Meta tags:

Headings (most frequently used words):

daniel, paleka, recent, posts, contact,

Text of the page (most frequently used words):
2025 (27), nov (24), the (11), you (8), and (6), posts (5), are (4), for (4), now (4), about (3), should (3), research (3), writing (3), what (3), #daniel (3), want (2), out (2), can (2), like (2), email (2), talk (2), your (2), takeoff (2), long (2), first (2), things (2), questions (2), underrated (2), more (2), recent (2), see (2), working (2), paleka (2), reach, anonymously, use, this, form, echoing, getting, something, send, message, twitter, standing, invitation, contact, all, jan, delay, engineering, heavy, light, automation, mar, gpt, draws, itself, consistent, type, guy, jul, memetic, optimization, brainrot, testing, could, lead, llms, retain, users, instead, helping, them, why, days, some, hidden, time, sinks, process, going, get, priced, best, coding, tools, publish, stack, slow, fast, everything, else, prioritize, base, maslow, hierarchy, two, types, llm, preferences, immortality, reproduction, incompatible, term, future, how, well, ais, not, that, important, right, inkhaven, retraction, obvious, ways, fail, hated, but, also, just, try, reinforcement, learning, contra, dwarkesh, sample, efficiency, via, information, theory, deal, with, forecasting, apocalyptic, residual, secret, third, way, save, world, pick, basketball, maximize, benefits, minimize, risks, board, games, when, don, know, even, superhuman, forecasters, only, good, reasons, thankful, edition, invoke, shoulder, people, month, sleep, intervention, experiment, public, still, fewer, page, trivia, started, phd, advised, september, 2022, security, failure, modes, artificial, intelligence, papers, florian, tramèr, newsletter, publications, setup, vibes, links, home,

Text of the page (random words):
daniel paleka daniel paleka home links now vibes setup publications newsletter hi i m daniel i started my phd advised by florian tramèr in september 2022 working on security and failure modes of artificial intelligence see my recent papers and what i m working on now see the trivia page for more recent posts fewer posts more posts writing in public is still underrated nov 30 2025 my month long sleep intervention experiment nov 29 2025 invoke shoulder people nov 28 2025 underrated reasons to be thankful ai research edition nov 27 2025 even superhuman ai forecasters are only as good as your questions nov 26 2025 board games are for when you don t know what to talk about nov 25 2025 pick up basketball to maximize benefits and minimize risks nov 23 2025 on the apocalyptic residual or the secret third way to save the world nov 22 2025 what s the deal with rl and forecasting nov 20 2025 contra dwarkesh on rl sample efficiency via information theory nov 18 2025 16 questions about reinforcement learning nov 17 2025 things i hated at first but now like and also things you should just try nov 15 2025 obvious ways rl can fail nov 14 2025 inkhaven s first retraction nov 13 2025 how well ais do ai research is not that important right now nov 12 2025 immortality and reproduction are incompatible in the long term future nov 10 2025 the two types of llm preferences nov 9 2025 prioritize the base of maslow s hierarchy nov 8 2025 slow takeoff for ai r d fast takeoff for everything else nov 7 2025 you should publish your stack nov 5 2025 you are going to get priced out of the best ai coding tools nov 4 2025 some hidden time sinks of the writing process nov 3 2025 why i am writing 30 posts in 30 days nov 2 2025 a b testing could lead llms to retain users instead of helping them nov 1 2025 memetic optimization 1 brainrot jul 10 2025 gpt 4o draws itself as a consistent type of guy mar 31 2025 you should delay engineering heavy research in light of r d automation jan 6 2025 all posts contact i am echoing the standing invitation i like getting email if you want to talk about something i do send me an email or message me on twitter if you want to reach out anonymously you can use this form