I guess given posts by Gowers like this in 2025, the 2026/2027 assertions don’t seem, to my naive intuitions, that implausible?

Obviously each field is different (eg biology may require wetlab data?), but directionally this feels feasible, so I’m curious why you think it’s likely not the case.

Can you explain why you think their timelines are too aggressive? Which of their assumptions do you think are wrong?

Gemma

Do you think his Canada approach is net positive for him?

Not at all clear this disproves the political strategy maxim. If anything, it highlights why Dems should say whatever it takes to get power, and then hash out policy once in office.

Why cover tech related topics for the @washingtonpost.com if you have general animus towards the tech industry and paint with this broad brush?

Are you truly able to call balls and strikes as you see them?

I think you can safely skip the “for a sense of scale…” thing when you’re talking about $81 trillion 😛

The reason is that the information ecosystem is warped and unbalanced in favor of those who want to dismantle (pointing out everything each round of “elites” are doing wrong) vs those who want to build. It’s always been harder to build institutions, but never by this much.

It works fine now, but there’s somewhat lesser engagement, and left wing politics dominates the chatter (though a decent number of AI academics on here posting about research etc)

Agree.

(And separately, I wish the AI community on Bluesky were a bit larger and enthusiastic about this sort of stuff!)

I had to log back on to Twitter to believe this was true (it sadly is).

I say the same thing, though, to people who say they're "just" into removing illegal aliens: if that's what you believe, you haven't seen the people in power recently.

Difference is that you might need a seat on the plane to go to some other experience you enjoy / need, but why go to theme parks if the base experience isn't worth it? Especially when the the number of competing entertainment options only increases.

I’m disappointed by how so many of my fellow leftists are so reflexively against new technology and see it as a threat (vs something to debate). You see it the replies to that post.

If pretty much anything about the future and about change scares us, we’re always going to be the losing side.

This is great, needed to be dug into.

I’m an amateur coder and I can vouch firsthand that these models are as frustratingly lacking in basic common sense as they are brilliant in spitting out code I otherwise couldn’t write.

Agreed. Policing language seems like the important thing here.