@Armavica @ploum the creator of Claude made a papel of Reinforcemenf Learning trough Human Feedback. You guys doing “fix this, solve this, solve that” are just paying to make THEIR product better adjusting the model so it fits in more use cases.