--- Log opened Fri Jan 06 00:00:06 2023
05:02 < kanzure> calling it "soft eugenics" is silly https://undark.org/2023/01/06/interview-matthew-cobb-on-the-ethics-of-genetic-engineering/
05:10 -!- darsie [~darsie@84-113-55-200.cable.dynamic.surfer.at] has joined #hplusroadmap
05:20 -!- yashgaroth [~ffffffff@2601:5c4:c780:6aa0::4249] has joined #hplusroadmap
06:58 -!- juri_ [~juri@84-19-175-179.pool.ovpn.com] has quit [Ping timeout: 252 seconds]
07:15 -!- juri_ [~juri@84-19-175-179.pool.ovpn.com] has joined #hplusroadmap
07:27 -!- Molly_Lucy [~Molly_Luc@user/Molly-Lucy/x-8688804] has quit [Read error: Connection reset by peer]
07:46 -!- saxo [~saxo@2001:19f0:6800:1102:5400:ff:fe11:39a1] has joined #hplusroadmap
08:49 -!- cthlolo [~lorogue@77.33.23.154.dhcp.fibianet.dk] has joined #hplusroadmap
09:00 < kanzure> what lately are the best designed biologically-accurate-but-not-neuronally-reductionist brain simulations?
09:15 -!- TMM_ [hp@amanda.tmm.cx] has quit [Quit: https://quassel-irc.org - Chat comfortably. Anywhere.]
09:15 -!- TMM_ [hp@amanda.tmm.cx] has joined #hplusroadmap
09:40 -!- darsie [~darsie@84-113-55-200.cable.dynamic.surfer.at] has quit [K-Lined]
10:33 < maaku> .tw 1611286048084041728
10:33 < saxo> When it comes to similarities between the brain and deep learning, what's really striking is that everything that was actually bio inspired (e.g. sigmoid/tanh activations, spiking NNs, hebbian learning, etc.) had been dropped, while... (Cont.) (@fchollet) psa from .tw cmd: fuck twitter
10:33 < maaku> lol what's with the saxo psa commentary
10:34 < kanzure> dpk hates elon i think
10:34 < kanzure> it's some sort of political protest
10:35 < kanzure> .tw https://twitter.com/fchollet/status/1611290359023157252
10:35 < saxo> Running tons of experiments while having very few priors about what the solution should look like is tremendously more effective than coming up with fancy theories about how the brain really works and repeatedly trying to prove those theories. Numenta also comes to mind here... (@fchollet, in reply to tw:1611288306574295042) psa from .tw cmd: fuck twitter
10:35 < kanzure> or, nobody has actually been looking at the brain because of the immense difficulty of acquiring dead human brains
10:36 < maaku> is it that hard to legally aquire a cadaver?
10:37 < maaku> idk I'm inclined to believe that brains just aren't even close to optimal
10:43 < lsneff> they’re just complicated
10:43 < lsneff> yeah, almost nothing from bio neural networks is still used
10:43 < lsneff> Too expensive
10:44 < maaku> evolution has the problem of having to have purely local solutions
10:44 < maaku> that's a massive constraint engineered systems don't need to have
10:45 < maaku> .t https://unchartedterritories.tomaspueyo.com/p/transportation-tech-shaped-empires
10:45 < saxo> How Transportation Technologies Shaped Empires
10:45 < kanzure> maaku: have you tried acquiring cadavers
10:45 < maaku> no
10:46 < kanzure> part of the issue is that human remains cannot be bought or sold in the united states
10:46 < kanzure> so no market for research cadavers can develop
10:50 < kanzure> https://braindonorproject.org/
10:51 < kanzure> federal brain tissue depository https://neurobiobank.nih.gov/about/
10:52 < kanzure> where do you even click to get shipped a brain
10:52 < kanzure> https://neurobiobank.nih.gov/researchers/
10:52 < kanzure> "Requests for tissue must come from an institution that has an active Federal Wide Assurance (FWA) with the U.S. Department of Health and Human Services (DHHS), Office of Human Research Protection (OHRP)."
10:52 < kanzure> office of what now
10:54 < kanzure> "Our networked brain and tissue repositories distribute thousands of samples per year" that's like nothing- there are 8 million scientists in the world
10:56 < kanzure> (2016) "The brain bank at Belmont’s McLean Hospital is facing a dire shortage of healthy gray matter — receiving just one nondiseased brain donation in nearly a year — and scientists say the lack of specimens threatens to stymie research"
10:57 < kanzure> "More than 3,000 brains are stored at the Harvard Brain Tissue Resource Center at McLean Hospital just outside Boston. It is one of the largest brain banks in the world."
10:58 < kanzure> this one stopped accepting brain donations because they reached capacity (less than 100,000 brains) https://www.abc.net.au/news/2022-01-04/sa-brain-bank-not-accepting-donations-for-science-research/100736278 (in fact it looks like only 380 brains)
12:21 < nsh> <homer> mmmm, 100,000 brain capacitor...
12:27 < kanzure> https://borretti.me/assets/content/eog581/The%20Epiphany%20of%20Gliese%20581.pdf
12:40 -!- Llamamoe [~Llamamoe@46.204.76.114.nat.umts.dynamic.t-mobile.pl] has quit [Quit: Leaving.]
12:47 -!- saxo [~saxo@2001:19f0:6800:1102:5400:ff:fe11:39a1] has quit [Remote host closed the connection]
12:47 -!- saxo [~saxo@2001:19f0:6800:1102:5400:ff:fe11:39a1] has joined #hplusroadmap
15:16 -!- cthlolo [~lorogue@77.33.23.154.dhcp.fibianet.dk] has quit [Read error: Connection reset by peer]
16:30 -!- L29Ah [~L29Ah@wikipedia/L29Ah] has joined #hplusroadmap
16:30 -!- L29Ah [~L29Ah@wikipedia/L29Ah] has left #hplusroadmap []
16:31 -!- L29Ah [~L29Ah@wikipedia/L29Ah] has joined #hplusroadmap
17:15 < kanzure> "RNA recording in single bacterial cells using reprogrammed tracrRNAs" https://www.nature.com/articles/s41587-022-01604-8
17:16 < kanzure> .tw https://twitter.com/Chunlei_Jiao/status/1611055354481188865
17:16 < saxo> In this paper, we transplant the Rptr concept in vivo to develop a direct RNA recording platform called TIGER. allowing us to record RNA in a programmable, quantitative, multiplexable way at a single-cell level. (@Chunlei_Jiao, in reply to tw:1611055351721525251) psa from .tw cmd: fuck twitter
17:16 < kanzure> "primate growth&reproduction biscuit" https://twitter.com/goth600/status/1611468406271922176
17:17 < kanzure> "Improved cytosine base editors are generated by phage-assisted evolution of a deoxyadenosine deaminase" https://www.nature.com/articles/s41587-022-01533-6
17:17 < kanzure> "A transcription factor atlas of directed differentiation" https://www.sciencedirect.com/science/article/pii/S0092867422014702 https://twitter.com/JuliaJoung/status/1611032251734822912
17:53 < kanzure> http://blog.agi.io/2015/12/how-to-build-general-intelligence.html
17:55 < kanzure> https://cerenaut.ai/research-roadmap/
17:59 < kanzure> this seems to be what became of leabra, under active development too https://github.com/emer/emergent
17:59 < kanzure> http://emersim.org/
18:00 < kanzure> textbook https://compcogneuro.org/
18:04 -!- yashgaroth [~ffffffff@2601:5c4:c780:6aa0::4249] has quit [Quit: Leaving]
18:23 < kanzure> "We will focus mainly on supporting Leabra and other related biologically-based models that currently don't fit well within the existing Tensor-gradient-descent paradigm as implemented in PyTorch and TensorFlow, etc. Leabra uses synchronous bidirectional activation dynamics, instead of the cascaded feed-forward dynamics of standard backpropagation models, and incorporates ion-channel-based ...
18:23 < kanzure> ...activation dynamics and various scopes of inhibitory interactions within layers. Furthermore, these models have sparse overall activity levels, and benefit greatly from sender-based computation, which is also highly beneficial for discrete spiking models."
--- Log closed Sat Jan 07 00:00:07 2023