Any deterministic system is hard to defend if models can hillclimb fast enough and generate them. It was several years ago when I first considered this possibility, and it’s been consistently true since then.
The gap is widening because a tiny group of people sits on frontier systems with near infinite quotas, treating capability discovery as a full-time job developing skills, intuition, and applications before anyone else can.
Dusting off this tweet from April because this gap in shared understanding of LLM capability is *widening*. It's now less "two groups of people speaking past each other" and more a sharp funnel.
- Napkin math somewhere around 6B people (~75% of the population) have barely come
I was thinking this could be explained by woowoo stuff until both Stanford and UCSF diagnosed me with Stage 1 cancer. Your brain truly denies that you can be sick in early stages. "What do you mean I have a survival change of 95%?? I'm feeling 100% okay."
“Look at me, I’m so smart. So smart, in fact, that I ignored my doctors when they tried to get me to undergo cancer treatment in 2003 and I insisted on alternative medicine instead.”
