Why Meta and Twitter’s AI and ML Layoffs Matter

venturebeat.com

4 pointsjensgk3 comments

From what I can tell, they laid off people on teams that were doing fluff (from the perspective of the CFO)- things that don't contribute to the bottom line of products making money and mainly exist to help the company maintain its repuation for privacy, security, technology or whatever.

I've seen a few teams like this before and usually they don't really contribute to moving the ML field forward but mostly set themselves up as gatekeepers who make it hard to launch products (for better or worse- I certainly know if I'm an eng deploying ML I don't need some external team to tell me to consider the larger implications).

I can see that argument for the Twitter team but does it apply to the Facebook one too? Is it just that the research they’re doing is so far from being productionised that it had no real value to management?

It's hard to say from the outside; certainly at Google there were teams similar to Probability that paid their way (Sibyl was a good example) but there were other teams that had lofty missions but zero impact. My guess is this team was probably at the very bottom of the list for ML impact. From looking at their various online literature they look a bit... theoretical?

Reply on news.ycombinator.com