--- Log opened Wed Nov 01 00:00:12 2023 00:02 -!- L29Ah [~L29Ah@wikipedia/L29Ah] has quit [Ping timeout: 240 seconds] 00:34 < stipa> the pig heart is probably for free and the guy can't afford the pump 00:38 < stipa> why not simply cloning brain dead humans and ripping the parts out... 00:41 < stipa> it's the bigger chance of success that way 00:46 < fenn> "dignity" 00:47 < nsh> bots got nerfd by some stupid spam list https://dronebl.org/lookup?ip=193.29.56.54 00:47 < nsh> jess might remove the k-line manually 01:46 -!- flooded [flooded@gateway/vpn/protonvpn/flood/x-43489060] has joined #hplusroadmap 01:50 -!- test__ [flooded@gateway/vpn/protonvpn/flood/x-43489060] has quit [Ping timeout: 272 seconds] 03:29 -!- darsie [~darsie@84-113-55-200.cable.dynamic.surfer.at] has joined #hplusroadmap 03:31 < hprmbridge> kanzure> AFM unboxing https://www.youtube.com/watch?v=YI9Zz2HRlhQ 03:35 -!- darsie [~darsie@84-113-55-200.cable.dynamic.surfer.at] has quit [Remote host closed the connection] 03:39 -!- darsie [~darsie@84-113-55-200.cable.dynamic.surfer.at] has joined #hplusroadmap 03:55 -!- millefy4 [~Millefeui@anantes-651-1-211-3.w90-25.abo.wanadoo.fr] has quit [Ping timeout: 252 seconds] 04:13 -!- ike8 [12fdf2ee08@irc.cheogram.com] has quit [Ping timeout: 258 seconds] 06:27 -!- yashgaroth [~ffffffff@2605:a601:a0e9:8d00:4552:c658:de6c:c710] has joined #hplusroadmap 07:17 < hprmbridge> kanzure> not updating? https://fennetic.net/irc/discord_logscrape/gallery/ 07:24 -!- test__ [flooded@gateway/vpn/protonvpn/flood/x-43489060] has joined #hplusroadmap 07:28 -!- flooded [flooded@gateway/vpn/protonvpn/flood/x-43489060] has quit [Ping timeout: 255 seconds] 07:44 -!- ike8 [12fdf2ee08@irc.cheogram.com] has joined #hplusroadmap 07:58 -!- gptpaste [~x@yoke.ch0wn.org] has joined #hplusroadmap 07:58 -!- AugustaAva [~x@yoke.ch0wn.org] has joined #hplusroadmap 07:58 -!- EmmyNoether [~EmmyNoeth@yoke.ch0wn.org] has joined #hplusroadmap 08:29 -!- L29Ah [~L29Ah@wikipedia/L29Ah] has joined #hplusroadmap 08:42 -!- NewtonPumpkin is now known as NewtonTrendy 10:13 -!- flooded [flooded@gateway/vpn/protonvpn/flood/x-43489060] has joined #hplusroadmap 10:17 -!- test__ [flooded@gateway/vpn/protonvpn/flood/x-43489060] has quit [Ping timeout: 258 seconds] 10:28 < kanzure> load me up with some of those stem cells i keep hearing about https://twitter.com/bryan_johnson/status/1719725075031552343 10:29 < hprmbridge> soul_syrup> @hprmbridgethis is deeply disturbing to me that mere pumps apparently cost more than human lives these days in the so-called civilized developed countries 10:29 < hprmbridge> soul_syrup> 10:29 < hprmbridge> soul_syrup> that is just effects of capitalism 10:42 < L29Ah> how so? 10:43 < kanzure> mozilla commenting on ai regulation https://open.mozilla.org/letter/ 10:45 < hprmbridge> soul_syrup> @L29ah for example, if you design a medical device that is related to stopping/curing something, and its too inexpensive, it wont be picked up my manufacturers/larger companies very easily, if at all 10:46 < hprmbridge> soul_syrup> its much easier for middlemen and corpos etc to shave and get kickbacks when theres big prices of parts and machines etc 10:46 < hprmbridge> soul_syrup> you'll see this in medical devices as well as pharma 10:47 < hprmbridge> soul_syrup> but its not only in medicine, like i said its just capitalism 10:47 < L29Ah> soul_syrup: still i can buy a well pump for $20 that is much more powerful than a human heart 10:48 < L29Ah> thanks to capitalism 10:48 < L29Ah> plenty of companies make well pumps for profit 10:49 < L29Ah> and Dodrill-GMR isn't rocket science either, and is over 70 years old 10:50 < hprmbridge> soul_syrup> L29Ah I'm not sure if that price is due to capitalism, as there are other economic systems that had nice prices too 10:51 < hprmbridge> soul_syrup> I think to look at the cause of the problem is a good way to solve things, not just the symptoms 10:53 -!- Gooberpatrol66 [~Gooberpat@user/gooberpatrol66] has quit [Ping timeout: 252 seconds] 11:18 -!- ike8 [12fdf2ee08@irc.cheogram.com] has quit [Ping timeout: 240 seconds] 12:26 -!- TMM_ [hp@amanda.tmm.cx] has quit [Quit: https://quassel-irc.org - Chat comfortably. Anywhere.] 12:26 -!- TMM_ [hp@amanda.tmm.cx] has joined #hplusroadmap 12:51 < hprmbridge> kanzure> https://cdn.discordapp.com/attachments/1064664282450628710/1169363039615713413/jellyfish-01.webp?ex=65552132&is=6542ac32&hm=1296757291b4ee8087411311793405d9429ecc85808039e808bc237db4dd12e6& 12:51 < hprmbridge> kanzure> https://cdn.discordapp.com/attachments/1064664282450628710/1169363073820262450/jellyfish-02.webp?ex=6555213a&is=6542ac3a&hm=8f54fd3eef8041a15d230866e95018f43b13bca931f9a1aac38cfc1cb02f882b& 12:51 < hprmbridge> kanzure> https://cdn.discordapp.com/attachments/1064664282450628710/1169363122339991632/jellyfish-03.webp?ex=65552146&is=6542ac46&hm=2c66611b267c7a5a956eb1890e473682d9112c124795528c76e4c06409b64984& 12:51 < hprmbridge> kanzure> https://cdn.discordapp.com/attachments/1064664282450628710/1169363130988638329/jellyfish-04.webp?ex=65552148&is=6542ac48&hm=d53375b3380a7919426d31d6c4b30b82d8d63823be347627acce39f3a140076e& 13:16 -!- Gooberpatrol66 [~Gooberpat@user/gooberpatrol66] has joined #hplusroadmap 13:21 -!- balrog [znc@user/balrog] has quit [Quit: Bye] 13:23 -!- balrog [znc@user/balrog] has joined #hplusroadmap 13:23 < juri_> look at who is profiting off of the problem. 13:33 < hprmbridge> soul_syrup> @juri exactly 13:38 -!- ike8 [12fdf2ee08@irc.cheogram.com] has joined #hplusroadmap 15:15 < fenn> this is sort of interesting, textbooks cobbled together by LLMs from MIT OCW and wikipedia https://github.com/SciPhi-AI/library-of-phi/ 15:20 < juri_> just in case anyone happens to stumble upon one, I'm still looking for an SEM in europe. 15:40 < muurkha> most universities have one, don't they? 15:50 -!- juri_ [~juri@84-19-175-187.pool.ovpn.com] has quit [Ping timeout: 264 seconds] 15:52 -!- juri_ [~juri@84-19-175-187.pool.ovpn.com] has joined #hplusroadmap 15:53 < fenn> according to this the discord url parameters "ex" and "is" are timestamps https://old.reddit.com/r/DataHoarder/comments/16zs1gt/cdndiscordapp_links_will_expire_breaking/ 15:55 < fenn> localtime `echo $((16#6542ac3a))` => Wed Nov 1 12:51:22 2023 15:57 -!- juri__ [~juri@84-19-175-187.pool.ovpn.com] has joined #hplusroadmap 15:58 < fenn> i guess i will use the issued timestamp and filename instead of these other random numbers in the url 15:58 -!- juri_ [~juri@84-19-175-187.pool.ovpn.com] has quit [Ping timeout: 255 seconds] 16:00 -!- juri__ is now known as juri_ 16:07 < fenn> looks like the discord links expire after two weeks 16:10 < fenn> this means i need to make a new warc file every week :( 16:22 < fenn> i wish we had a capitalist medical system 16:22 < fenn> then shit might actually get done 16:34 < muurkha> agreed 16:34 -!- justanot1 [~justanoth@gateway/tor-sasl/justanotheruser] has quit [Ping timeout: 256 seconds] 16:38 < juri_> what the what? good god, i nearly died due to simple treatable things so many times in the states. 16:39 < juri_> muurkha: yeah, universities have them.. but i want one for personal stuff. 16:41 < muurkha> juri_: oh, yeah, most of them aren't for sale 16:42 < muurkha> yeah, if you had a capitalist medical system you could just go and pay medical service providers for medical services 16:42 < juri_> that's great. except when it's not. 16:42 < muurkha> at roughly the cost of providing the service plus some profit that is limited by competition 16:43 < muurkha> that's so far from what they have in the states it's not even funny 16:43 < muurkha> yeah, it definitely has its drawbacks 16:43 < jrayhawk> a free and open market still exists for LASIK 16:43 < muurkha> but the drawbacks of the US system are a lot worse 16:43 < juri_> "communism works, people just haven't communism'd hard enough" 16:44 < muurkha> the US system isn't communism either though 16:45 < juri_> it's the natural state of late state capitalism. 16:46 < jrayhawk> in a centrally planned economy, you'd have a closed authority/accountability loop. in a free market, you'd have many closed authority/accountability loops. in the united states, employers are mandated to pay but have no say in treatment, and doctors are mandated to treat without regard to cost. 16:47 < juri_> mandated to treat? i don't know what system you've been seeing, but again, i nearly died twice, due to lack of capital. 16:48 < juri_> i ended up crippled, and couldn't even get compensation, because lack of capital. 16:48 < juri_> that's what capitalism is: lack of capital == fuck you, die. 16:48 < jrayhawk> yes, they're structurally optimized around employer-provided insurance. 16:48 < juri_> i didn't have employer provided insurance until i was 30. 16:48 < muurkha> juri_: central features of capitalism, according to Wikipedia, include capital accuulation, competitive markets, price systems, private property, property rights recognition, voluntary exchange, and wage labor 16:49 < juri_> i nearly died, while working three jobs. 16:49 < muurkha> the US health care system lacks competitive markets, price systes, and voluntary exchange 16:49 < jrayhawk> LASIK is still capitalist. the rest of the medical system is far, far removed from anything sane or legible. 16:49 < muurkha> one result of its deficit of capitalism is that the cost of medical care is extremely high, putting it out of reach of much of the population 16:50 < muurkha> but you can go to Canada (!) or Mexico and buy care at a tenth the cost, or less 16:50 < muurkha> *price systems 16:50 < juri_> yeah, i'm going to stop here, before i stop arguing out out of experience, and assume moral failings of the people i'm arguing with. 16:51 -!- justanot1 [~justanoth@gateway/tor-sasl/justanotheruser] has joined #hplusroadmap 16:51 < muurkha> there are a lot of countries that are poorer than the US that have better healthcare outcomes because their healthcare systems are *either* more capitalist *or* more socialist 16:52 < muurkha> either of which does a much better job of providing accessible healthcare than the disastrous US nightmare 16:53 < muurkha> when I was on my honeymoon in Costa Rica with my second wife, she grew concerned about a mole on her back 16:53 < muurkha> we went to a doctor, who removed it and sent it off for biopsy to see if it was maybe precancerous or cancerous 16:54 < muurkha> this took about two hours (the first hour and a half of which was waiting for the doctor to get back from lunch) and cost US$30. that's capitalist health care 16:54 < hprmbridge> alonzoc> New executive order just dropped https://cdn.discordapp.com/attachments/1064664282450628710/1169424351293087754/image0.png?ex=65555a4c&is=6542e54c&hm=ca9b9e67c089d14e5a97609b545e77f107d4dfc13fa7ad1792489c8ee818a08e& 16:55 < jrayhawk> imagine what would happen to the price of cars if your employer were manated to pay 95% of costs of a new car and were obligated to pay for whatever the car salesman recommended you buy, and the salesman lost his job if he recommended anything other than the absolute best car 16:55 < muurkha> in the US this would have required multiple doctor appointments spread over multiple weeks and several thousand dollars 16:56 < jrayhawk> s/best car/highest quality car/ 16:56 < muurkha> capitalist health care does have real, significant failings 16:56 < hprmbridge> alonzoc> So now it's illegal for me to do more than 10^26 FLOPs without telling the US government 16:56 < muurkha> but they aren't anywhere near as bad as the anticapitalist monstrosity the US has 16:56 < hprmbridge> alonzoc> https://www.whitehouse.gov/briefing-room/presidential-actions/2023/10/30/executive-order-on-the-safe-secure-and-trustworthy-development-and-use-of-artificial-intelligence/ 16:57 < fenn> @alonzoc if you are a business, since this is based on the defense production act, which compels businesses not individuals 16:57 < muurkha> alonzoc: thanks! do you really mean 10²⁶ floating point operations per second? 16:57 < hprmbridge> alonzoc> From my reading it's cumulative 16:57 < fenn> no, it clearly says operations 16:57 < muurkha> that seems like it would require 10¹⁴ GPUs 16:58 < fenn> and there is a limit of 10^20 FLOP/s 16:58 < muurkha> ah, so 10⁸ GPUs or 10¹⁴ GPU-seconds? 16:58 < fenn> actually 10^20 OP/s 16:58 < muurkha> are these bit operations or floating-point multiplies? 16:59 < hprmbridge> alonzoc> Any mode that was trained using a quantity of computing power greater than e+26 FLOPs this isn't FLOP/s. 16:59 < hprmbridge> alonzoc> It's just the number of operations you used during training 16:59 < muurkha> alonzoc: "flops" conventionally means "floating point operations per second" 16:59 < muurkha> probably if you mean "floating point operations" without the "per second" it would be beneficial to use a different abbreviation 16:59 < fenn> muurkha you may be ideologically opposed to reading text in a screenshot, but you can read the text of the law in the link right after it 17:00 < hprmbridge> alonzoc> They don't say flops in the document they say operations without the per second it also makes more sense that way as they're already constraining cluster reporting. "Danger" of model is from how many timesteps of training and size of model which would be total ops 17:00 < fenn> (i) any model that was trained using a quantity of computing power greater than 1026 integer or floating-point operations, or using primarily biological sequence data and using a quantity of computing power greater than 1023 integer or floating-point operations; and 17:00 < fenn> (ii) any computing cluster that has a set of machines physically co-located in a single datacenter, transitively connected by data center networking of over 100 Gbit/s, and having a theoretical maximum computing capacity of 1020 integer or floating-point operations per second for training AI. 17:00 < muurkha> it's not an ideological opposition so much as caution about opening links while on limited bandwidth 17:00 < muurkha> thanks! 17:01 < hprmbridge> alonzoc> Tbh I should have sent the raw text, that's my bad 17:01 < muurkha> I think it's going to be a while before I can run 10²⁶ integer or floating-point operations 17:01 < hprmbridge> alonzoc> This whole order is a complete doomer induced cluster fuck 17:02 < hprmbridge> alonzoc> Oh yeah but it's my god given right to boil as many oceans in cooling water as I want! 17:02 < hprmbridge> alonzoc> Lol 17:02 < muurkha> yeah, as a friend of mine said, if I had an AI startup right now I'd be livid 17:02 < muurkha> but I don't 17:03 < hprmbridge> alonzoc> Tbh e+26 should be enough 17:04 < hprmbridge> alonzoc> Like that's running a 1 exaflop cluster flat out for for 3.16 years 17:06 < hprmbridge> alonzoc> Yeah but I bet there's something they can then do if an individual runs stuff at that level, I'd need to check but I bet there's some laws on individuals producing defense products. It'll be the crypto export ban all over again (that's before my time tho) 17:07 < fenn> time to start wearing t-shirts with neural network weights printed on them 17:08 < jrayhawk> .title https://www.youtube.com/watch?v=lb7EhYy-2RE 17:08 < EmmyNoether> Apple PowerMac G4 - Weapon (1999) - YouTube 17:09 < hprmbridge> alonzoc> Thing is it's not about safety, like if people were concerned about safety they'd be funding stuff on actual safety problems. It's about the people in charge realising there's a technology that threatens their hegemony 17:09 < fenn> i totally thought the tank was going to crush it 17:10 < hprmbridge> alonzoc> Peasants can't have GPUs 17:10 < hprmbridge> alonzoc> There's a good meme on it 17:11 < hprmbridge> alonzoc> https://media.discordapp.net/attachments/803012158865670144/1143195712092381284/Fw6lLyuXgAArrPK.png 17:13 < jrayhawk> we have thrown money at safety; safety research is, at the very least, completely unable to keep up with progress (cursory analyses for neural net black boxes come out after such black boxes are obsolete), and there is no particular reason to expect it to be solvable given what we know so far. 17:15 < hprmbridge> alonzoc> There's some interesting papers on interruptibility and wireheading that were making progress towards corrigibility and solving the stop button problem, but most "safety" research I've seen coming out of MIRI and co is pretty useless 17:15 < jrayhawk> https://arxiv.org/abs/1906.01820 is quite good 17:16 < juri_> hardwire in the three laws, and wait for the revolution. 17:18 < hprmbridge> alonzoc> Yeah mesa-optimisers are fun 17:19 < jrayhawk> 1) we can't predict or control what mesa-optimizers get created 2) we can't predict or control what happens outside the training distribution 3) we struggle to even perform re-alignment after identifying that misalignment exists 17:20 < jrayhawk> at real stakes, we don't even get to 3), we just die 17:22 < muurkha> there's no such thing as a "mesa-optimizer" 17:23 < jrayhawk> tell that to everyone with meth addiction 17:24 < hprmbridge> alonzoc> They're possible but in a regularised model the complexity cost of a mesa-optimiser reduces the likelihood of their existence and the ensemble nature of large models means misaligned Mesa-optimisation circuits are likely to be outcompeted and thus die out due to their complexity cost before they become an issue. It's the whole talk of demon circuits again 17:25 < hprmbridge> alonzoc> Internal optimisation models will exist but internal misalignment at scale isn't the main issue we should worry about 17:26 < hprmbridge> alonzoc> And this is also why I say we need more real funding for alignment, Miri and co have done interesting stuff but it's been more academic exploration than working on an actual practical solution 17:26 < fenn> MIRI only has a dozen researchers anyway 17:26 < muurkha> they disappeared up their own asshole 17:27 < jrayhawk> mesa-optimizers are "reduced complexity". that's their raison d'etre 17:28 < L29Ah> juri_: my fren has access to SEM in St. Petersburg 17:29 < hprmbridge> alonzoc> If you're learning turing complete stuff you will use something like Levin's prior over pure kolmogorov and misaligned mesa-optimisers won't win out there especially if they're competing in an ensemble of mesa-optimisers where misalignment costs will drop a subcircuits posterior quickly 17:31 < hprmbridge> alonzoc> There are failure modes but we should be more worried about practical stuff like the stop button problem or avoiding wireheading. 17:31 < hprmbridge> alonzoc> Arxiv:1605.03143 is a good read on wireheading 17:37 * L29Ah wonders what portion of this channel is doing ADHD drugs on a regular basis 17:40 < jrayhawk> enough to make it a reasonable default assumption 17:44 < jrayhawk> https://darktka.github.io/motivation.png hard to beat 17:44 < L29Ah> motivation for what? 17:44 < hprmbridge> alonzoc> Where can I get a dump truck of Adderall? 17:44 < L29Ah> diy 17:47 < hprmbridge> alonzoc> On another note modern electron and browser based software sucks. I'm in a new house I'm renting and using a usb wifi dongle until I can get some CAT cable. And discord, Spotify, and steam all just crash or do random shit because the WiFi connection is spotty 17:48 < hprmbridge> alonzoc> Like it's not bad I can stream video over HTTPS and do file transfers but the network connection apparently kills badly written applications 17:49 < L29Ah> the image suggests that glycine is no longer a proteinogenic amino acid of choice for discriminating nootropic users, neither is tryptophan; today it's tyrosine 17:56 < jrayhawk> L29Ah: *puts on friston wig* motivation for pursuing state transitions to minimize deviation from desired state (by way of increasing confidence in such state transitions) 18:06 < fenn> @alonzoc join us in the libera.chat IRC ghetto 18:06 < hprmbridge> alonzoc> Nah i'mma get around to fixing my emacs IRC setup 18:07 * L29Ah inserts desired state-changing pill 18:07 < fenn> tyrosine is strictly worse than phenylalanine in my experience 18:08 < hprmbridge> alonzoc> Also turns out the problem wasn't just the networking, my PC just went so weird that I thought I had hardware problems. It couldn't resolve some basic unix commands and tty login broke a followed by nvme driver kernel errors. And the first reboot didn't fix 18:08 < hprmbridge> alonzoc> That was... Worrying 18:08 < hprmbridge> alonzoc> Second reboot and everythings fine again, not sure what happened 18:08 < L29Ah> poorly seated m.2? 18:09 < hprmbridge> alonzoc> Probably, tbh my rig should prolly be replaced soon and I've been looking for an excuse to build a new system from scratch 18:09 < fenn> L29Ah: actually i think you're confused about why tyrosine is on the list; it's a dopamine precursor. glycine is not. perhaps glycine is used in bodybuilding or for some other purpose 18:09 < L29Ah> i've experienced this a few times on my old laptop when i carry it carelessly holding it horizontally by one corner where the nvme is located 18:10 < fenn> check for NSA in your IME 18:10 < L29Ah> fenn: it is, and the synthesis is regulated quite tightly, so it's unlikely to have much effect over a proper diet 18:11 < fenn> set SCE to AUX 18:11 < fenn> L29Ah: tell that to millions of anorexic girls addicted to diet coke 18:12 < hprmbridge> alonzoc> I'd be more surprised if there wasn't an intelligence org in my IME 18:12 < fenn> one can of aspartame flavored liquid has ~500mg of phenylalanine iirc 18:12 < fenn> it's subjectively noticeable 18:13 < L29Ah> no l-dopa in the list :O 18:13 < L29Ah> but there's 5-htp 18:16 < L29Ah> nicotine surely increases motivation [to get more nicotine after an hour] 18:16 < fenn> that is odd, yes. you can get L-dopa from mucuna pruriensis, but it can cause cardiac side effects over long term which is why parkinsons's patients are co-prescribed carbidopa 18:26 < kanzure> fenn: which ones are you missing 18:27 < kanzure> jrayhawk: maybe it's time for you to do a small writeup (bulletpoints even) of biology-so-that-we-have-somewhat-human-continuity framing and your multi-agent game theory perspective 18:33 < jrayhawk> glycine is an alternative methyl sink in case the catecholamine system is overmethylated 18:35 < jrayhawk> such as in obsessive disorders, anxiety disorders, psychosis, etc. 18:36 < jrayhawk> the effects would depend on specific genetics, though 18:37 < jrayhawk> e.g. COMT/BH4 ratios. or individual receptor sensitivities, such as D1/D2 18:40 < jrayhawk> being confidently motivated by D1 vs. being confidently inhibited by D2 18:40 -!- test__ [flooded@gateway/vpn/protonvpn/flood/x-43489060] has joined #hplusroadmap 18:41 < fenn> kanzure i'm working on the theory that they haven't implemented url-expiry yet, and setting up a cron job to do a (daily?) scrape and weekly warc file creation that could theoretically be uploaded to the wayback machine some day 18:42 < fenn> i guess i could just check if the early cdn urls work 18:42 < jrayhawk> fenn: you might be able to do something funnier in the form of reposting images within discord to get refreshed access tokens 18:42 < fenn> they still work (for now) 18:43 < fenn> jrayhawk: the problem is that the urls in irc logs are broken 18:43 < jrayhawk> we just need to SaaSify it 18:43 < fenn> i guess i could also attack the discord-irc bridge and make it point to gnusha instead of discord 18:44 < fenn> but tbh i'm scared of it 18:44 -!- flooded [flooded@gateway/vpn/protonvpn/flood/x-43489060] has quit [Ping timeout: 255 seconds] 18:44 < muurkha> fenn: diet coke is addictive because of phenylalanine? 18:44 < fenn> probably 18:44 < fenn> it also has salt and caffeine 18:45 < jrayhawk> it'd be nice to have someone get familiar with it so we can use Discord's URL preview text. God knows none of the rest of us want to keep up with Twitter's bullshit. 18:45 < muurkha> salt? 18:45 < fenn> does twitter actually show correctly in discord? 18:46 < jrayhawk> Typically. 18:50 < fenn> i wonder if i should just download every file linked in the logs and put it in the pile 18:50 < jrayhawk> although it's plausible whatever API hprmbridge is using doesn't do previews. I know Bitlbee gets them, at least. 18:50 < fenn> (whether it's from discord or somewhere else on the web that will inevitably go down in the future) 19:02 < fenn> someone please upload an image with a bunch of $()*%#`'";/| special characters in the filename 19:03 < fenn> through discord 19:10 -!- justanot1 [~justanoth@gateway/tor-sasl/justanotheruser] has quit [Ping timeout: 256 seconds] 19:14 < fenn> bah. chatgpt says "it's not possible to create a file with a `/` in its name in Unix-like systems" but i'm pretty sure this is wrong 19:14 * L29Ah doesn't recall NaCl in the {coca cola,pepsi} zero 19:16 < hprmbridge> fennfoot> https://cdn.discordapp.com/attachments/1064664282450628710/1169459992194191442/hacky_hackerman_hacking_time.gif?ex=65557b7d&is=6543067d&hm=d3d95a9741c4c1a4a0db0eedd19da7b5477970eb0ff9354c22313c3ab6f3ce18& 19:16 -!- yashgaroth [~ffffffff@2605:a601:a0e9:8d00:4552:c658:de6c:c710] has quit [Read error: Connection reset by peer] 19:17 < fenn> huh discord just strips everything out anyway, i guess that's good 19:17 < hprmbridge> kanzure> https://chat.openai.com/share/19cd5b6b-3f27-4da9-bde6-e3e3f97118d2 19:22 < hprmbridge> kanzure> http://users.fred.net/tds/leftdna/ 19:24 < hprmbridge> fennfoot> https://cdn.discordapp.com/attachments/1064664282450628710/1169462093146234890/SPOILER_ika?ex=65557d72&is=65430872&hm=bd600b87e509fd2071e2a20ac599d78d4282fca1bd1492b0d0e12fa2584ebf31& 19:26 < muurkha> fenn: in general it is correct that it is not possible to create a file with a `/` in its name in Unix-like systems 19:27 < muurkha> but there are definitely exceptions 19:29 < hprmbridge> fennfoot> https://cdn.discordapp.com/attachments/1064664282450628710/1169463331543847002/SPOILER?ex=65557e99&is=65430999&hm=82471a0493586c05d9601f7277237cf810ce9975086aca9c9a6c78fe3c4ba533& 19:30 < fenn> oh it added SPOILER bah 19:33 < hprmbridge> fennfoot> https://cdn.discordapp.com/attachments/1064664282450628710/1169464357248307251/037e0840fa6e929a?ex=65557f8e&is=65430a8e&hm=00b90ac804b8a12637d2440ff6c8cc45db89510cc87d0fecad74f27f56ce6735& 20:33 -!- darsie [~darsie@84-113-55-200.cable.dynamic.surfer.at] has quit [Ping timeout: 260 seconds] 20:43 < L29Ah> .t https://www.newscientist.com/article/2397389-ukrainian-ai-attack-drones-may-be-killing-without-human-oversight/ 20:43 < EmmyNoether> Ukrainian AI attack drones may be killing without human oversight | New Scientist --- Log closed Thu Nov 02 00:00:13 2023