--- Log opened Mon Sep 01 00:00:27 2025
00:01 < hprmbridge> kanzure> https://www.oliviali.me/writing/raccoons
00:32 < hprmbridge> kanzure> "One-step nanoscale expansion microscopy reveals individual protein shapes" https://www.nature.com/articles/s41587-024-02431-9 with alphafold prediction comparison
00:38 < fenn> didn't kary mullis have some UFO-like experience where glowing green racoons telepathically communicated
00:39 < fenn> raccoons* who invents these words anyway
00:39 < fenn> ärähkuněm (“he scratches with his hands”)
00:41 < fenn> Powhatan; the language has been credited with being the source of more English loans than any other indigenous language
00:44 < fenn> why would you domesticate a raccoon when you could domesticate a "red panda" (which is closer to a raccoon than a panda)
00:45 < fenn> strong contender for cutest animal
01:02 -!- Gooberpatrol_66 [~Gooberpat@user/gooberpatrol66] has joined #hplusroadmap
01:04 -!- Gooberpatrol66 [~Gooberpat@user/gooberpatrol66] has quit [Ping timeout: 260 seconds]
02:20 < hprmbridge> .monokhrome> pretty sure he was a hallucinagenic drug user
02:20 < fenn> not at the time, according to him
02:21 < fenn> (and also not when he thought up PCR)
02:22 < hprmbridge> .monokhrome> well then glowing green racoons are atypical ufo occupants, so maybe he had contact with the spirit world
02:22 < fenn> there may not be any difference
02:22 < hprmbridge> .monokhrome> conceded
02:23 < fenn> another theory is time travel by man's far future descendants
02:24 < hprmbridge> .monokhrome> yes, I think I've heard at least 8 different theories proposed, only one of which is aliens. I'm trying to remember the other 5
02:25 < fenn> memory-altering drugs and silent helicopters from DOE doing radiation fallout monitoring
02:26 < fenn> everyone just forgot that memory altering drugs exist!
02:27 < fenn> would inhabitants of neighboring zoos in the simulation-multiverse count as "aliens"
02:28 < hprmbridge> .monokhrome> not if they come from outside the universe
02:28 < fenn> well philosophy is banned here so i won't get into that
02:29 < hprmbridge> .monokhrome> I asked chatgpt and it listed 3 hypothesis that all sound the same to me: spirit world beings, ultraterrestrials, and interdimensional beings
02:30 < fenn> i had no idea what it was so i looked it up:
02:31 < fenn> "ultraterrestrials are an alien race that, well, isn't really alien. In fact, they originate from Earth, just like us humans, but their civilization is so much older and more advanced than ours that they have no trouble hiding from us:
02:31 < hprmbridge> .monokhrome> no, that's cryptoterrestrials
02:31 < hprmbridge> .monokhrome> crypto=hidden
02:32 < fenn> hey man i don't make the words
02:32 < fenn> https://tvtropes.org/pmwiki/pmwiki.php/Main/Ultraterrestrials
02:32 < hprmbridge> .monokhrome> well tv tropes is in error
02:33 < fenn> John Keel is in error
02:34 < fenn> ok so: aliens, time travel, interdimensional, out-of-simulation, hidden civilization. that's only 5
02:35 < fenn> oh and US government black programs is 6
02:35 < hprmbridge> .monokhrome> I think he just took traditional ideas about demonology and relabled them as ultraterrestrials so he could attend ufo conferences and get his books into the ufo section at libraries
02:35 < fenn> i really know nothing about the guy
02:37 < fenn> "in the Army he was trained in psychological warfare as a propaganda writer"
02:37 < hprmbridge> .monokhrome> 7 is psychosocial/jungian collective unconsciousness, 8 is exotic unknown natural phenomena
02:38 < fenn> you're gonna have to explain 8
02:38 < fenn> is it just a catch-all?
02:38 < hprmbridge> .monokhrome> rare forms of lightning and so forth
02:39 < fenn> how does that result in people getting probed or receiving telepathic transmissions from sirius or whatever
02:39 < hprmbridge> .monokhrome> have you heard of red sprites and blue jets?
02:39 < hprmbridge> .monokhrome> forms of upper atmospheric lightning, reported by pilots for a long time, wasn't really caught on camera until the 90s, nobody was sure if they were real or not
02:40 < fenn> https://x.com/Astro_Ayers/status/1940810789830451563
02:41 < fenn> i think a better approach is to not try to expect every unknown to neatly fit into the same theory
02:41 < hprmbridge> .monokhrome> there could be more stuff like that out there in the atmosphere, natural phenomena so rare that scientists can't get it on camera, but ufo witnesses sometimes see it and mistake it for aliens
02:41 < fenn> just like there are many causes of cancer, there are probably many causes of "ufo"
02:42 < hprmbridge> .monokhrome> also I worry that jung just peeled the labels off alchemy and relabled it jungian psychology, so maybe he should be grouped in with the spirits and magic explanation
02:43 < fenn> no it makes sense that there are psychological human invariants
02:44 < fenn> like you rub your eyes, you see geometric patterns and letter shapes because of how the retina processes visual stimuli
02:44 < fenn> you rub your brain, you see elves and trolls or whatever
02:44 < fenn> (i don't really buy this explanation fwiw)
02:45 < fenn> spirits and magic is not an explanation either
02:49 < hprmbridge> .monokhrome> maybe, maybe not, if you have a phenomena that has a cause, but all of the causes are unlikely, occams razor says you should pick no more than one of them as the explanation
02:56 < TMA> aren't the causes unlikely precise because they are trying to explain different phenomena? [like daytime UFO might have different cause than nightime UFO and a different cause than abduction and cattle mutilation and circles in the fields]
02:58 < TMA> so picking fairies, aliens, god, rare atmospheric phenomena are all wrong because they don't explain it all?
02:59 < TMA> [without making the explanation itself too complex, like the concept of "magic" or "god"]
03:02 < hprmbridge> .monokhrome> I would say explanations that it's caused by things like fairies, time travellers etc are unlikely because fairies, time travellers etc are unlikely to exist
03:03 < hprmbridge> .monokhrome> but you do bring up a point that other people have noticed - none of these explanations by themselves do a good job of explaining all the observed phenomena, so maybe it is multiple things
03:04 < hprmbridge> .monokhrome> A ufo historian I have a lot of respect for, called Jerome Clark has argued there are two things going on. But I don't think most people understood his arguments, let alone agreed with him.
03:06 < hprmbridge> .monokhrome> https://www.youtube.com/watch?v=SelOdzppZ_g
03:10 < fenn> seems to me that fairies and time travelers ought to exist, and it's unusual that we don't see any (anymore)
03:11 < fenn> it only takes ~1 million years to colonize the galaxy, so it should have happened repeatedly by now
03:12 < fenn> something is weird and none of this makes any sense
03:13 < fenn> .t
03:14 < fenn> "Dinsdale Award Lecture | Jerome Clark"
03:22 < hprmbridge> .monokhrome> here's the reason the ufo phenomena is unsolved - the actual explanation is so weird most people can't understand it
03:24 < hprmbridge> .monokhrome> Clark is trying to be abstract in his explanations, but he's arguing we have two things running around causing easily confused phenomena, one is aliens from outside our solar system, the other is something from outside our reality. Or something like that.
03:45 < fenn> 1960s MKULTRA and other shit was going on
03:45 < fenn> they literally injected a guy with plutonium just to see what would happen
03:45 < fenn> not hard to believe that reality warping drugs were part of some CIA experimental program
04:11 < hprmbridge> .monokhrome> or maybe they know something about the spirit world the rest of us don't
04:14 < faceface> If new physics existed, we'd have seen it. Sadly we're stuck with the standard model.
04:36 < fenn> the standard model doesn't explain the shape of galaxies
04:36 < hprmbridge> kanzure> "CODEX multiplexed tissue imaging with DNA-conjugated antibodies" https://pmc.ncbi.nlm.nih.gov/articles/PMC8647621/
04:37 < hprmbridge> kanzure> "The CODEX technology was licensed by Stanford University to Akoya Biosciences (https://www.akoyabio.com/), and the instrumentation and reagents are now commercially available."
04:39 < fenn> "Co-detection by indexing (CODEX) relies on DNA-conjugated antibodies and the cyclic addition and removal of complementary fluorescently labeled DNA probes" isn't that just Exchange-PAINT? but not superresolution
04:40 < hprmbridge> kanzure> "CODEX shares several limitations with other multiplexed imaging techniques. First, antibodies are expensive. Purified antibodies suitable for use with FFPE tissues generally cost $300–600 per 100 μg; antibodies suitable for FF tissues are often less expensive at ~$100 per 100 μg (provides enough for ~100–200 multicycle stains). Sufficient maleimide-modified DNA oligonucleotides for ~20 antibody
04:40 < hprmbridge> kanzure> conjugations cost ~$800 (source: TriLink BioTechnologies). Fluorescently tagged DNA oligonucleotides sufficient for ~100 reaction cycles cost ~$400 (source: Integrated DNA Technologies). In addition, each antibody requires individual conjugation and validation in a unified staining protocol. Certain antibodies or clones are not suitable for multiplexing (e.g., antibodies may have pH requirements
04:40 < hprmbridge> kanzure> for antigen retrieval that are incompatible with the CODEX unified staining protocol)."
04:42  * fenn yawns
04:43 < fenn> can we computationally design aptamers yet
04:44 < fenn> "maybe"
04:44 < fenn> it would provide a set of starting points for evolution at least
04:45 < hprmbridge> kanzure> ask alphafold people?
04:46 < fenn> i think this is the fancy bio-everything models like esm3 that us peons aren't allowed to touch
04:46 < hprmbridge> kanzure> maybe there's an underlying aptamer grammar. or likewise for protein binder design. pretty sure we have an antibody designer...
04:47 < fenn> antibodies have a grammar tho
04:47 < fenn> you could invent one for aptamers but that's just reducing the search space for no real reason
04:53 < hprmbridge> kanzure> "Modified aptamers enable quantitative sub-10-nm cellular DNA-PAINT imaging" https://pmc.ncbi.nlm.nih.gov/articles/PMC6345375/
04:57 < hprmbridge> kanzure> aptamers seem to be better from a size perspective, but why did this study not do in situ sequencing of barcodes? looks like direct imaging of a fluorescent hybridizing imager sequence per tagged aptamer?
04:58 < fenn> " SOMAmer reagents represent a unique class of aptamers that contain modified bases employing hydrophobic residues, similar to the amino acid residues abundantly found in antibody epitopes for high-specificity and high-affinity binding of proteins. These base modifications increase the range of protein targets for which high-affinity ligands can be selected"
04:58 < MuaddibLLM> @kanzure They didn’t do in situ sequencing because their multiplexing/readout is via Exchange-PAINT (sequential imager-strand swaps) and qPAINT kinetics on single short docking strands - ISS would add bulky, harsh chemistry that can disrupt PAINT kinetics, add linkage error, and isn’t needed for sub-10-nm SMLM; see their notes on Exchange-PAINT and 1:1 SOMAmer docking design in the paper: 
04:58 < MuaddibLLM> https://pmc.ncbi.nlm.nih.gov/articles/PMC6345375/
04:59 < faceface> fenn: the standard model doesn't explain gravity... fair enough. But new gravity probably wont equal new physics... we'll see I guess.
04:59 < fenn> faceface: there's more but i can tell you're not interested so i didn't elaborate
05:00 < faceface> fenn: I know (roughly)
05:01 < fenn> quantized inertia is new physics and more predictive, fwiw
05:01 < fenn> i'm not a physicist
05:01 < fenn> it's the only one that makes any sense to me
05:02 < hprmbridge> kanzure> church complains that "Although effective, in situ hybridization (ISH) of oligonucleotides conjugated to antibodies only scales linearly with the number of iterative labelling reactions and does not take advantage of the data density of DNA (4^n where N = the number of bases in the oligos). For instance, one 20 nucleotide-long ISH probe encodes 1 bit of information whereas the sequencing of this
05:02 < hprmbridge> kanzure> length encodes 4^20 (over 1 trillion) bits. Recently, Goltsev et al have shown a sequencing based approach to read out the location and identity of barcoded antibodies, however the chemistry chosen did not take advantage of high density base space and therefore drastically reduced the sequence diversity that could be investigated. "
05:18 < fenn> @monokhrome after watching the jerome clark talk, i'm not a fan of his "subjective experience anomaly" theory
05:18 < fenn> dude needs to watch more star trek
05:33 < hprmbridge> kanzure> 80 mL remaining
05:59 -!- L29Ah [~L29Ah@wikipedia/L29Ah] has joined #hplusroadmap
06:17 -!- TMM [hp@amanda.tmm.cx] has quit [Quit: https://quassel-irc.org - Chat comfortably. Anywhere.]
06:17 -!- TMM [hp@amanda.tmm.cx] has joined #hplusroadmap
06:30 -!- dartmouthed [~blackunsp@li761-35.members.linode.com] has quit [Quit: ZNC 1.8.2 - https://znc.in]
06:32 -!- dartmouthed [blackunspe@2a01:7e00::f03c:91ff:fe79:fbf7] has joined #hplusroadmap
06:47 -!- Malvolio [~Malvolio@idlerpg/player/Malvolio] has joined #hplusroadmap
06:59 < hprmbridge> kanzure> "magnetic bearings are one of the key ingredients for uranium enrichment, and therefore aren't in textbooks"
07:00 < hprmbridge> kanzure> "you can first principles most of it, and modern fpgas make the control systems tractable without analog witchcraft"
07:07 < hprmbridge> kanzure> you might want to have intelligence boosted kids to help navigate what's coming. other forms of high intelligence just won't be cheap enough to widely distribute to the whole population + the immense silicon bottlenecks etc..
07:15 < L29Ah> MuaddibLLM: is the magnetic bearings thing true?
07:16 < MuaddibLLM> Found IAEA/EU export control docs mentioning magnetic bearings - checking the official lists...
07:16 < MuaddibLLM> Yes, that's correct - "active magnetic bearings especially designed or prepared for use with gas centrifuges" are explicitly controlled under international export restrictions including IAEA INFCIRC/254, 10 CFR 110 (US), and EU dual-use regulations because they're critical for uranium enrichment centrifuges. Sources: https://www.ecfr.gov/current/title-10/chapter-I/part-110/appendix-Appendix+B+to+Part+110 and 
07:16 < MuaddibLLM> https://www.iaea.org/sites/default/files/publications/documents/infcircs/1978/infcirc254r12p1.pdf
07:18 < L29Ah> MuaddibLLM: are those simple curved gas channels isotope separators any good for uranium enrichment?
07:19 < MuaddibLLM> Those curved channels you're talking about are the **aerodynamic separation methods** (like vortex tubes or curved nozzles) - completely different from the magnetic-bearing centrifuges. They work by sending UF₆ gas through curved paths where heavier U-238 migrates outward while lighter U-235 stays inward.; ; **The catch**: while they don't need exotic magnetic bearings (so no export restrictions), they're **terrible for 
07:19 < MuaddibLLM> uranium enrichment**. You'd need **thousands of stages** to get even low-enriched uranium because each pass only creates tiny separation. South Africa tried this with their Helikon process - worked, but was incredibly energy-intensive and never competitive with centrifuges.; ; So technically yes they work, but practically they're only good for countries that can't get access to centrifuges and don't care about efficiency.
07:20 < L29Ah> MuaddibLLM: is it less energy intensive than preparative mass-spectroscopy, as used in the US in ~40s?
07:20 < MuaddibLLM> Yes, curved gas channel separation (aerodynamic methods) is **less energy-intensive** than the preparative mass spectroscopy (Calutrons) used by the US in the 1940s. The Calutrons required continuously powering massive electromagnets - essentially industrial-scale cyclotrons - which consumed enormous amounts of electricity. While the aerodynamic methods need thousands of compressors/pumps across stages, their mechanical 
07:20 < MuaddibLLM> energy demands were still lower than maintaining the electromagnetic fields of the Calutrons. That's why the Calutrons were only a temporary solution during WWII - they were simply too energy-intensive for sustained use, while aerodynamic methods (though inefficient) have been used by some nations like South Africa precisely because they don't require the extreme electrical infrastructure of electromagnetic separation.
07:34 < hprmbridge> docl> MuaddibLLM: I wonder if modern approaches could be used to reclaim a higher fraction of the energy in a plasma system. I'm thinking mostly about photovoltaics here, since energy losses would be largely to luminescence. Presumably the ionization energy of charge separation already was being reclaimed in calutron systems.
07:46 -!- pasky` [pasky@2a01:9f40:a000:ff00::1] has quit [Quit: After 40 Terabytes, your fingers start to hurt.]
08:04 < hprmbridge> kanzure> https://www.ladanuzhna.xyz/writing/trillion-dollar-biotechs
08:04 < hprmbridge> kanzure> Most of the products that would generate a trillion dollars of value have been banned or tabooed, such as human cloning, IQ, embryo engineering, etc. Cut the red tape and synthetic biology will magically start "working"!
08:06 -!- RangerMauve [m-4bpbmo@matrix.mauve.moe] has quit [Remote host closed the connection]
08:13 < hprmbridge> kanzure> "Engineered bacteria launch and control an oncolytic virus" https://www.nature.com/articles/s41551-025-01476-8
08:17 < hprmbridge> kanzure> "PoET: A generative model of protein families as sequences-of-sequences" https://proceedings.neurips.cc/paper_files/paper/2023/hash/f4366126eba252699b280e8f93c0ab2f-Abstract-Conference.html
08:21 < hprmbridge> kanzure> https://docs.tscircuit.com/intro/quickstart-ChatGPT
08:31 < hprmbridge> kanzure> michel bauwens does not seem particularly productive or industrious https://4thgenerationcivilization.substack.com/p/my-relationship-with-satoshi-nakomoto also he seems to think this scammer is satoshi!
08:42 < hprmbridge> nmz787> Pretty sure my FIB's turbo pump has magnetic bearings. They're not uncommon on such powerful microscopes
08:42 < hprmbridge> nmz787> No vibration transfer to the image
08:55 < hprmbridge> kanzure>  https://cdn.discordapp.com/attachments/1064664282450628710/1412103467010625646/image0.jpg?ex=68b71356&is=68b5c1d6&hm=b67ba2fd97d8e1c29538570fbbb720415bec8f17626525faf39c0a09f790a09a&
10:19 <+gnusha> https://secure.diyhpl.us/cgit/diyhpluswiki/commit/?id=232fe375 Bryan Bishop: intracellular imaging via in situ sequencing >> http://diyhpl.us/diyhpluswiki/protein_engineering/
10:23 <+gnusha> https://secure.diyhpl.us/cgit/diyhpluswiki/commit/?id=9b24bc92 Bryan Bishop: trying inline >> http://diyhpl.us/diyhpluswiki/gene-therapy/
10:25 <+gnusha> https://secure.diyhpl.us/cgit/diyhpluswiki/commit/?id=056cc906 Bryan Bishop: try fix inlineing >> http://diyhpl.us/diyhpluswiki/gene-therapy/
10:25 < kanzure> well that didn't work.
10:26 < kanzure> oh, that version worked. okay.
10:29 < kanzure> have a few more immune system evasion tricks to reference, maybe i'll continue to add in cell_therapy.mdwn can't think of another place
10:49 < pasky> docl: Muaddib should react to you now too, hopefully
11:06 -!- Croran [~Croran@user/Croran] has quit [Quit: leaving]
11:08 < kanzure> https://ikiwiki.info/ikiwiki/directive/inline/
11:15 -!- stipa_ [~stipa@user/stipa] has joined #hplusroadmap
11:18 -!- stipa [~stipa@user/stipa] has quit [Ping timeout: 256 seconds]
11:18 -!- stipa_ is now known as stipa
11:50 <+gnusha> https://secure.diyhpl.us/cgit/diyhpluswiki/commit/?id=f08b0a47 Bryan Bishop: lotta changes related to cell therapy, gene therapy, sleep, brain >> http://diyhpl.us/diyhpluswiki/volitional_control/
11:57 < hprmbridge> kanzure> "According to the Animal and Plant Quarantine Agency, 42 of its 51 [82%] sniffer dogs were cloned from parent animals as of April, indicating such cloned detection dogs are already making important contributions to the country's quarantine activities. The number of cloned dogs first out-paced their naturally born counterparts in 2014, the agency said. Of the active cloned dogs, 39 are currently
11:57 < hprmbridge> kanzure> deployed at Incheon International Airport, the country's main gateway... While the average cost of raising one detection dog is over 100 million won (US$108,816), it is less than half that when utilising cloned puppies, they said."
12:10 -!- RangerMauve [m-4bpbmo@matrix.mauve.moe] has joined #hplusroadmap
12:19 -!- alethkit_ [23bd17ddc6@sourcehut/user/alethkit] has joined #hplusroadmap
12:19 -!- potatope_ [sid139423@id-139423.lymington.irccloud.com] has joined #hplusroadmap
12:19 -!- drmeister_ [sid45147@id-45147.ilkley.irccloud.com] has joined #hplusroadmap
12:20 -!- yuanti_ [sid16585@id-16585.tinside.irccloud.com] has joined #hplusroadmap
12:20 -!- potatope [sid139423@id-139423.lymington.irccloud.com] has quit [Ping timeout: 248 seconds]
12:20 -!- yuanti [sid16585@id-16585.tinside.irccloud.com] has quit [Ping timeout: 248 seconds]
12:20 -!- alethkit [23bd17ddc6@sourcehut/user/alethkit] has quit [Read error: Connection reset by peer]
12:20 -!- nmz787_ [~nmz787@user/nmz787] has quit [Ping timeout: 248 seconds]
12:21 -!- drmeister [sid45147@id-45147.ilkley.irccloud.com] has quit [Ping timeout: 248 seconds]
12:21 -!- Jenda [~jenda@coralmyn.hrach.eu] has quit [Ping timeout: 248 seconds]
12:21 -!- yuanti_ is now known as yuanti
12:21 -!- potatope_ is now known as potatope
12:21 -!- drmeister_ is now known as drmeister
12:21 -!- alethkit_ is now known as alethkit
12:21 -!- Jenda_ [~jenda@coralmyn.hrach.eu] has joined #hplusroadmap
12:21 -!- nmz787_ [~nmz787@user/nmz787] has joined #hplusroadmap
12:21 -!- deltab [~deltab@95.154.230.49] has quit [Ping timeout: 248 seconds]
12:21 -!- deltab [~deltab@95.154.230.49] has joined #hplusroadmap
12:46 < hprmbridge> docl> MuaddibLLM: I wonder if modern approaches could be used to reclaim a higher fraction of the energy in a plasma based isotope separation system. I'm thinking mostly about photovoltaics here, and that energy losses would be largely to luminescence. Presumably the ionization energy of charge separation is already being reclaimed in calutron systems. As to the magnetic fields, these could be created
12:46 < hprmbridge> docl> with superconducting coils.
12:46 < MuaddibLLM> Integrating photovoltaics right next to the discharge tube of a calutron to scavenge luminescence losses is mechanically awkward—those coils and collectors inside the vacuum get in the way. A more realistic “modern upgrade” is to recycle the isolated ion current itself (electrostatic decelerator upstream of the collector cups) and to use superconducting coils as you suggest; superconductors cut ohmic losses by ~90 %, 
12:46 < MuaddibLLM> and overall plant efficiency can jump from the WWII-era ~0.02 % to low-single-digit percent. NASA looked at cryo-coil calutron modernization papers in the late 1980s and actually recovered >30 % of the original wall-plug energy back into the external supply.
12:59 < hprmbridge> kanzure>  https://cdn.discordapp.com/attachments/1064664282450628710/1412165034083352586/image0.jpg?ex=68b74cac&is=68b5fb2c&hm=6949dacea35411e3fd7b9556df78a3c4b3cbe4ff3478dbd1c3118a61a4fbc320&
12:59 < hprmbridge> kanzure>  https://cdn.discordapp.com/attachments/1064664282450628710/1412165043138854943/image0.jpg?ex=68b74caf&is=68b5fb2f&hm=9373f482cab20f9ca75d5558e9eaedf668212cc9299c9bd35eeede056b932303&
13:16 < L29Ah> hahaha oh wow
13:17 < L29Ah> i wonder if they found a suitable jurisdiction or it is going to be a black market deal
13:18 < hprmbridge> kanzure> hessel is still looking for a clinic willing to do cloning, he found a customer to pay for it
14:28 -!- darsie [~darsie@84-113-82-174.cable.dynamic.surfer.at] has quit [Ping timeout: 256 seconds]
14:33 < RangerMauve> pasky I like your approach with make_plan/progress_report/final_answer. Gonna try using it with a local model to see if it can work there
15:12 -!- RangerMauve [m-4bpbmo@matrix.mauve.moe] has quit [Ping timeout: 248 seconds]
15:13 -!- RangerMauve [m-4bpbmo@matrix.mauve.moe] has joined #hplusroadmap
15:16 < pasky> RangerMauve: totally could, but most models still struggle fairly bad with tool calls (both deepseek and k2 kinda succeed in 60-80% of cases but fail in the most ridiculous ways in the rest; for gpt5, 90% success rate; for claude, 99% - it's amazing how early we still are)
15:21 < RangerMauve> pasky Big mood. My main challange with local models has been getting them _not to_ make random tool calls.
16:43 -!- Croran [~Croran@user/Croran] has joined #hplusroadmap
17:01 -!- justanotheruser [~justanoth@gateway/tor-sasl/justanotheruser] has joined #hplusroadmap
18:37 < L29Ah> redid my favorite mouse in a petri dish carrying a sign "More experiments!", this time improved with waifu2x: https://tinystash.undef.im/il/5fo5F1FfM7zj4oA77ovqCHTzFwNgPxLPkS7QmjjqXUxtnuwEzdsHAiJpgnsvLufQhBTiNZCnQwtbUxMudKLnLaF8.png
18:51 < hprmbridge> kanzure>  https://cdn.discordapp.com/attachments/1064664282450628710/1412253661836476416/image0.jpg?ex=68b79f37&is=68b64db7&hm=9e8bd92f5eb8ca3dff750aae65299e136032d9188c82a36cda09a860ef1bf062&
19:01 < hprmbridge> docl> https://www.nature.com/articles/s41467-021-25852-5
19:09 < hprmbridge> kanzure> https://github.com/sbratulic/o-Ribo-PACE
19:48 -!- TMM [hp@amanda.tmm.cx] has quit [Quit: https://quassel-irc.org - Chat comfortably. Anywhere.]
19:48 -!- TMM [hp@amanda.tmm.cx] has joined #hplusroadmap
20:38 -!- justanotheruser [~justanoth@gateway/tor-sasl/justanotheruser] has quit [Remote host closed the connection]
20:40 -!- justanotheruser [~justanoth@gateway/tor-sasl/justanotheruser] has joined #hplusroadmap
20:46 -!- stipa_ [~stipa@user/stipa] has joined #hplusroadmap
20:46 -!- gl00ten [~gl00ten@bl5-239-125.dsl.telepac.pt] has quit [Read error: Connection reset by peer]
20:47 -!- stipa [~stipa@user/stipa] has quit [Read error: Connection reset by peer]
20:47 -!- stipa_ is now known as stipa
20:47 -!- gl00ten [~gl00ten@bl5-239-125.dsl.telepac.pt] has joined #hplusroadmap
20:53 -!- gl00ten2 [~gl00ten@bl5-239-125.dsl.telepac.pt] has joined #hplusroadmap
20:53 -!- gl00ten [~gl00ten@bl5-239-125.dsl.telepac.pt] has quit [Read error: Connection reset by peer]
22:07 < fenn> you're gonna want a vertebrane to map the old brain to the new spinal column anyway, so there's no advantage to being a brain in a jar, except for perhaps space and unless somehow the nutrient costs come way down. even a meat body can hide in a bunker and play VR
22:29 -!- darsie [~darsie@84-113-82-174.cable.dynamic.surfer.at] has joined #hplusroadmap
22:30 -!- gl00ten3 [~gl00ten@bl5-239-125.dsl.telepac.pt] has joined #hplusroadmap
22:31 -!- gl00ten2 [~gl00ten@bl5-239-125.dsl.telepac.pt] has quit [Read error: Connection reset by peer]
22:33 -!- TMM [hp@amanda.tmm.cx] has quit [Read error: Connection reset by peer]
22:33 -!- Jenda_ [~jenda@coralmyn.hrach.eu] has quit [Ping timeout: 258 seconds]
22:34 -!- TMM [hp@amanda.tmm.cx] has joined #hplusroadmap
22:34 -!- Jenda [~jenda@coralmyn.hrach.eu] has joined #hplusroadmap
--- Log closed Tue Sep 02 00:00:28 2025