I have been focusing a lot on how Large Language Models (LLMs) can improve our daily work for personal benefit - why we haven't yet seen the dramatic productivity improvements many predicted.