Post #84428

2025-12-19 14:52 UTC

A good piece on how GenAI is flooding the field. I too have worked with ML for a while and feel similarly. "Having done my PhD on AI language generation (long considered niche), I was thrilled we had come this far. But the awe I felt was rivaled by my growing rage at the flood of media takes and self-appointed experts insisting that generative AI could do things it simply can’t, and warning that anyone who didn’t adopt it would be left behind." https://www.technologyreview.com/2025/12/15/1129179/generative-ai-hype-distracts-us-from-ais-more-important-breakthroughs/ #AI #GenAI #ML #CompSci

Replies (6)

@veronica@mastodon.online 2025-12-19 15:09

I think we're seriously due for less marketing hype and a more sensible and grounded academic discussion of AI and Machine Learning in general. Some of the more extreme hype on AI is bordering on superstition and has religious overtones. I suspect some of this deliberately exploits the tendency of humans to anthropomorphise anything that mimics human behaviour.

Open ##1516820
@jaredwhite@indieweb.social 2025-12-19 15:37

@veronica “chief ethics scientist at AI startup Hugging Face” 🤨 Yes, I agree OP made some good points. Unfortunately I couldn't get past the breathless lauding of even the predictve AI we have which in some cases is very helpful (mainly personal data analysis & search) and in other cases horribly anti-social (like sentiment analysis, mass surveillance, etc.)

Open ##1516833
@DanielaKEngert@hachyderm.io 2025-12-19 15:50

@veronica Thank you for sharing this article! My diploma thesis (in EE sic!) in 1988 was about a classification problem. It used all computing power I could get hold off to run reasonably-sized neural networks and train them. The results were disappointing. Nearly 20 years later, I applied a ML-based clustering algorithm to learn all currently detectable radio emissions from low short wave up to multiple GHz (a SIGINT system on a submarine). The results were good enough to meet expectations. Nearly 20 more years later, such useful ML systems are utterly conflated with overhyped Gen AI, or the fabled AGI. Too many people can't tell those apart, thereby enshittifying our world.

Open ##1516836
@vla22@tech.lgbt 2025-12-20 12:36

@veronica I should have guessed the article was by Margaret Mitchell! Thanks for bringing it to our attention

Open ##1516837
@godofbiscuits@sfba.social 2025-12-20 14:05

@veronica @nicklockwood does it still steal from everyone and use up electricity like a monster?

Open ##1516838
@_michaljan@mastodon.social 2025-12-21 19:30

@veronica saved for later! As a little apropos, this website is a great example of #enshittification -- I had to close 4 pop ups to start reading!

Open ##1516844