It accelerates coding, but it does hit walls rather quickly and you’d need the adjustment phase where it’s just as challenging as when we hit walls before LLMs