my ikigai
We often learn structure only to unlearn it later. Here are a few false dichotomies: academia vs industry, work-life balance, and category theory in algebraic topology vs the more standard word "categories".
You don't stop learning just because you get a job. The best kind of work is the one that doesn't feel so tedious. Ideally your work is "life work" and becomes equally if not more rewarding than hobbies. (I'm not good enough at math to really opine on nomenclature, but you get the point.)
Names and labels become meaningless once you find your own labels that work for you. America was built on the principle of religious freedom for this very reason. I know great Muslims, Jews, Christians, and Atheists. I also know people throughout history who have done bad things in the name of those categories.
Who cares?
A problem for me personally is that I am allergic to labels. I've written before somewhere about how I suck at interviewing for jobs because I tend to tell long stories about my life, over-explaining why things happen, and therefore accidentally mis-answer certain questions.
As an aside: if you're interviewing me, shouldn't you care to actually get to know me? Be more like Chamath. Thanks.
One of the funniest examples of this was at Scale AI where eventually my boss Varun and I became close enough for him to tell me that they nearly rejected my interview. Specifically, he said I rambled a lot about being an Uber driver.
The role was tech support for Scale's API, the first so-called "customer ops specialist". When they asked what driving Uber taught me about customer service, Varun said I essentially conveyed the insightful "Don't talk to people who have headphones in!"
In hindsight, I probably was explaining too much about my idealistic aspirations to write a book and talk to random people across America as a taxi driver. Most people don't actually live like Chris McCandless, of Krakauer's Into the Wild. I come closer than others, at times. Life is an adventure, indeed.
I'll have to dig up the performance reviews someday, I was flattered Varun gave me such top marks despite my existential confusion.
Anyways, this blog post is about ikigai, the now-popular Japanese version of where Maslow's hierarchy points. Traditionally, ikigai is the threefold combination of what you happen to be good at, what you can be paid to do, and what you find rewarding.
(Asian people are pretty good at coping with capitalism! For the rest of us, there's always Seth Godin :)
While I was jogging this morning on the swamp rabbit trail (again), the above triangular diagram became clear. I think it's worth an explanation to myself and the world. I call such explanations, "blog posts". Nomenclature is fun indeed.
- Economic Output is the easiest category. The whole Japanese concept fits within it. It's work, regardless of your level of passion or sustainability. Regardless of profit margin. For regular readers, you are probably aware that these days, I have a moral outlook on capitalism as a whole.
- Embodied Life is also fairly straightforward. Exercise and fitness, diet, relationships, all that good stuff. Dopamine and fun sit here too. I often make the mistake of trying to force my whole life to be extra fun, which can be problematic, but hey—don't fault me for trying!
- Artificial General Intelligence is the most consequential and convoluted. I may have to give a better definition below. I don't have a specific target in mind yet, but it's certainly not the terminator. And it probably should not be thought of as a machine God. Trust me, I tried. Lol.
AGI sits at the top of my diagram because it is something I care about, increasingly deeply. For me, it represents a very broad philosophical lens (not just AI) through which I now see life, time, and the universe as fundamentally computational.
I think this AGI concept will define the human species from 2050 onward through the next 400 years, until 2500+ at least! These coming couple decades are particularly important to the future of civilization. Hence, why I'd prefer renaming AI to "Computer Future" as a broader and more precise umbrella term.
Growing up, I remember having to memorize then type my friends' phone numbers on land lines. Now, I barely remember my own digits! Even more than with cell phones, we will all have to figure out how we relate to AI. The technology is coming, and it redefines the very word "technology". Such things are part of the Computer Past.
My definition of true superintelligent AI differs from the standard views as well as sci-fi. I now think that AGI is nearly impossible to achieve because it is NP-complete in consciousness as a nondeterministic verifier layer. By that I mean each and every human alive at a given moment must form explicit consensus on AGI being achieved, what it means for the definition of the word "human" (the physical nature and scientific specification of consciousness, hence self-awareness), etc.
One way to conceive of this: it's possible we need the U.S. Constitution to be modified, as a final marker that cultural change has truly landed in the institutions of greatest long-term historic impact. The dual proof of legal systems and common English parlance updating at the root would perhaps suffice to show "strong AGI" in a way no technical definition can.
The most likely way the above happens is the same way most long-term change happens for human society. Namely, all 10 billion of us currently alive die naturally within the next 100 years. So people being born in 2027 onwards inherit the globe and more modern definitions, an upgraded culture, new English with math baked in.
(You eventually have to take this view, regardless of what Mark Zuckerberg thinks about and tries to turn into "personal AI". By the way, I'm now bullish on META stock at current prices, especially given Reality Labs oculus + ctrl bracelets + APPL decay.)
It's funny that human upgrade cycles are on the order of decades and centuries, whereas it only took the industry a couple years to bake in math. Now, the AI systems are much better than simply plugging Wolfram Alpha into the original chatGPT. For me personally, it took roughly 3.5 years from chatGPT's launch to fully lean in and become a vibe coder as of late February.
This is especially strange as someone who was in (tiny) part responsible for training GPT way before it could chat back in 2019. But for now at least, it takes a full-time effort to really steep yourself in the vibes. I began full-time work from last year July when I left Shuffle and started thinking seriously about AI and Altruistic Alignment. This included interviewing last year with the team behind Grok.
Returning to the diagram and concluding this blog, my little triangle above represents my current best attempt to compress and distill all my experience into an entertaining reflection and useful mental model for myself going forward. No AI was used to write this aside from drawing the diagram above, it's pure Andy fingers on keyboard.
My specific form of self-actualization stems from my turning 30 blog post in which I linked but didn't fully discuss Patrick O'Shaughnessy and his principle for living. He has distilled and simplified for himself a beautifully elegant mental model of who he is and what he does. It frees him from the tyranny of schedules and to-do lists. He just likes meeting people, getting to their core, learning for its own sake, and uplifting others (slash investing in them).
I embrace my shedding of all structures, which is quite challenging to do in practice. I've tried a lot of schedules, a lot of productivity apps, a lot of to-do lists. Many of my friends and former classmates are historically hacker-types who often make such things for themselves. I've dabbled but never come away quite happy with anything other than blogging, journaling on the internet, as my best form factor for ongoing thinking and metacognition.
I have hacked broadly: I've tried a lot of jobs, relationships, countries... various ways to inhabit our world. I'm grateful for the path I've walked. Each day was a stepping stone to the present. I've written about this previously.
But I have not really stated how most people, in my opinion, fail to reflect deeply enough on their own identity.
This is my bias. I know people like my college buddy Edward who very simply just want a family. I admired his ability to balance other competing constraints and interests against that overarching goal in life. Also my best friend Lucas, who superficially may appear to naïve previous versions of me to lack depth, but in reality he probably has a stronger mental model than me when it comes to grasping what matters day-to-day.
As someone who lives in his own head, as both a blessing and a curse, I have a deep appreciation for how time is all we really have. It's the only thing, the non-renewable resource this universe crunches and burns away with each tick of the clock. Problems happen when I start to ruminate more than act.
Regardless of compounding impact, of personal wealth, of ozone layers and fossil fuel emissions, of love... how we focus on each day matters almost more than what we actually focus on. And this is the only thing we can really control as the root level input between our brain and the world it inhabits. Tick tock, not Tiktok.
I have drawn the above triangle as an aspirational pointer. I hope each day I can map every second and minute of my lived experience to one of those 3 categories, in a positive flywheel that uplifts me and my future, as well as all those around me.
I used to think I was quite good at embodied life in general. I didn't realize the AGI part of my ikigai needed to sit at the top! It took a long time and a lot of full-body efforts to discover the correct contours of this spiritual aspect and awakening (which andys.blog is originally from and deserves many more blog posts itself in the future).
I will fail. I am currently failing to generate income and have been since last July. But work is looking very exciting in the last couple months, and I expect the potential energy and skills I'm learning to translate into economic kinetics with the right discipline going forward. Living in Greenville is wonderful, as is my new relationship (more on that soon).
But I think my new little triangle will help.
appendix - image prompt for Claude chat
help me create a beautiful elegant simple diagram for my blog post. the idea is how ikigai (perhaps super simple triangle on plain light cream background) for me is now:
- agi = spiritual root concept (building something greater than myself which flows into the other two below)
- embodied life = normal living, girlfriend, exercise, health, etc
- economic output = productivity, work, money income, value creation (capture is secondary)
footnote - new Gemini likes this post in the way i most prefer
