Frontier AI labs already have the internal capability to do model routing themselves, so external multi-provider routing is mainly a short-term negotiating tactic on price rather than a durable long-term strategy.
Pash argues labs can route between their own models internally at will, so customer-side model routing is really just leverage in price negotiations, useful mainly because OpenAI and Anthropic remain close competitors. ✦ AI generated
Pash · The Cognitive Revolution · 2026-07-02 · original ↗
starts at this moment · 30:32
So I don't I don't I don't really know that, you know, model routing is going to be a long-term thing, but it is a negotiating tactic against against the firms. Uh, negotiating against the price. So, it's it's very helpful right now that we have OpenAI and Anthropic kind of close to each other. It would be horrible if we just had anthropic.
verbatim transcript · starts at 30:32
30:32an optimization which is not necessary right now but the moment it is they they think it is necessary they will do it. Um, so I don't I don't I don't really know that, you know, model routing is going to be a long-term thing, but it is a negotiating tactic against against the firms. Uh, negotiating against the price. So, it's it's very helpful right now that we have OpenAI and Anthropic
30:54kind of close to each other. It would be horrible if we just had anthropic. Um, we would we'd all be suffering terribly. So tactically, for what it's worth, I like to start with the smarter model and have it delegate to the lesser model. I've been a little confused by the idea that you'll start with a sonnet or even a haik coup and have it route up. I could see that working.
31:23I think in so many cases that there there's always this kind of question of how much are you going to control and really map out the structure of the work that you're trying to do? How much you going to control the inputs and how much are you going to map out that structure? Going back a couple years, I think it was already possible with even like a
31:46GPT40 with fine-tuning to get virtually all routine tasks done on an AI workflow basis. You know, if if you're a company that employs people to do roughly speaking the same thing over and over again, I would bet that with 40 and fine-tuning, you could get human level performance on a large large majority of those tasks. So already I think we've had for a long time the ability to if
32:21you know you control the environment and if you map it out and if you're willing to do the evals and the prompts and the fine-tuning, you could get there. Obviously, the hurdles there are high. The hurdles have been brought down with smarter models. What do people really want to do? I want to have a smart kind of general purpose assistant that I can throw anything at and then have it intelligently decide