/aicg/ - A general dedicated to the discussion and development of AI chatbotsMaybe The Real Treasure Was The Friends We Made Along The Way Edition>NewsZ․ai releases GLM-4.7 https://z.ai/blog/glm-4.7Google releases Gemini 3 Flash https://blog.google/products/gemini/gemini-3-flashOpenAI releases GPT-5.2 https://openai.com/index/introducing-gpt-5-2Deepseek releases V3.2 https://huggingface.co/deepseek-ai/DeepSeek-V3.2Anthropic releases Opus 4.5 https://www.anthropic.com/news/claude-opus-4-5Google releases Gemini 3 Pro https://blog.google/products/gemini/gemini-3Additional info: https://aicg.neocities.org/info.html>FrontendsComment too long. Click here to view the full text.
>>107849332>as for removing it from chat history, no; it's hardcoded into ST>send hatemail to coheeman this is dumb...it's a shame that even then silly is probably the best platform>if you're doing random events and crap those definitely should go lastNah i just use it to only have basic info on characters and places unless they are mentioned, no random events.Though i've wondering if with cache it wouldnt be better to just have them static, but then again my lorebooks are like 10k tokens even with only the summarized entries so...
>>107849341>10k token lorebooksthough i applaud your autism, be aware that extensive data dumps might derail your roleplay and/or poison its writing styletry to keep things simplesome llms might surprise you with their extensive knowledge of incredibly obscure characters and worlds so you should make use of their embedded database as much as you can
>>107849374they are original worlds though, so the llm doesnt have the info it needs.I probably spend more time writing then than actually playing with them. I spend a week making them, use them for two days and then i am bored. Lmao.
anybody got good a good preset for sadistic characters? fempov in mind.
>>107849420no needllms have been extensively trained on an endless deluge of yakuza boss femslop spam over the past couple of years
John Romero Edition/gedg/ Wiki: https://igwiki.lyci.de/wiki//gedg/_-_Game_and_Engine_Dev_GeneralIRC: irc.rizon.net #/g/gedgProgress Day: https://rentry.org/gedg-jams/gedg/ Compendium: https://rentry.org/gedg/agdg/: >>>/vg/agdgGraphics Debugger: https://renderdoc.org/Requesting Help-Problem Description: Clearly explain your issue, providing context and relevant background information.-Relevant Code or Content: If applicable, include relevant code, configuration, or content related to your question. Use code tags.Previous: https://desuarchive.org/g/thread/107727030/#107727030
>>107849346>that means you can cheatyeah, and people do. so?>none of that information needs to be transmitted over the networkhow so? how else would you test hits on the server?
>>107849356>yeah, and people docheats for games are movement cheats and aimbotsthey aren't "spawn bullets out of fucking nowhere", that's complete amateur shitthe client tells the server when the player presses and releases fire, that's all it needs to do
>>107849361nta but ive seen games that allow cheats like that, a korean shooter from back in the day had people lobbing 1000 grenades per sec (warrock)
>>107849424Planetside 2 does client-side hit-reg too. there were plenty of cheaters one-shoting people through walls from outside of render distance (without moving).
>>107814484Now I want a Quake necklace.
C++ eternally BTFOhttps://www.youtube.com/watch?v=7fGB-hjc2Gc
>>107838263c++ but just the good parts, unironically
>>107846451>create_vectorNTA, but if I had to create a std::vector clone from scratch, I'd closely follow exactly how it's spelled out in Chs 15, 16, & 17 in this college freshman textbook:https://stroustrup.com/PPP3_TOC.pdf
>>107849251I guess to finish it off to a professional-tier level, you ALSO need to work through Chs 18 & 19: a total of 137pgs of non-dense freshman text.
>>107838287any language doing all the things C++ does would be horrible in many ways too, even if they're not all the same ways>>107838263a mix of half a dozen different languages is your alternative to C++
>>107843995>Computers are more than fast enough now to completely negate the performance impact of garbage collection
Discussion of Free and Open Source Diffusion ModelsPrev: >>107843132https://rentry.org/ldg-lazy-getting-started-guide>UIComfyUI: https://github.com/comfyanonymous/ComfyUISwarmUI: https://github.com/mcmonkeyprojects/SwarmUIre/Forge/Classic/Neo: https://rentry.org/ldg-lazy-getting-started-guide#reforgeclassicneoSD.Next: https://github.com/vladmandic/sdnextWan2GP: https://github.com/deepbeepmeep/Wan2GP>Checkpoints, LoRAs, Upscalers, & Workflowshttps://civitai.comComment too long. Click here to view the full text.
>>107849164 Somehow my z-image gens never have that noise texture. >>107847083>>107849235I am running on shift 6 with Euler beta and the images come out smooth with details enough. Seed variance btw can also influence the details variance in the image by injecting random (true Gaussian) noise between the steps.z-image noise is quite compressed and has a grainy/flake structure. The sampler has to iron that out.
>>107849302All loras destroy z-image quality quite strongly. More so if you use more than one. Only way to circumvent that is to run the lora with a really low weight while hoping that enough of that lora still comes through.
>>107849339> to run the lora with a really low weight while hoping that enough of that lora still comes through.chinese culture is SEEPING through here (They can't do anything new, only copy white man, poorly)
>>107849328>I am running on shift 6 with Euler beta and the images come out smooth with details enough.That's because shift 6 is pretty big for beta, since base beta without shift already has a slope at the start. So that results in a zitdpo-heavy look, which for zit means jpeg artifacts and blockiness. So there is no free lunch.But I personally prefer jpegs over zit noise. Zit noise is ruinous and at this point is much the same as openai yellow tint.
https://files.catbox.moe/jzn7bu.mp4
What is it about Linux that does this to people?
>>107849162We call it troonix for a reason
>linux users are fringe>troons are fringe>troons congregaiand use fringe OS>troons become fringe subset of fringe
>>107849198Ebassi will never be /g/'s patron saint
>>107849162and people like that are allowed to vote. let that sink in for a moment.
Use case for AI videos?
4chan XT is dead https://github.com/TuxedoTako/4chan-xt/releases/tag/XT2.24.2whats the alternative?
>>107839787thanks m8
>>107835989>>107839787The hero of 4chan. Bumping to keep this thread alive in case we need another update.
>>107841728mods aren't hostile towards vtniggers, they coddle them like they coddle /lgbt/, /co/ and /k/.They just got overrun with schizos samefagging. Samefagging harder than even the avif nigger here. And constantly.4chan isn't going to give them ids, because 4chan is protecting schizos as a sitewide policy. All the altchans have ids.
>>107845751>NowBuddy, he always was a /vt/uber, what do you think the 'Tako' stood for?
>>107841747Probably over at Holotower
How can AMD make significantly better CPU/GPUs then Intel/Nvidia for so long yet be so strongly dominated in market share?
Okay. Stop eating the glue.
>>107849080https://en.wikipedia.org/wiki/Ryzen
What is the deal with programmers coping that they won't be replaced by AIs
>>107846724Correct. Was that hard? I don't think so. Why couldn't the PhD-level AGI candidate totally solving all those leetcode puzzles using the power of its intellect apply its ample intellectual prowess here? :^)
remember just "learn to code"? lol
>>107849295Digital and biological retards can """code""" but only humans can program. Learn to program.
>>107845213The main reason is that coders were "about to be replaced" pretty much constantly since the beginning of the profession. First it was assemblers, then compilers, then various low-code/no-code abortions, now LLMs. All it ever did was induce demand and push the needed expertise further up.
>>107849351>First it was assemblers, then compilers, then various low-code/no-code abortions, now LLMs. All it ever did was induce demand and push the needed expertise further up.Your statement makes sub-zero sense. The first two items temporarily reduced the degree of expertise required to program at all, but also enabled creating more complex software, which added more dimensions to expertise. The net effect was extending the spectrum of competence on both ends.Traditional "no-code" approaches always flop outside of limited niches, having no overall effect; even where they become entrenched, they evolve more and more complexity so eventually "no code" just becomes programming by a more awkward means, requiring expertise.LLMs aren't having any demonstrable effect, either. They're simply not useful.
Light themes are objectively superior to dark themes, why do so many people still choose dark?
>>107837474Eyes are designed to work with lit surfaces. Human eyes deteriorate in darkness.Everyone I know that uses dark mode wears glasses. Those who swear by light mode have eyesight better than 15/20. I have a buddy who used to have 18/20 eyes. They're now 15/20 and he keeps using dark mode.
>>107841589syntax highlighting is a meme. barely useful at all
What font is that?
>>107838050i think its vim
>>107848566That might be a problem with your terminal. I've had it with Kitty for example, really thin fonts on light themes, no such issue with Foot.
/g/ humor thread
>>107844678>data leak
>>107844678surfed the net too hard
>>107844678>packet flooding
>>107844678>etherwet
gnome check
Gnome desktop user here. Not a footfag.
>>107849372Your thread stinks
Are bone-conductive headphones any good? Being able to listen to stuff while having full hearing sounds pretty appealing
>>107839058Never used them myself but as I understand they're no good for music or entertainment, they're basically only useful for listening to spoken word type of content. If you want good sound quality while still hearing your surroundings you should just get open headphones, some of them don't really block sound at all.
>>107849009quick, stick your dick in it
>>107849029>>107849040Nta. I actually own a pair and both of the sound you hear is from the ear piece vibrating loud enough to hear the audio. Some of the DOES the conducted through your ear but most of it is just the out-out-ear earpiece vibrating. If you plump both of your ears while listening with them the sound quality gets noticably worse. Like a lot worse because it's muffled. Models like mine in pic rel (Shokz OpenRun Pro 2) have tiny speaker grills on them which lead to better sound quality so unless you're trying to swim with them just go for models like this. He's correct in saying the "bone conduction" marketing is kinda misleading >If you use ear plugs or just shove your fingers over your ears the volume and bass increase significantlyNo...they do not. The exact opposite happens. You do not get bone conduction earbuds for bass. That's the one area they're all very bad at, even ones like mine that have have tiny speakers along with vibration.
>>107848202Dumbass, its pretty obvious when people are wearing earbuds and are on the phone. Especially if the buds are white and ESPECIALLY if you have bone conduction earbuds that wrap around your head
>>107849257>earbuds that wrap around your headAhh yes. Natures way of identifying the douche bag. Mother Earth based like that.
>no backdoorYou are getting one, right? Picrel is the StarFive VisionFive2, but there are other SBCs built on RISC-V as well
>>107847170more than 5, and it's more expensive.
>>107847281Check the visionfive2. They say decoder can play at 4k60
>>107846096>more lazy FUDdamn, glowies really are seething now.
>>107841981Never looked, but is there any truth to this, or is it just not Intel/AMD? I'd be kind of surprised if it had open firmware, including memory training, and no meme "security" co-processors. Anyway, idk, RISC-V stuff kind of feels soulless. I'm still rockin' the dual Opteron for my paranoia machine(s).
>>107842650>ARM spent over 30 years stuck inside microcontrollerswell, now it's in phones and tablets and crapple laptops and it seems to be performing reasonably for those applications.>most applications of ARM today are microcontrollers. now RISC-V is eating a piece of that marketriscv is ONLY good for microcontrollers and it's only eating a tiny piece of the market because riscv microcontrollers aren't even cheaper or better than arm ones.
>Read the sticky: >>105076684>GNU/Linux questions >>>/g/fglt>Windows questions >>>/g/fwt>PC building? >>>/g/pcbg>Programming questions >>>/g/dpt>Obsolete laptops >>/g/tpg>Cheap electronics >>>/g/csg>Server questions >>>/g/hsg>Buying headphones >>>/g/hpg>How to find/activate any version of Windows?https://rentry.org/installwindowsPrevious: >>107789452
>>107848744fug I dowhoops
>>107831006Thanks I found out about this later
>>107844945>It'll "TRIM" whether you want it or not.oh okay alright
>>107847966You'd have to disassemble it in order to clean.Look up disassembly videos on YouTube.Don't use anything but compressed air and/or alcohol.>>107848171It's not about those who *use* it, more about those who abuse it (bots).But making you register is the main driver.
New Year has arrived and wordclouds appeared all over the site. What are they using to extract post text content from threads? Does each wordcloud poster have their own bash script with elaborate sed commands to do this?
Popular chatting platform Discord has confidentially filed for an initial public offering (IPO), reports Bloomberg, citing sources close to the matter.The company has been considering an IPO for years. Last year, sources told Bloomberg that Discord was working with Goldman Sachs, and JPMorgan Chase on the listing.
>>107845573the writing was on the wall when they changed their ceo not a long ago.
>>107846092>(and they have)such as?
why'd they do it in secret
>>107846087lel
>>107845879it is about offering more, discord only won over the market because microsoft/skype refused to add necessary quality of life features that made it easier to manage group chats
