This is a key point on AI agents "Brains" matter for agents. As long as the scaling law holds, you get much better results out of using bigger frontier models as the main brains of your agent, though it might be able to delegate to lesser AIs Labs with the best model get the

→ View original post on X — @emollick