Sam Altman expects AGI any year now? Like, possibly 2023 or something? That's interesting.

"Decades and decades" was a pretty reasonable sentiment in 2016, I think. I myself probably would have expected AGI in either the 2030s or 2040s back then. But now... nah. AI has advanced too much during the 20s, already reached proof-of-concept levels of sophistication and generalization, and each consecitive year makes a bigger difference than the last. It's just... mathematically impossible at this point to have 7+ major leaps forward and not end up with AGI. The gap between modern AI and AGI is not large enough to have seven years on par with 2022 and not end up with AGI (and future years won't be "on par," 2023 will make more progress than 2022, 2024 more than 2023).

Anyway though, apologies to CriminalizeGolf. It's unfair of me to be an asshole when he was perfectly respectful and polite. I'm just fractious after 11 years of dealing with tens of thousands of skeptics and "realists" who are snotty and condescending.