What Bittensor is trying to do

This lesson stays away from money and mechanics and asks only what the network is for.

Bittensor pays for useful work without a company in the middle deciding who gets paid. A chain of computers keeps score: it tracks who took part, measures how much agreement there is about who did good work, and hands out the network's own currency, TAO, accordingly. [source]

The "work" is what the network's own documentation calls a digital commodity: a specific, measurable product — a prediction, a translation, a trained model — the same idea as a measurable good in the ordinary economy, just produced by software instead of a factory. The point of the whole system, in that documentation's own framing, is deciding how to fairly split TAO among whoever is creating that kind of value. [source]

Bittensor isn't one single competition. It is organized into many separate subnets, each running its own contest for a different kind of work, each with its own participants and its own rules for judging quality. [source]

Inside every one of those contests, some participants do the work and others check it, and the network pays both sides for doing their part well — who does which job, and how the checking actually happens, is what the next two lessons cover. [source]