Comment by binarymax

LLMs don’t learn. They’re static. You could try to fine tune, or continually add longer and longer context, but in the end you hit a wall.