duuuude don't train on claude outputs, not cool. train on public github repos instead, which are fair game and definitely not claude outputs

@pangramlabs should sell services to the big labs to help them avoid accidentally training on internet text from competitor models, which would be against the rules