Claude code is already amazing. It can handle more or less complex tasks, and even if it fails at first, it keeps iterating until it’s done. However, if you ask it to do the same high-level task again later, it will not have learned from its previous work. This will only be a matter of time before the model “learns skills” while performing work that it can later replicate. That alone will unlock another explosion of capabilities. Exciting and scary.(koeppelmann.eth 🦉💳)