Twelve theses and predictions on "AGI" (falsely so-called)

2 wyclif 1 5/1/2025, 6:42:58 PM programnotes.blog ↗

Comments (1)

techpineapple · 19h ago
I wonder if one of the reasons thought leaders like and are bullish on AI is that one of your jobs is to simplify things. Act at high levels of abstraction. The details don’t matter. Anyone who has ever worked with an executive knows that you’re trained to talk to them almost like kids (especially outside of their domain expertise) because they’re trying to manage so broad a context.

And so AI is great at this.

So the things that may bother an expert, like the fact that AI hallucinates, often seems to have no idea what it’s talking about and lacks nuance doesn’t bother them. In fact the things that bother most experts about AI are anti-patterns to like VCs. They literally can’t see how AI may be exceptional at helping them but not others.

Like the story about how SBF hated reading books, everything should be a like 5 bullet point blog post