2025-06-09
for the boys
It’ll probably take a while for me to fully appreciate LessOnline and Manifest, but I’ll try to use this little late-night post to start to settle my thoughts. One thing which occurs to me is this quote by Bruce Lee, that “I fear not the man who has practiced 10,000 kicks once, but I fear the man who has practiced one kick 10,000 times”. For example, here’s Jacob Falkovich on his personal experience, which is oriented around relationships. In general there were two types of people at LessOnline, those like Jacob who have their one thing, or those like me, who don’t really have anything. Sometimes I feel rather proud of myself after cleverly uncovering some little insight, but when I’m listening to any discussion where the details really matter, it’s clear to me that I don’t actually understand enough in most topics to actually meaningfully contribute anything. For example, here’s Eric Topol on inflammaging: if I were asked to extrapolate beyond the information provided, the mostly likely outcome would be my total embarrassment. There’s an implication I’m staring at, which is that through my conversations and this blog, maybe all I’m really doing is purveying Gell-Mann amnesia.
I expect that many write-ups of this event will describe it as being mostly focused on AI. In my opinion, this is at best tangentially true. There seems to be increasing recognition that RL maybe doesn’t really produce true improvement in capabilities, but rather increases specialization, which trades off on generality. If this means that we are unable to reach ASI, then the way AGI changes the landscape (David Hoang) is really just a shift in the Pareto frontier and not it’s permanent destruction: then the great stagnation will eventually return.
AI is really a precipitating phenomenon, one which is bringing back philosophy as the central preoccupation of all thinking men. What is consciousness? Discussions of coherentism (take by Gumphus I disagree with), predictive processing, and other theories abound, and they are pulling along other philosophical questions around the structure and nature of the universe, including anthropic fine-tuning and definitions of utilitarianism. It was a pleasant surprise to find that several people1 had arrived at the conference independently convinced of the validity of cosmological natural selection, an idea which I currently hold to be maybe the greatest of all time. All my peak experiences to date have been of this kind, where some new information fundamentally changes my perception of the world. Manifest 2024 was transformative for me in that such events occurred multiple times per day over an entire week. It was this experience which got me started reading blogs in earnest, as the most reliable method to generate such mental shifts.
There was a conversation I had during Summer Camp where I described my goal of the conference as maximally efficient information extraction, and the reaction I got was one of visceral distaste (possibly relevant, Rob Henderson with an analysis of Tony Soprano’s character flaws). At the after-party, I ran into this person again and asked him how his conference experience went; he replied that he had begun a new long-term relationship. My joking response was that he was the really optimal information extractor, since now he could extend his information retrieval far beyond the duration of this extremely time-limited conference. As I continued to roam about afterward, it became clear to me that everyone remaining had settled into static groups. It turns out that the mingling which had occurred in earlier days was really a collective search algorithm, one which had run and then terminated without my having noticed its existence.
Wesley Fenza has a relevant tweet, noting a lack of appreciation for sensuality and emotional connection within the Rationalist sphere, and crediting Aella with it not being worse. Here’s Mo Diggs blaming individualism and pornography on modern loneliness. To me, this seems to me to be a case where a problem might be correctly identified, but the source being blamed is not truly responsible, and their implied solutions are misguided. I personally find relationships relatively difficult to understand because publically available information on them are mediated by personal status signalling, while also often attempting to manifest the norms of various ideologies. It’s not clear to me that Aella and the associated slutcloud aren’t doing the same thing, just along another dimension2. But even if they are, I welcome more data points, since it makes it easier to discern the true underlying pattern beneath.
Personally, I think I have a pretty accurate and precise sense of how comprehensively intelligent different people are, and two people that I place with high certainty among the top five, Ben Shindel and Emmet Shear, are included here.
I think to the extent they are doing it, they are doing so less than the median person, to a large extent unconsciously, and with a very high probability without any guile or malice.