The fascination of HN for Habermas is really astonishing. According to Algolia we are going towards 20 posts (https://hn.algolia.com/?dateRange=all&page=0&prefix=true&que...), more than three times the number for for Tony Hoare. That's actually quite concerning.

For me (60 too) it's both, the journey and the destination. LLMs not only help me get around the boring stuff so I have more time for the things I really want to design and build, but they also open areas for me in which I always wanted to go but for which it was very time consuming or difficult to get the required knowledge (e.g. tons of academic papers which are neither well written nor complete and sometimes just wrong). The LLMs help me to explore and find the right way through those areas, so these adventures suddenly become affordable for me, without doing a PhD on each subject. And even if the LLM generates some code, it still needs a "guiding hand" and engineering experience. So for me, no, AI doesn't kill my passion, but offers a very attractive symbiosis, which makes me much more efficient and my spare-time work even more interesting. I find myself watching fewer and fewer streaming videos because exploring new areas and collaborating with the LLM is much more entertaining.

Sure. I could delegate the boring stuff to an agent and focus on what I'm interested in. It would make me more productive, if I could e.g. just let the agent generate an optimizer and code generator for any required target for my intermediate language. I have not yet found a system though which can do that without significant effort from my side; currently I don't seem to spare time; I just spend it differently (controlling the agent and getting aroung its weaknesses).