... presumably for doing just that with AI systems that Google had already monetized. (Relevant: she's a highly successful Black woman in a very white male dominated field, who's also been fighting against the use of facial recognition among other things.)
This comes the same day as the Trump-appointed National Labor Relations Board found that Google had also illegally fired/spied on employees trying to unionize/organize last year.
Anyways, AI twitter is going wild but curious to hear y'alls takes. Personally I'm not too surprised. AI Technology has the potential to generate so much profit for these companies (pretty much as close as you can get to value in the Marxian sense), so there are strong economic incentives to silence anyone doing good work exposing problems with these systems. The only thing we can do as (tech) workers is organize.
We already invented a paperclip maximizer; it's capitalism
Except instead of paperclips it's electric scooters and bikes
Ya know that Canary in the coal mines, and how when it dies it means bad shit is coming. Well this is it.
The unionizing driver inside google started when they got a defense department contract. So ya can bet what kind of evil shit they wanna do.
it sounds weird but I'm kind of optimistic. Google has a very good reputation IMO in these AI research circles as a good place to work. Hopefully that illusion is gone now and tech workers are closer (maybe by a very small amount?) to building organized power.
another way of thinking about it: if Google wasn't scared they wouldn't have fired her. They're doing it out of weakness.
The only thing we can do as (tech) workers is organize.
Yeah, I think we also need to start having conversations about ethics.
Like if you work for Uber now, you're just scum. You're just pure scum and there is no excuse for it.
Agreed! But IMO we're having conversations about ethics, it's just those conversations aren't translating into what these companies are doing (bc that's not profitable). Or maybe that's just 'we' as certain AI researchers, idk.
Her twitter feud with Yann LeCun was crazy.
Her lecture at the Fairness Accountability Transparency and Ethics in Computer Vision tutorial (CVPR 2020) is good. Part 1, Part 2, Part 3 , Q&A , Timnit Gebru research, Black in AI
