Anyone else noticing that cutting edge models (5.6 Sol Ultra, Fable 5 Ultracode) have a tendency to waste massive amounts of tokens on the wrong tasks? Experts at missing the point?
For example in one instance, an agent wasted 1b+ tokens building a perpetually re-optimized
Just for posterity the other side of this argument is:
1. The critical teams from these labs are still colocated in the bay. Same as it’s always been - Google keeps a presence but the important work isn’t done in nyc. Choosing to live there means opting out of exciting projects.
For years, the advice was to move to San Francisco if you wanted a career in tech. That advice is now wrong. New York has become the most exciting place in the world to build.
A few reasons why:
1. The AI labs are here. OpenAI, Anthropic, Google, and hundreds of others all run