--- Log opened Sat Sep 30 00:00:37 2023 00:02 -!- Llamamoe [~Llamamoe@46.204.77.176] has joined #hplusroadmap 00:06 -!- Llamamoe [~Llamamoe@46.204.77.176] has quit [Client Quit] 00:18 -!- Gooberpatrol66 [~Gooberpat@user/gooberpatrol66] has quit [Remote host closed the connection] 00:18 -!- Gooberpatrol66 [~Gooberpat@user/gooberpatrol66] has joined #hplusroadmap 01:30 -!- darsie [~darsie@84-113-55-200.cable.dynamic.surfer.at] has joined #hplusroadmap 01:31 < fenn> some light reading: https://old.reddit.com/r/aliens/comments/14rp7w9/from_the_late_2000s_to_the_mid2010s_i_worked_as_a/ 01:31 < fenn> .t 01:31 < EmmyNoether> From the late 2000s to the mid-2010s, I worked as a molecular biologist for a national security contractor in a program to study Exo-Biospheric-Organisms (EBO). I will share with you a lot of information on this subject. Feel free to ask questions or ask for clarification : aliens 01:51 < sphertext_> cool science fiction 02:13 < fenn> http://people.eku.edu/ritchisong/554images/Respiration/Respiration-bird-mammal-insect.gif 02:36 -!- sphertext_ [~sphertext@user/sphertext] has quit [Ping timeout: 255 seconds] 02:39 -!- sphertext_ [~sphertext@user/sphertext] has joined #hplusroadmap 04:29 -!- yashgaroth [~ffffffff@2601:5c4:c780:6aa0:4907:3721:a261:6e96] has joined #hplusroadmap 04:31 -!- ike8 [12fdf2ee08@irc.cheogram.com] has quit [Ping timeout: 255 seconds] 04:56 -!- stipa [~stipa@user/stipa] has joined #hplusroadmap 07:52 -!- ike8 [12fdf2ee08@irc.cheogram.com] has joined #hplusroadmap 09:51 -!- ike8 [12fdf2ee08@irc.cheogram.com] has quit [Ping timeout: 255 seconds] 10:10 -!- ike8 [12fdf2ee08@irc.cheogram.com] has joined #hplusroadmap 10:33 -!- ike8 [12fdf2ee08@irc.cheogram.com] has quit [Ping timeout: 240 seconds] 12:30 -!- TMM_ [hp@amanda.tmm.cx] has quit [Quit: https://quassel-irc.org - Chat comfortably. Anywhere.] 12:30 -!- TMM_ [hp@amanda.tmm.cx] has joined #hplusroadmap 13:10 < kanzure> "Loss of Notch signaling in skeletal stem cells enhances bone formation with aging" https://www.nature.com/articles/s41413-023-00283-8 13:58 -!- ike8 [12fdf2ee08@irc.cheogram.com] has joined #hplusroadmap 14:51 < fenn> this is a really impressive integration of a bunch of random AI stuff such as chatGPT, google image search, and good old fashioned bot scripting https://www.youtube.com/watch?v=0wTf_bbkW2U&t=19m20s 14:51 < fenn> .t 14:51 < EmmyNoether> Creating a Self-Aware Lara Croft that Plays Tomb Raider - YouTube 17:34 < sphertext_> does anybody remember those transdisciplinary papers in complexity science, where they used something from thermodynamics, idk i think it was brownian motion, and applied it to patterns of human migration, or something like that? 17:38 -!- darsie [~darsie@84-113-55-200.cable.dynamic.surfer.at] has quit [Ping timeout: 272 seconds] 17:41 < ike8> you can search the logs 17:41 < ike8> https://gnusha.org/logs/ 17:42 < sphertext_> i doubt the logs capture exhaustively the past 15 years in the history of science 17:42 < sphertext_> i was just wondering if anyone else here have heard about those papers, as they were kinda famous/interesting 17:45 < ike8> ah, gotcha. I assumed you spoke with someone here about those papers. 17:45 < kanzure> "To eliminate the repeat that causes the major deletion from human mitochondria" 17:45 < kanzure> "To break up both the FOXO and TOR networks into 2 or more independent sub-networks that are independently regulated, so you don't have to turn the whole mTOR network off to do any of the FOXO3a longevity-enhancing expression. But it's hard to imagine how to do that without synthesizing new DNA sequences for large-ish chunks of the genome. Even then, it wouldn't be easy." 17:46 < sphertext_> hehe nope. just a random question 17:46 < kanzure> "Replace many genes for constructing the mitochondria and for doing oxidative phosphorylation with genes derived from birds, whose mitochondria are much more robust to oxidative damage. 17:46 < kanzure> " 17:47 < Mabel> dunno about papers but there's Complexity Zoo which is a wiki on complexity of all things~ https://complexityzoo.net/Complexity_Zoo 17:48 < sphertext_> Mabel complexity sciences != computational complexity 17:49 < kanzure> a look from 10 years ago at aging: https://diyhpl.us/~bryan/papers2/longevity/aging_roadmap.txt 17:49 < kanzure> er, 13 years. 17:54 < kanzure> https://news.ycombinator.com/item?id=37714703 17:54 < kanzure> "“Chatbot That Gives Detailed Instructions on Murder, Ethnic Cleansing” Is this supposed to manipulate me? Just low effort drooling “think of the children!” nonsense? Who is at their laptops thinking, “damn I wish I could ethnically cleanse people but gosh darn I just don’t know where to begin”?" 17:56 -!- A_Dragon is now known as DemonDerg 17:57 < sphertext_> i honestly hoped we'd be over LLMs by the 4th quarter of this year 18:01 < kanzure> https://news.ycombinator.com/item?id=37721158 18:09 < sphertext_> i can't be the only one who thinks that the debate about LLM censorship is just a business tactic to stir ppl up, so that LLMs continue to capture their attention? 18:12 < kanzure> i don't understand, like you specifically don't believe that tech companies have been under pressure from the government to moderate content and all those reports are false or something? 18:15 < kanzure> (insert snarky remark about section 230 here) 18:18 < kanzure> really? "Llama.cpp inference on fast cpus is perfectly usable for 70B parameter models" 18:18 < sphertext_> "On one side are AI companies like OpenAI, researchers, and users who believe that for safety reasons, it is best for AI to be developed behind closed doors, where professionals and experts can decide what AI tools should and should not be able to do for their users." 18:18 < sphertext_> from the article you posted 18:18 < sphertext_> doesn't say anything about gov regulation 18:18 < sphertext_> it's OpenAi that is trying to stir up artificial panic 18:19 < L29Ah> all the base LLM-training crowd does it apparently 18:20 < L29Ah> at least Facebook + all the web-available interactive LLM services 18:21 < kanzure> sphertext_: the yudkowskyites are very interested in winning some extra territory from this so you will see lots of pro-regulation commentary 18:21 < L29Ah> the only publicly accessible large language model that wasn't messed up like this is llama, today maybe some fully-open-source-open-data effort have catched up to it 18:22 < kanzure> and unfortunately there doesn't seem to be much of a structured effort to stop ceding ground to lesswrong dwellers 18:22 < L29Ah> kanzure: are there that many yudkowskyites? 18:22 < kanzure> through effective altruism yes 18:22 < kanzure> e/acc has not done much 18:22 < kanzure> but maybe someone will solve this 18:26 < L29Ah> https://manifold.markets/RobinGreen/will-any-prominent-notkilleveryonei 18:27 < sphertext_> there's no danger in a broken tool; but if you keep yelling about regulating it, you might end up distracting everyone from how broken it is lol 18:27 < kanzure> yes what a surprise, the people who are ideologically aligned with eliezer who created a polling system are also the same people who would vote for an outcome that aligns with the ideological alignment thing. wow! amazing outcome! 18:28 < kanzure> "examine the arguments in detail" == curtail to the gishgallop 18:28 < L29Ah> that's supposed to be a prediction market, not a polling system 18:28 < kanzure> whatever, it's still the same people 18:30 < kanzure> "wow! the dominant god internet cult is precisely represented on a website marketed and made to cater to members of that dominant machine god religion!" 18:30 < kanzure> *machine god 18:30 < sphertext_> lol 18:30 < kanzure> what is this, anyway? what happened to metaculus or predicit- didn't predictit get reinstated after kalshi showed how to be regulated? 18:35 < kanzure> also: i don't believe that harassing everyone into participating in the machine god cult is an appropriate game theoretic strategy. i know it seems appetitizing and appealing but usually people fight pretty hard against that kind of totalitarianism. 18:35 -!- yashgaroth [~ffffffff@2601:5c4:c780:6aa0:4907:3721:a261:6e96] has quit [Quit: Leaving] 18:39 < sphertext_> riling ppl up is precisely the point imho. riling ppl up about trivialities has been a dominant tactic in public relations for societal control the past decade 18:41 < L29Ah> poked manifold for some time now, poking metaculus, apparently there are plenty of small differences, i couldn't pinpoint a crucial one yet, metaculus has more varied question types yet lacking tags for questions' (subscriptions/filtering) 18:43 < L29Ah> apparently on manifold one can insert monies and spend monies in pre-approved charities 18:43 < kanzure> https://accelerator.github.com/ 18:43 -!- TMM_ [hp@amanda.tmm.cx] has quit [Quit: https://quassel-irc.org - Chat comfortably. Anywhere.] 18:43 -!- TMM_ [hp@amanda.tmm.cx] has joined #hplusroadmap 18:45 < sphertext_> interesting that github is preserving dignity after the microsoft acquisition 18:48 < kanzure> "US government backs thymus rejuvenation project with $37m ARPA-H funding[..] Thymus Rejuvenation project is led by Thymmune Therapeutics, a George Church..." 18:54 < L29Ah> sphertext_: they've closed registrations with anonymous e-mail providers 18:55 < sphertext_> that sucks. there's gitlab, but i hate their front end, because it doesn't work without javascript.. 18:56 < sphertext_> so you can't read project readmes/documentation/wikis in console browsers 18:57 < kanzure> https://www.palladiummag.com/2020/04/01/the-bison-sphere-manifesto/ 19:00 < kanzure> is there a specific argument for a large-scale space structure that is manageable in the near-term (next 20-30 years) that could also support a population size of several billion people? 19:13 -!- ike8 [12fdf2ee08@irc.cheogram.com] has quit [Ping timeout: 255 seconds] 20:08 < sphertext_> so there's an app called "too good to go" where local restaurants/cafes/bakeries/grocery stores go and say "hey we have this leftover food today, you can have it for a couple of currencies" 20:09 < sphertext_> and you can reserve it, then go and pick it up 20:09 < sphertext_> it's intended to mitigate food waste. i think it's a fantastic app 20:10 < sphertext_> problem is that it's super popular, so ppl keep refreshing the app and when something comes up, they reserve it quickly, before others do 20:10 < sphertext_> so i mitm'ed the app traffic to find the API calls it makes, so i can write a script to monitor for updates automatically :D 20:12 < sphertext_> ik it's kinda basic, but it's one of those neat pay offs that come from having a bit of technical knowledge 20:13 < sphertext_> most of the stuff is discounted in the range of 50-75% so it's a big win 20:26 < kanzure> have you ever considered using your skills to just make more money so that you don't need a discount of 75% to eat human food 20:29 < sphertext_> i have more important things to do with my life, than becoming an agent of capitalism, especially since i can already afford most of what i need, with a few tricks 20:29 < sphertext_> why would i let capitalism exploit me, when i can exploit it instead? 20:30 < kanzure> exploit you? 20:31 < sphertext_> for labour 20:31 < sphertext_> usually towards questionable goals 20:33 < sphertext_> do i really want to build another cookie tracker to increase sales of toxic junk food? 20:33 < sphertext_> also, i have very low neurological resilience to stress and struggle with patience in social interactions, so i don't cope well in high pressure workplace environments. 20:33 < sphertext_> dont judge me :( 22:48 < muurkha> kanzure: " specific argument for a large-scale space structure that is manageable in the near-term (next 20-30 years) that could also support a population size of several billion people" is a possible summary of https://dernocua.github.io/notes/backward-exponential.html 22:51 < muurkha> but I don't know if converting asteroid Pallas into separate O'Neill colonies is what you have in mind 22:52 < muurkha> since it's numerous space structures rather than just one impractically enormous one 23:02 -!- sphertext_ [~sphertext@user/sphertext] has quit [Ping timeout: 255 seconds] 23:06 -!- sphertext_ [~sphertext@user/sphertext] has joined #hplusroadmap --- Log closed Sun Oct 01 00:00:38 2023