Prof. Anima Anandkumar

Pretty sure DeepMind didn't do any PR for this - I think this is probably because I wrote up 'StreetLearn, in the newsletter a few weeks ago. It's unclear to me what you're requesting - any writeup of a paper should have three other papers referenced not from primary authors?

So I can understand being infuriated, but it doesn't seem like there's any kind of wizard behind the curtain here. Context is expensive, so we need to fix economics of journalism to get what you want, I think

Mark *is* trying to fix the economics of journalism by calling it out. Why push back against that?

(I don't understand how calling this out fixes this - it just drives economic benefit to a problematic article.)

Hmm, suppose enough prominent AI researchers all agreed to consistently call out such articles... you don't think that will create incentives for journalists to try something else? (The other economic issues notwithstanding -- I agree there are other issues at play.)

Maybe. It could also be a disincentive to covering the area, as you (if you're starting out) have an unpredictable risk of public criticism. It feels odd to me as I'm sure there's a way to improve things that doesn't require public disapproval, which feels inherently adversarial