Reposted by Tim
One thing I didn’t have space to fit here: some people, on both sides of the debate, like to pretend that the only AI risk worth caring about is superintelligence. I think that’s wrong, and misses a huge swathe of ways that improvement in AI capabilities could be really bad for humanity at large.
![Himid has lived in Britain for more than seven decades, grown up here, been educated here, built a life and a career here
- and yet the question of what "here" actually means, of who gets to belong and who is merely tolerated, has never left her work. "We do a thing, Magda and I, if we're booking a table on the phone, I will always do it," says Himid, whose RP-ish English has little trace of either London or Lancashire. "But when we go to a place we don't know, I always want Magda [who was born in Poland and is white] to go first. She signifies one thing visually and I signify another thing audibly. Together, we are incredibly belonging and unbelonging."](https://cdn.bsky.app/img/feed_thumbnail/plain/did:plc:yhbp52uzo4vsn5hlkbrfvxoq/bafkreieix4etf3ysmqh3q7byirpq77phrqnu46njn466psddyk46qhi46i)
