[sisve]

Do you think we will get a self-improving model in 26 or 27? Maybe not a native one but some kind of hack so a model will learn something without loosing part of the context window?