Back to fitness.

I know that a lot of people look unfavourably at AI (or what's currently being described as AI: large language models, or LLMs). I admit that there are issues that I have with it, too. My job is increasingly at risk due to AI and I think it's entirely likely that I will be made redundant within the next five years as a direct result.

And don't get me started on AI-produced "art".

But it can have its uses if you know how to utilise it. That's half the battle. Creating a good prompt can yield incredible results, but writing a good prompt is a skill in itself.

For example, I could ask Google about chronic back pain.  Which would give me thousands of results.

Or I could write a comprehensive prompt, which is what I did. I described my pain, how long I've had it, the fact that I've not yet had a GP diagnosis, my own suspicions about my sedentary lifestyle being a factor and it give me a result.

To be fair, it did advise going to see a GP as one of its first comments, but it also suggested that my theory about my inactivity could be a factor. I then asked if there were exercises that I could do to help ease or remove the pain and it updated its response with a 6 - 8 week plan. I then added that I had some basic exercise equipment and listed it to see if that would change things, and the LLM made slight adjustments to add the equipment after a few weeks.

And finally, I asked it to create a collated collection of its responses with detail descriptions of the exercises in a printable format for offline reference. And it did that, too.

I still have to contact my GP, and it's still hugely advisable to double-check what an LLM tells you (as should be the case for any internet search, especially medical) but a lot of the basic exercises are the same as those in my Yoga programme. A cursory search looks like there's nothing massively incorrect in what the LLM has given me, so that's something to get on with, I think.