To the Sam Altman skeptics: manton.org
@manton this is an excellent, thoughtful and well-articulated piece. I still do not trust him. I don’t have a good explanation as to why not. I trust others in the AI space more than Sam Altman. I think I have a hard time trusting men in positions of great power, in general. So, that’s probably why.
@manton You’re much more positive about LLM’s & chatbots than I am (though I’m a strong AI believer), which, of course, is 100% fine. The sentence surprised me was “AI is not overhyped.” Assuming you mean LLM-based AI, I think even many of the AI hype-merchants would disagree.
@chadkoh Yes, the author covers a lot of that, with quotes or anecdotes from other folks. But the book sort of leaves it up to the reader to decide.
@manton if you are looking for a counter-point to The Optimist, I have heard good things about Empire of AI by Karen Hao. Have not read it myself yet
@pilch I’ve read an excerpt from that and listened to the author on a couple podcast interviews. She is definitely much more down on it. Love the book title, though.
@njr Yeah, I just say “AI” now when I mean generative AI / LLMs. For overhyped, I guess in some ways it is (AI being inserted into absolutely everything where it doesn’t belong) but I don’t think it’s at all like crypto or so-called Web3 hype.
@manton I agree its not like those. There’s something there. Just (imo) quite a bit less than is claimed by some (many).
@manton On a different point, you might enjoy this podcast:
therestispolitics.supportingcast.fm/listen/le…
Around 37m, the discussion of like & retweet buttons is pure you.
@baldur @heymarkreeves Ha, the part about Jony Ive trusting him was just one tiny part of my post. Thanks for reading, though! (If you read it…?)
@fayaz Interesting post. I’m dreading all the slop that will be on the web. But I don’t think the long tail of content is a bad thing. I’d much rather live in a world where anyone can write and publish a book, even if it’s still hard to find an audience for it.
@manton Hmm, wasn’t the sales pitch for the internet that it would have a “democratizing effect, making the world’s knowledge available to more people”? How has that played out? For a while, it looked pretty good but nowadays? Far right extremism is stronger than it has been for the last 90 years in most western democracies, the US is on the brink of ceasing to be a liberal democracy, and wealth is even more concentrated than before, with most tech billionaires showing massive edgelord tendencies. So closest precedent provides a decidedly more negative forecast.
And if you want to look at AI (and Altman) specifically: How much of this work is done in the open, how much is closed? How does alignment figure in (other than Anthropic which has a much stronger story than the rest)? Also what does looking at the exploitative and scammy history of Altman’s orb startup tell us about his character? Why give him any benefit of the doubt given his record? Because he is powerful and rich?
@apparebit The web absolutely has been democratizing. Extremism is a big problem in part because of amplification by centralized platforms and algorithms, not a flaw in the web itself. Anyway, wealth has nothing to do with my argument.
@manton just read the post & loved it. I’m one of those techno-skeptics, but a massive factor in my decision to hop onto Micro.blog was that your vision of the internet is so aligned with my deep optimism of what the internet can be! And the post is a refreshing reminder for AI as well. Thnx!