People have been rewriting history and saying that "everyone has always believed that LLMs alone wouldn't be AGI and that extensive scaffolding around them would be necessary". No, throughout most of 2023 (the "sparks of AGI" era) the mainstream bay area belief was that LLMs were *already* AGI, and that merely scaling their parameter count and training data size by ~2 OOM without changing anything else would lead to super-intelligence.
@AirCanada You said you apologize for the inconvenience of my almost 5 hour delay. I don't believe you. You lie. You're not sorry. You're laughing at me. Jokes on you. I'm flying @JetBlue next time
@edzitron Infinite energy to enable cheating on homework and lowering support costs. Good thing we've already solved for hunger, universal healthcare, war, etc.
@mattjohnstonca The national child care system and dental benefits plan provide tangible benefits to so many. It's sad that they're so rarely mentioned amidst the endless rage farming.