--- Log opened Wed Apr 17 00:00:29 2024
01:16 -!- darsie [~darsie@84-112-12-36.cable.dynamic.surfer.at] has joined #hplusroadmap
02:29 -!- justanot1 [~justanoth@gateway/tor-sasl/justanotheruser] has quit [Ping timeout: 260 seconds]
03:37 -!- millefy [~Millefeui@91-160-78-132.subs.proxad.net] has quit [Ping timeout: 252 seconds]
05:16 < kanzure> hmph
05:36 -!- darsie [~darsie@84-112-12-36.cable.dynamic.surfer.at] has quit [Quit: Wash your hands. Don't touch your face. Avoid fossil fuels and animal products. Have no/fewer children. Protest, elect sane politicians. Invest ecologically.]
05:37 -!- darsie [~darsie@84-112-12-36.cable.dynamic.surfer.at] has joined #hplusroadmap
05:44 -!- darsie [~darsie@84-112-12-36.cable.dynamic.surfer.at] has quit [Quit: Wash your hands. Don't touch your face. Avoid fossil fuels and animal products. Have no/fewer children. Protest, elect sane politicians. Invest ecologically.]
05:46 -!- darsie [~darsie@84-112-12-36.cable.dynamic.surfer.at] has joined #hplusroadmap
07:00 -!- cthlolo [~lorogue@77.33.24.3.dhcp.fibianet.dk] has joined #hplusroadmap
08:15 -!- cthlolo [~lorogue@77.33.24.3.dhcp.fibianet.dk] has quit [Quit: Leaving]
08:21 -!- CryptoDavid [uid14990@id-14990.uxbridge.irccloud.com] has joined #hplusroadmap
08:41 < kanzure> future of humanity institute (the doomerism one) has closed down, this is probably good for the world
08:43 < kanzure> "varieties of doom" https://gist.github.com/JD-P/915ab877c20774d0d8690bd0b95ab061
08:54 < hprmbridge> fodagut> Is this the Oxford one? Maybe this frees up time for Drexler.
09:27 < hprmbridge> alonzoc> They obviously ran out of money once Bankman stopped donating
09:38 < hprmbridge> kanzure> bostrom had other funding before SBF tho
09:39 < hprmbridge> alonzoc> I should have marked that with a sarcasm or half-joking decleration, I am curious why it actually closed though.
09:39 < hprmbridge> alonzoc> From what I can gather it started all the way back in 2020 with a hiring freeze which extended into contracts not being renewed
09:40 < hprmbridge> alonzoc> However Bankmans public fraud and it's relation to EA has caused interest and support of EA to dry up a few pro-EA people on campus have lamented as such to me
09:59 < hprmbridge> fodagut> Unfortunately on the AI front in particular vitalik has been providing plenty of funding.
10:23 -!- justanotheruser [~justanoth@gateway/tor-sasl/justanotheruser] has joined #hplusroadmap
11:05 -!- justanotheruser [~justanoth@gateway/tor-sasl/justanotheruser] has quit [Remote host closed the connection]
11:07 -!- justanotheruser [~justanoth@gateway/tor-sasl/justanotheruser] has joined #hplusroadmap
11:10 -!- CryptoDavid [uid14990@id-14990.uxbridge.irccloud.com] has quit [Quit: Connection closed for inactivity]
11:40 < hprmbridge> kanzure> https://adprofectum.substack.com/p/the-case-for-biological-enhancement
11:42 < hprmbridge> kanzure> https://www.thinkingcomplete.com/2018/04/a-model-of-intelligence.html?m=1
11:46 < hprmbridge> kanzure> https://www.nooceleration.com/p/the-biosingularity-is-near
12:51 -!- millefy [~Millefeui@91-160-78-132.subs.proxad.net] has joined #hplusroadmap
13:11 < kanzure> "Construction of arbitrarily strong amplifiers of natural selection using evolutionary graph theory" https://www.nature.com/articles/s42003-018-0078-7
13:16 < kanzure> "Complex computation from developmental priors" https://www.biorxiv.org/content/10.1101/2021.03.29.437584v3.abstract
13:17 < kanzure> "Encoding innate ability through a genomic bottleneck" https://www.biorxiv.org/content/10.1101/2021.03.16.435261v2.abstract
13:50  * fenn chuckles at "embryo thunderdome"
13:55 < kanzure> funny to look back on this 2014 page and see what we got wrong https://diyhpl.us/wiki/projects/human-like-cognitive-abilities/
13:56 < kanzure> stochastic parrot is in there, but also it says "must not require impossibly huge amounts of computational resources"
13:56 < kanzure> "Implementations must run on commodity computing hardware."
13:59 < hprmbridge> Lev> fwiw i think the bitter lesson post applies to the "no emergence" part too
14:00 < hprmbridge> Lev> Like by far it seems the solution is "pretrain" and not "design" with few real exceptions
14:06 < kanzure> more on the shutdown of the future of humanity institute https://news.ycombinator.com/item?id=40065862 including some complaints about bostrom and his 'superintelligence' book
14:06 < kanzure> "future of humanity institute final report" https://static1.squarespace.com/static/660e95991cf0293c2463bcc8/t/661a3fc3cecceb2b8ffce80d/1712996303164/FHI+Final+Report.pdf
14:08 < kanzure> the "where we failed" section needs more introspection and self awareness
14:08 < kanzure> they spent their whole existence basically fighting against the future for humanity, and therefore don't deserve the title they claimed.
14:12 < fenn> "
14:12 < fenn> "Implementations must run on commodity computing hardware." turned out to be correct. imagenet didn't happen until fly brain-level GPU was below $2000
14:12 < fenn> the price a grad student could personally afford
14:54 -!- Llamamoe [~Llamamoe@188.146.91.151] has quit [Quit: Leaving.]
17:07 < hprmbridge> fodagut> The "where we failed" section was rediculously short.
17:08 < hprmbridge> kanzure> right? that's not a retrospective.
17:08 < hprmbridge> kanzure> we did everything perfectly, it is the world and/or society that is wrong!
17:12 < hprmbridge> kanzure> haha https://groups.google.com/d/msgid/bitcoindev/fb52ccb5-9942-4db8-b880-3c06ebc47cd1%40achow101.com
17:13 < hprmbridge> kanzure> "I would like to personally raise caution against Kanzure. I'm sure he's more than capable if so many people believe so but just want to highlight that based on his own public twitter thread, he claimed that he has better things to do."
17:17 < hprmbridge> alonzoc> Well they have no good metrics for a retrospective. The spent their time pontificating without producing much of value to the issues they wished to resolve.
17:22 < hprmbridge> alonzoc> Yeah, upon rereading the where we failed seems to be mostly "Why we were shut down" which is obviously their only failing.
17:22 < hprmbridge> alonzoc> - Bad communication with the rest of academia
17:22 < hprmbridge> alonzoc> - Failure to establish office political footholds beyond our department/institute
17:22 < hprmbridge> alonzoc> - To informal to mill out papers sufficient for a group their size
17:38 -!- darsie [~darsie@84-112-12-36.cable.dynamic.surfer.at] has quit [Quit: Wash your hands. Don't touch your face. Avoid fossil fuels and animal products. Have no/fewer children. Protest, elect sane politicians. Invest ecologically.]
17:51 < docl> well there's still the future of life institute... for better or worse
19:03 < fenn> this is why academia is crap
19:04 < fenn> produce a stream of incremental research updates, or else
19:08 < fenn> .m https://twitter.com/prash_singh/status/1780522588873040316
19:08 < AugustaAda> ​https://video.twimg.com/ext_tw_video/1780521119721668608/pu/vid/avc1/1920x1080/QM_Ai_UISNaaW65T.mp4?tag=14
19:23 < hprmbridge> nmz787> fenn are you talking about that twitter post?
19:23 < hprmbridge> nmz787> or the future of life institute?
19:23 < hprmbridge> nmz787> re incmremental crap
19:39 < fenn> the future of humanity institute. someone said they didn't mill out enough papers
19:43 < hprmbridge> alonzoc> I agree the milling out papers is a flaw in academia but it's still a game a research institute working on university money has to play regardless.
19:49 < hprmbridge> alonzoc> To be honest they were probably the best longtermist institute when it came to instrumentally useful results such as their interruptibility paper.
20:55 -!- mxz__ [~mxz@user/mxz] has joined #hplusroadmap
20:56 -!- mxz_ [~mxz@user/mxz] has quit [Ping timeout: 256 seconds]
20:57 -!- mxz [~mxz@user/mxz] has quit [Ping timeout: 268 seconds]
20:57 -!- mxz__ is now known as mxz
23:02 -!- mxz_ [~mxz@user/mxz] has joined #hplusroadmap
--- Log closed Thu Apr 18 00:00:30 2024