Epiplexity

Sign in<br>Subscribe

Tim Ferriss introduced me to Jim Collins, the man behind one of my favorite business classics, Good to Great. They discuss how Jim dissects his behaviors, literally calling himself a bug to be studied.<br>i prefer Jim's affectionate version over AI doom and alien invasions, don't you?Blogging has created my own "bug book" of sorts. The latest surprise result to announce is that I'm very likely going to get a job again after 6 beautiful years of funemployment, aka founder journeying.<br>To contextualize the significance of this new chapter, I'd go back and revisit two other key turning points of similar magnitude in my professional arc:<br>Landing my first real job at Scale AI. After college, I bummed around for a year—as data officer for a charter school, a math tutor, chess coach, and Uber driver, then a contract worker for Namebase and Clever for about 2 days each. Scale was the first role in which I felt true fit as a professional and it led to immense personal growth and learning as well as all subsequent opportunities and my current, hopefully highly employable, résumé.<br>Funemployed founderhood culminating with Shuffle. After covid lockdowns decimated San Francisco into something worse than a ghost town, I bundled my geographic angst with my desires for entrepreneurial success and accidentally resigned a cushy tech life for a solid year and change. Thus my payments startup in Ecuador was born, failing gracefully into a mild million-dollar success with Shuffle Dating reaching profitability on a national scale. During my tour of duty, we served tens of thousands of customers, many of whom got matches and a handful even ended up married. Yay!<br>In hindsight, both of these chapters were somewhat reactionary, spontaneous and unplanned. That doesn't make them any less lovely or wonderful, just that I had not yet matured into a strategic operator and long-term business thinker. I took immense initiative in fairly uncalibrated ways.<br>I'm still learning and growing, of course, but now I can more confidently 80-20 things and predict what matters on an 18mo-5yr time horizon rather than being limited to 18 days of frequent whiplash or 5-week planning cycles.<br>Last year July, I began transitioning away from Shuffle in earnest. As CEO at the time, I invited Austin to step back into running the business himself as we clarified a few fundamental dynamics:<br>the business had stopped growing and<br>neither of us was particularly excited to expend massive efforts pursuing the appropriate paths to take things to the next level, in large part because<br>leaning into AI presented the macro opportunity of a lifetime, and<br>splitting the net income was not sustainable or principled given our off-peak numbers. Austin needed the monthly income more than me. I could walk away more cleanly.<br>So walk away I did.<br>Towards what, exactly? AI in general, but in specific I was not yet sure. I enshrined my best-guess hypothetical solution to the infamous AI alignment problem at andytrattner.com/alignment and didn't enjoy interviewing as a Grok eng. I found the political and cultural aspects of AI most interesting, rather than directly building another business or launching a new product into this space.<br>I also knew I wasn't technical nor inclined enough to dive into the actual mechanics of machine learning, having already dropped out of Leslie Kaelbling's course and research seminar as an undergrad. I played with the latest chat toys and early command-line agents like Indent, moved back to SF, started discovering more about the developing AI ecosystems. I tried contributing with &U grants, Capitalism Unlocked content on X and Youtube, zScore.info, etc.<br>This may have seemed like a random smattering of activities from the outside, but for me it felt like a reasonably principled experimental progression from the highest-level differentiated macro ideas I could distill to tactical instantiations of how I might execute some kind of master plan to ultimately generate value for society at scale. I have yet to see genuine evidence that my Future of Loving Grace essay is significantly off the mark, long-term.<br>So why am I getting a job now? What went wrong with my self-employment plans?<br>For one thing, I haven't made much income at all since leaving Shuffle last year. My Future of Loving Grace isn't as actionable as Leopold's Situational Awareness! (and or i'm less skilled at action lol)<br>More broadly, as I wrote in this post's companion substack essay, moving to Greenville earlier this year represented a shift in my personal life and mindset towards stability. With relationship and geography constrained to the good, my foremost and final frontier towards full-on adulting is a sustainable bank balance.<br>I've also developed a deeper understanding of how history hinges on products, as they are a physical manifestation of change. Essays are OK, and teaching people directly is always wonderful, but great work gets recognized measurably more than...