I read Rodney's past predictions on LLM's. There's not a _single_ one that was high information. Not a single one that made me go "hey this was actually a good call, it was very hard to catch that in hindsight".
Almost all his predictions were low signal, just correcting the hype and doom from extremes.
If I'm wrong, please show me one thing he said (almost everyone has one thing tbh but lets see) that would make everyone go: yeah this guy had foresight.
Edit: he made some directional claims on LLM's: that they were so hyped that OpenAI was not even worth the $30 billion valuation in 2024[1]. It is now worth ~$1 trillion. For a pundit who puts his reputation on the line on single falsifiable claims like this, he has a really bad prediction pattern. Like, the one high signal claim he made turned out to be 25x incorrect!
What annoys me is that this archetype of anti-hypers had ONE job. That is to get the ONE hype correct. But they still miss that. Anyone can keep correcting the hype and they would be right 90% of the time.
[1] https://spectrum.ieee.org/gpt-4-calm-down
Almost all his predictions were low signal, just correcting the hype and doom from extremes.
If I'm wrong, please show me one thing he said (almost everyone has one thing tbh but lets see) that would make everyone go: yeah this guy had foresight.
Edit: he made some directional claims on LLM's: that they were so hyped that OpenAI was not even worth the $30 billion valuation in 2024[1]. It is now worth ~$1 trillion. For a pundit who puts his reputation on the line on single falsifiable claims like this, he has a really bad prediction pattern. Like, the one high signal claim he made turned out to be 25x incorrect!
What annoys me is that this archetype of anti-hypers had ONE job. That is to get the ONE hype correct. But they still miss that. Anyone can keep correcting the hype and they would be right 90% of the time.
[1] https://spectrum.ieee.org/gpt-4-calm-down