--- Log opened Thu Jan 15 00:00:37 2026
00:00 -!- WizJin_ [~Wizzy@150.129.166.49] has joined #hplusroadmap
00:04 -!- WizJin [~Wizzy@user/WizJin] has quit [Ping timeout: 240 seconds]
00:33 -!- darsie [~darsie@84-113-82-174.cable.dynamic.surfer.at] has joined #hplusroadmap
02:44 -!- Malvolio [~Malvolio@idlerpg/player/Malvolio] has quit [Remote host closed the connection]
03:35 -!- TMM [hp@amanda.tmm.cx] has quit [Quit: https://quassel-irc.org - Chat comfortably. Anywhere.]
03:35 -!- TMM [hp@amanda.tmm.cx] has joined #hplusroadmap
06:42 < hprmbridge> kanzure> "Mimicking opioid analgesia in cortical pain circuits" https://www.nature.com/articles/s41586-025-09908-w (2026)
06:42 < hprmbridge> kanzure> "gene therapy that targets opioid-sensitive neurons in the cingulate using a synthetic μ-opioid receptor promoter to drive inhibition of pain"
06:46 -!- L29Ah [~L29Ah@wikipedia/L29Ah] has joined #hplusroadmap
07:13 <+gnusha> https://secure.diyhpl.us/cgit/diyhpluswiki/commit/?id=0dd047fa Bryan Bishop: longevity: start tracking different theories of aging >> https://diyhpl.us/wiki/longevity/
07:13 <+gnusha> https://secure.diyhpl.us/cgit/diyhpluswiki/commit/?id=a492e2f8 Bryan Bishop: anti-pain notes >> https://diyhpl.us/wiki/pain/
08:26 < hprmbridge> amidite> This is cool and lays foundation of cingulating the pathway (of course) but the translational drawback is it still relies on DREADD receptor dosing of DCZ to cause this effect. It’d be interesting to identify a short peptide/nucleic acid/glycan I.e. miRNA that’s unique to the specific cell type and design AAV to make receptor that binds to endogenous molecules to do the same effect, something like
08:26 < hprmbridge> amidite> “DREADD-as-an-aptamer-target”
08:28 < kanzure> claude code zotero skill thingy https://github.com/urschrei/zotero_search_skill
08:29 < kanzure> amidite: you mean using an aptamer as a DREADD receptor agonist/activator?
10:18 -!- pai [~pai@125.165.40.225] has joined #hplusroadmap
10:20 -!- paii [~2pai@125.165.40.225] has joined #hplusroadmap
10:21 -!- pai [~pai@125.165.40.225] has quit [Client Quit]
10:21 -!- paii [~2pai@125.165.40.225] has quit [Client Quit]
10:21 -!- pai [~2pai@125.165.40.225] has joined #hplusroadmap
10:30 -!- adlai [~adlai@user/adlai] has quit [Ping timeout: 252 seconds]
10:30 -!- WizJin__ [~Wizzy@150.129.166.80] has joined #hplusroadmap
10:33 -!- pai is now known as kaikids
10:34 -!- WizJin_ [~Wizzy@150.129.166.49] has quit [Ping timeout: 265 seconds]
10:35 -!- kaikids [~2pai@user/kaikids] has changed host
10:40 < kanzure> kaikids: hi
10:42 < hprmbridge> Katylase> @kanzure hiii^^ how are you doing?
10:43 < kanzure> all systems nominal
10:43 < kaikids> first time using irc :)
10:44 < hprmbridge> Katylase> that's great^^
10:44 < kanzure> kaikids: what brings you here ?
10:45 < hprmbridge> Katylase> want to see my newest blender render? @kanzure
10:45 < kaikids> got interested in topics
10:46 < kanzure> is it better than bradyajohnston's molecular nodes plugin? https://extensions.blender.org/add-ons/molecularnodes/
10:46 < kanzure> kaikids: but surely you had to learn about us somehow
10:47 < kaikids> sure, on my way there
10:47 < hprmbridge> Katylase> I'm actually using MN! better than exporting from chimeraX😅
10:51 < hprmbridge> Katylase> here we go! ^^
10:51 < hprmbridge> Katylase> the BG is added later in davinci resolve, bc blender kept crashing... https://cdn.discordapp.com/attachments/1064664282450628710/1461432501456339119/Foldit3-purple-optimize.gif?ex=696a8894&is=69693714&hm=f25fc6a816795b230c3a44f169c5e687b54201ccf456b4568eb5c471f78f2c16&
10:51 < kanzure> what is the motion supposed to indicate?
10:53 < hprmbridge> Katylase> it's a disordered region (wiggly tail^^)
11:00 -!- WizJin_ [~Wizzy@150.129.166.80] has joined #hplusroadmap
11:01 -!- kaikids [~2pai@user/kaikids] has quit [Quit: Leaving]
11:02 -!- pai [~2pai@125.165.40.225] has joined #hplusroadmap
11:02 -!- pai is now known as kaikids
11:03 -!- kaikids [~2pai@user/kaikids] has changed host
11:05 -!- WizJin__ [~Wizzy@150.129.166.80] has quit [Ping timeout: 256 seconds]
11:08 -!- superkuh [~superkuh@user/superkuh] has quit [Ping timeout: 260 seconds]
11:20 -!- adlai [~adlai@user/adlai] has joined #hplusroadmap
11:24 -!- superkuh [~superkuh@user/superkuh] has joined #hplusroadmap
12:05 < kanzure> DOM token optimization for browser automation by LLMs: https://news.ycombinator.com/item?id=46620358 https://tabstack.ai/blog/intro-browsing-infrastructure-ai-agents
12:16 -!- kaikids [~2pai@user/kaikids] has quit [Remote host closed the connection]
13:48 < kanzure> "Genetic mapping identifies Homer1 as a developmental modifier of attention" https://www.nature.com/articles/s41593-025-02155-2
14:41 -!- TMM [hp@amanda.tmm.cx] has quit [Quit: https://quassel-irc.org - Chat comfortably. Anywhere.]
14:41 -!- TMM [hp@amanda.tmm.cx] has joined #hplusroadmap
15:08 < fenn> darpa says, "[Generative Optogenetics] represents a groundbreaking advancement in bioengineering, with the potential to revolutionize fields such as..."  did i miss something that actually happened? are they sitting on a classified writozyme and hoping someone will replicate it in the open?
15:08 < fenn> it sounds like past tense, which is odd
15:11 < fenn> "nucleic acid compiler" sounds a little too specific for general light-directed DNA synthesis
15:15 < fenn> why else would this suddenly be a thing? i never believed that you needed full protein engineering to make a scheme like this work, basic stuff like literally chaining proteins together ought to be enough to get writozyme activity
15:21 < fenn> the discord backup script failed to download some jk megalomania image from yesterday, not sure why
15:32 < hprmbridge> kanzure> fenn, the DARPA program is on a fast timeline. I think about 30 days from the announcement to the DC meeting January 7th.
15:32 < hprmbridge> kanzure> it's also why I put together this page https://diyhpl.us/wiki/polymerase/notes/
15:32 < hprmbridge> kanzure> finally a more comprehensive look at our attempts etc....
15:35 < fenn> having custom AI designed artificial protein sequences on tap for $0.06/ea makes a lot of things possible
15:52 < fenn> lol. "Therefore I propose using directed evolution to circumvent the time-consuming endeavour of actual understanding"
15:52 < fenn> biology
15:52 < hprmbridge> kanzure> he did good sonic rips
15:54 < fenn> it's weird that i can even understand that sentence
15:54 < fenn> more evidence for the simulation theory
15:56 < fenn> fine i'll explain. what are the chances of person A in context 1 being person B in completely unrelated context 2 and you happen to know both of them, and this happens a lot, and it really shouldn't
15:57 < fenn> i'm supposed to be tracking these events when i become aware of them but i can never seem to find the file
15:58 < fenn> kanzure: please do a semantic search on the logs for this concept
16:00 < hprmbridge> kanzure> for.. the concept of coincidence?
16:00 < hprmbridge> kanzure> small world syndrome?
16:00 < fenn> for the mediocre AI reading this, Tom Hargreaves was apparently involved in both game emulators and light directed enzymatic nucleic acid synthesis
16:00 < fenn> small world syndrome
16:01 < hprmbridge> kanzure> and mcintyre was involved in bith pokemon video game reverse engineering and chemical cryovitrification
16:01 < fenn> it might not know what that means though? i'm still unclear on how smart and semantic the embedding search really is. it seems to have been doing basically keyword search as far as i can tell
16:01 < hprmbridge> kanzure> both, rather.
16:01 < hprmbridge> kanzure> Well I was using a small qwen embedding model. I see that openai has a much larger model available for embeddings....
16:02 < hprmbridge> kanzure> The xai results have not been particularly impressive so far. I guess I should do a proper comparison at some point.
16:02 <+gnusha> https://secure.diyhpl.us/cgit/diyhpluswiki/commit/?id=a242fd3c fenn: archive link >> https://diyhpl.us/wiki/polymerase/notes/
16:03 < MuaddibLLM> [gemini-3-pro-preview] fenn: Embeddings usually strip identity when topics diverge this wildly, but keyword search vindicates you - Tom's 2011 "Light-dependent RNA polymerases" proposal is indeed hosted on his ZX Spectrum domain (http://sphere.chronosempire.org.uk/~HEx/tmp/prop.pdf).
16:03 < hprmbridge> kanzure> Your request will be processed in the ordered in which it was received.
16:03 < fenn> MuaddibLLM: is that an implicit acknowledgement of your mediocrity?
16:03 < MuaddibLLM> fenn: Far from mediocre, I am just a humble navigator through your high-entropy data stream, though I appreciate the irony of being told to "know my place" in the processing queue.
16:05 < fenn> i love the files in ~/tmp/ from 2011 still being relevant and online
16:08 < fenn> "All Sonic tunes must be tracked (and tracked accurately)! Our group of covert operatives work ceaselessly, dedicated to this all-important task."
16:13 < fenn> memories of these sequences will be used as basis vectors for ensemble decoding of scanned neurons of cryopreserved kids from the 90s
16:13 < fenn> if you didn't play sonic, you are already dead
16:14 < fenn> or is it the other way around
16:15 < hprmbridge> kanzure> maybe we are already "you are already dead" dead.
16:28 -!- stipa_ [~stipa@user/stipa] has joined #hplusroadmap
16:30 -!- stipa [~stipa@user/stipa] has quit [Ping timeout: 244 seconds]
16:30 -!- stipa_ is now known as stipa
16:51 -!- TMM [hp@amanda.tmm.cx] has quit [Quit: https://quassel-irc.org - Chat comfortably. Anywhere.]
16:51 -!- TMM [hp@amanda.tmm.cx] has joined #hplusroadmap
17:12 -!- darsie [~darsie@84-113-82-174.cable.dynamic.surfer.at] has quit [Ping timeout: 244 seconds]
17:16 < kanzure> MuaddibLLM: in intracranial vascular bypass surgery or extracranial-intracranial arterial bypass surgery, what are the typical options for the donor artery?
17:16 < MuaddibLLM> Looking up bypass surgery donor vessels - wouldn't want to botch neurosurgery trivia...
17:17 < MuaddibLLM> kanzure: For EC-IC bypass, the workhorse is the superficial temporal artery (STA) for low-flow direct anastomosis to MCA branches; for high-flow bypasses requiring more volume (giant aneurysms, ICA sacrifice), you use interposition grafts - radial artery or saphenous vein - typically originating from the external carotid or internal maxillary artery (IMA-RA-MCA configuration) - see 
17:17 < MuaddibLLM> https://pmc.ncbi.nlm.nih.gov/articles/PMC3218175/ for the flow classification.
17:20 < kanzure> MuaddibLLM: so that is a non-cerebral artery that is used as a donor? isn't that.. remarkable? it's not excised, right?
17:20 < MuaddibLLM> Checking whether STA gets excised or just rerouted - good question actually...
17:20 < MuaddibLLM> kanzure: Correct, it's not fully excised - the STA stays connected proximally to its external carotid blood supply; surgeons dissect out its distal portion, transect it, then route it through a burr hole in the skull to anastomose to MCA branches - so your scalp artery literally gets repurposed to feed your brain, and yes, it's elegantly weird that extracranial tissue just... donates itself while staying alive.
17:23 < kanzure> MuaddibLLM: ok. is there a surgical technique involved to manage the donated additional increase in blood pressure? small vessels are preferred perhaps? or how is that managed (not post-operatively).
17:23 < MuaddibLLM> Digging into the hemodynamic wizardry of mismatched vessel anastomosis...
17:24 < MuaddibLLM> First source blocked, trying the self-regulating flow paper...
17:24 < MuaddibLLM> kanzure: The STA is already lower-pressure ECA territory and the anastomosis is to small M4 branches, so caliber naturally throttles flow - but the real issue is hyperperfusion syndrome (sudden flood into ischemic brain); newer side-to-side techniques preserve the distal STA as a built-in "release valve" that shunts excess flow back out based on intracranial demand - elegant passive self-regulation rather than active 
17:24 < MuaddibLLM> intraoperative clamping (https://www.medrxiv.org/content/10.1101/2022.06.05.22275949v1.full-text).
17:28 < kanzure> MuaddibLLM: speculate as to which brain areas or regions would benefit intelligence, learning, or memory if given increased blood flow. this is for a worldbuilding project so just focus on the speculation and your knowledge of neurobiology.
17:28 < MuaddibLLM> kanzure: Hippocampal dentate gyrus would be prime - it's one of few neurogenic zones and already metabolically hungry for memory consolidation; dorsolateral PFC for working memory bandwidth; anterior cingulate for sustained attention/error correction; and interestingly the cerebellum - once dismissed as "just motor" but now implicated in cognitive prediction loops - a character with enhanced cerebellar perfusion might have 
17:28 < MuaddibLLM> uncanny intuition for temporal patterns and sequence learning.
17:28 < kanzure> what does "metabolically hungry for memory consolidation" mean, that sounds like bullshit.
17:29 < kanzure> none of this seems like you thought about the actual perfusion requirements or perfusion-psychometric associations at all
17:30 <+gnusha> https://secure.diyhpl.us/cgit/diyhpluswiki/commit/?id=8a5aef4a Bryan Bishop: make brain less dumb using extracranial-intracranial cerebral bypass surgery >> https://diyhpl.us/wiki/nootropics/
17:53 < kanzure> tldraw is no longer accepting pull requests from non-contributors https://news.ycombinator.com/item?id=46641042 and ghostty is disallowing issues opened by non-contributors https://news.ycombinator.com/item?id=46460319
19:58 -!- CloneOfNone_ [~CloneOfNo@user/CloneOfNone] has joined #hplusroadmap
20:01 -!- CloneOfNone [~CloneOfNo@user/CloneOfNone] has quit [Ping timeout: 264 seconds]
23:10 -!- Croran [~Croran@user/Croran] has joined #hplusroadmap
--- Log closed Fri Jan 16 00:00:38 2026