Gregory's ServerFor more than a year, policy makers have been worried about the consequences of AI getting too powerful.
But it’s time to start worrying about the consequences of AI staying as dumb it currently is.
My latest for NYT Opinion (gift link):
 7 comments
  @Julia I'm tempted to talk about the real AI threat being corporations, but this probably isn't the space for that.  @Julia I don't know if the AI investment bubble *will* burst soon, but I 100% believe it should. It is a powerful technology when applied to the right problems in the right ways, but imo those make up a vanishing fraction of the ways it's being used today. (Side note: I've been ruling out so many job postings for use cases I don't trust are good applications of AI, y'all. So many things I know how to do but think shouldn't be done in the first place. It's painful.)  @Julia Those of you in the anti-AI league are on the losing side. Painters wept inconsolably for the loss of romanticism when photography was discovered, etc. etc. etc. (give thousands of examples here). And the funny thing is: eventually everything will fall into place. AI is here to stay. It is wonderful. You have to adapt or get screwed.  @Julia Well said. One thing AI may be useful for is finding out which memos and documents we don’t actually need, because AI can both write and read them. If we learn to stop wasting time on those, the effort will not have been wasted. |
@Julia Thank you, this is really good!
I hope very much that we also address another issue at some point: there are authoritative structures that we could be using to build better search, and we're just not! There ARE alternatives!
(I've written essays about this but my essays don't get into the NYT, lol)