Also, I the the problems of mental self-modification have some relevance to computer self-modification. How do you decide what an improvement might be? How do you evaluate the safety of a new method of change? I'm inclined to think these questions are actually somewhat easier for humans, and still very hard.

I certainly hope so! These are problems I have wrestled deeply with for much of my life, and they definitely SEEM relevant - so maybe people can learn from my failures as much as our host's successes. :)

Would you be willing to describe the early modifications you tried to make, and what you were hoping to get from them?

Absolutely. Assuming feedback continues to be positive on this post, I'll start my... er... 'series' with an attempt at a historical analysis of my prior attempts at self-modification - methodologies, successes, failures, etc.

Have you tried anything on the physical side? This would include psych meds, dietary change, exercise, and various ways (Alexander Technique, Feldenkrais method, yoga, etc.) of improving coordination. That last may sound weird, but I have some experience with the way habitual patterns of muscle tension are part of emotional and mental habits.

Yes, but I don't know how rigorously I've tried them - part of my problem is that it's very easy for me to dupe myself into BELIEVING I'm performing rationally, when I'm not - which is doubly frustrating because I have the math to know better; it's just expensive to compute in-the-moment. So I'll gladly include those in my first/'historical' report, but with lots of caveats of "I don't know if I did this right or not".