I think this is really fair pushback, thanks! Skeptical coverage of AI development is legitimate. I think the way I wrote this over-implied that these articles is a failing of journalism—the marketing hype claim is not baseless.
But I’m torn. I still think there’s something off about current AI coverage, and this could be a valid reason to want more journalism on AI. Most articles seem to default to either full embrace of AI companies’ claims or blanket skepticism, with relatively few spotlighting the strongest version of arguments on both sides of a debate.
Also, I think my core point stands without conditioning on object-level views: we need more journalists who can dig deep into AI development. More investigation and scrutiny from all angles would serve us better than our current situation of relatively thin coverage.
“Most articles seem to default to either full embrace of AI companies’ claims or blanket skepticism, with relatively few spotlighting the strongest version of arguments on both sides of a debate. ” Never agreed with anything as strongly in my life. Both these things are bad and we don’t need to choose a side between them. And note that the issue here isn’t about these things being “extreme”. An article that actually tries to make a case for foom by 2027, or “this is all nonsense, it’s just fancy autocomplete and overfitting on meaningless benchmarks” could easily be excellent. The problem is people not giving reasons for their stances, and either re-writing PR, or just expressing social distaste for Silicon Valley, as a substitute.
I think this is really fair pushback, thanks! Skeptical coverage of AI development is legitimate. I think the way I wrote this over-implied that these articles is a failing of journalism—the marketing hype claim is not baseless.
But I’m torn. I still think there’s something off about current AI coverage, and this could be a valid reason to want more journalism on AI. Most articles seem to default to either full embrace of AI companies’ claims or blanket skepticism, with relatively few spotlighting the strongest version of arguments on both sides of a debate.
Also, I think my core point stands without conditioning on object-level views: we need more journalists who can dig deep into AI development. More investigation and scrutiny from all angles would serve us better than our current situation of relatively thin coverage.
“Most articles seem to default to either full embrace of AI companies’ claims or blanket skepticism, with relatively few spotlighting the strongest version of arguments on both sides of a debate. ” Never agreed with anything as strongly in my life. Both these things are bad and we don’t need to choose a side between them. And note that the issue here isn’t about these things being “extreme”. An article that actually tries to make a case for foom by 2027, or “this is all nonsense, it’s just fancy autocomplete and overfitting on meaningless benchmarks” could easily be excellent. The problem is people not giving reasons for their stances, and either re-writing PR, or just expressing social distaste for Silicon Valley, as a substitute.