ivan
@IvanVendrov
solve cooperation, use it to solve everything else. collective intelligence research @ midjourney longer essays at http://nothinghuman.substack.com
Von Neumann on existential risk: Progress is chaotic so investments in planning are much less useful than investments in better control systems. Don't build Maginot lines, develop better command structures. Shorten your OODA loop. Response speed is everything.
mine: very hard, worth it: math (concepts), writing essays, coliving, chatting up strangers, being direct, deadlifts, love very hard, not worth it: math (proofs), chess, learning musical instruments very hard, ??: meditation, inner work
very hard, worth it in retrospect: math, treating others courteously, some examples of helping friends very hard, not worth it in retrospect: thinking about morality
More concretely: - Many western govts are actively undermining democracy in the name of democracy (and most non-western govts act even more perversely) - Many academic fields are actively suppressing the most important insights from their discipline (e.g. they’re blank-slatists,…
visited goodreads. feeling a strange nostalgia for the early internet, the awkward teenage phase of social media. simpler times
What other human cultural data would be highest value to collect / OCR / speech-to-text so it can be part of our language models?
Insurance is an underrated way to unlock secure AI progress. Insurers are incentivized to truthfully quantify and track risks: if they overstate risks, they get outcompeted; if they understate risks, their payouts bankrupt them. 1/9
Ok so this is going to be a VERY nerdy thread, but this is an important foundational point on mechanism design that I still haven't seen anyone articulate after years of expressing it privately to people, and I don't have the time to write a paper on, so let me throw it out there
Tinfoil hat: are these acquihires an early reflection of collapsing Coasian transaction costs under AI - that the (application layer software) firm basically does not exist?
Hypothetical question, inspired by current events: if you’re a CEO and you tell a company: don’t acquire us, just hire me and my top people for half the price. That’s a breach of fiduciary duty to the shareholders right?
I find political discourse confusing because it tends to conflate 1. how power should be structured (institution design) 2. which people should make decisions (personnel) 3. which concrete decisions should they make (policy / values) (probably reinventing something obvious…
Aligning an AI system, or a recommender system, in isolation is playing whack-a-mole. The real issue we're facing is "full-stack", and requires solutions that tackles the problems on all levels
In 2017, I was working to change FB News Feed's recommender to use “thick models of value” (per the paper we just released). @finkd even promised he'd make Facebook “Time Well Spent”. That effort was thwarted by the (1) market dynamics of the attention economy, (2) the US…