I decided to give it a go both ways and I found GPT a huge impediment with day one. It did a terrible job and I found it far easier to do myself. I think it's possible to finesse but it'll take more effort than simply "solve this." It didn't help that part two of day one is a fantastic example of a spec causing ambiguity through a lack of detail – I actually wonder if this was deliberate to throw LLM-based approaches off. (Note: There is zero threat of me ever being on the leaderboard as I will never be awake at 5am in December.)

See FAQ on using AI to get to the leaderboard: https://adventofcode.com/about#faq_ai_leaderboard.

And I wouldn't say day 1 leaderboard is surprisingly fast compared to other years. Time will tell, but I think LLMs will fall apart on hard problems. Typing speed will not be a limiting factor there.

The “learn a new language” is how I used it casually for a handful of years (and it’s great for that).

Last year is the first year I put serious effort into completing all of it (driven by a private leaderboard which made the accomplishments more personally rewarding).

If you’ve got a group of friends/colleagues who could use a little competitive motivation, consider making and joining a private leaderboard (free as in beer).