Finally understand how Bittensor subnets actually function

The key insight for me was that subnets are not just categories, they are individual incentive economies with their own entry costs, competition dynamics and reward structures. Participating in a less competitive subnet with high demand for its output can be significantly more profitable than competing in a crowded one even if the crowded one is more famous.

This changes how I think about the Bittensor model as a whole. Network-level participation through TAO gives broad exposure to the ecosystem. But understanding which subnets are most in demand and least saturated seems like it matters a lot for anyone thinking about where to actually participate.

Do you evaluate specific subnets separately or just look at Bittensor as one ecosystem?