@cwebber Extra painful: bigger tech companies can afford to pay for plans that limit use of context data to train future versions of an LLM service's models so THEIR work is "protected" while their employees consume the commons. But smaller companies and individual users will be giving up their data.