i transcribed every journal i've ever written along with this blog and stored it in a giant-context LLM conversation thread.

when i began asking it questions, it was like staring into an ultra-8k lossless mirror.

after the awe goes away, and you've finished asking the mirror shallow questions in hope to flatter yourself, you roll up your sleeves.

this thing could run complex multi-step analyses and report with confidence intervals.

if you're not familiar, one of the most intriguing results of LLM's is that they serve as sort of a connector between straight text (tokens) generation, reasoning steps, and ML tools. Machine learning has been cooking for decades. but it was never quite accessible. it would take a team probably 8 months to train a model for image recognition. now, you unbox it out of the OpenAI API. and a realistic sounding friend can do the analysis.

but enough of that. point is, i could now run complex psychological analysis at a large scale on all of the things i've written by hand or typing.

what i discovered, was.. actually surprising.

i like to think i'm mentally tough, (and the results are actually corroborating that) but when you deconstruct mental toughness into its components, you find that the actual answer of how mentally tough you are is nuanced.

in terms of resiliency, the measure of how fast it takes for you to return to baseline after a negative event, i'm ranked somewhere between 9-10/10.

in terms of fortitude, the measure of how hard you fall after a negative event, im ranked 2-3/10.

in terms of adaptability, the measure of what baseline you return to after recovering, im ranked 11/10. this also means that each time i get knocked down i actually return to a higher baseline. hard times beats strength into me.

this would actually quite track with the story of this blog.

there are ways to boost your fortitude. for me, one of the main things will be decoupling all of my identities and my ego. i, about as anyone else, have around 5-6 identities but they are all intertwined with each other, and my ego. so when one identity takes a hit from a negative event, all other identities including my ego go offline. in systems design, you would have never designed such an important system to be this messy and spaghettified. each component must be responsible only for itself, so a failure of a part never equals a catastrophic collapse of all other parts.

i think this is by far the highest leverage move i can make to increase fortitude. increase the number of identity bets running in parallel.

curious enough, it looks like i've actually gotten good at reinventing myself (ie. adaptability 11/10), so i could deconstruct how i've done that and then do it all at once across all identities to basically give myself lives.