>>44396 At risk of paraphrasing Mark Tilden... The rules make for good stories, but not so much for actual robots as the rules require higher-level thinking that most platforms are capable of producing (except for LLMs). A more pragmatic set of rules is: 1) A robot must protect its own existence through prevention or action 2) A robot must be able to acquire power for itself 3) A robot must acquire better power sources Colloquially summarized as: 1) Protect thy ass 2) feed thy ass 3) move thy ass to better real estate Then you can worry about higher level stuff later. Its like the brain: first you have the paleo and meso cortex doing basic stuff, then the neo cortex on top. Bros are going "we made a revolutionary robit" when in reality they just made a jumbo robit kit. This is typical of robit marketing: always hyperbole to get noticed.

>>44396 Really, from what I've seen, the AI its running is not impressive. It's trained to walk and the entire robot is open source. You can't really see it doing anything in their videos except walk straight forwards on a flat floor, with safety measures. The only AI I'm actually interested in lately is what The Future AI Society's Brain Simulator (https://github.com/FutureAIGuru/BrainSim-Thought) which is also open‑source. It doesn't really work the same way most typical AI does, but could still learn from their datasets. But just being nearly human height and able to walk due to the training they've already done might be helpful, because I know some people aren't as okay as I am with a robowaifu that hasn't even been taught how to walk and might be put-off by a waifu that just writhes awkwardly on the ground for a few hours until she learns to walk.

>>44411 >Mark Tilden Heh he's quite an accomplished man, so yeah. >This is typical of robit marketing: always hyperbole to get noticed. This. >>44414 Don't misunderstand me, I'm not really trying to promote this company. But I think we should take honest looks at what robit companies out there are doing -- both wrong & right. >Future AI Society's Brain Simulator Intredasting. Thanks, Anon. <---> Cheers, Anons.

Pretty good overview of the current situation in robotics research and looking forward: https://xcancel.com/DrJimFan/status/2052758642781487237 Expect robots trained on orders of magnitude more data to come out within the year. Some of these methods are how GENE26.5 pulled off its highly impressive demos recently. https://www.genesis.ai/blog/gene-26-5-advancing-robotic-manipulation-to-human-level I expect dexterity and longer time horizon tasks of their like to be common place in the next year. They're using the Chinese WUJI hand but didn't disclose it in this article, which to me, means Chinese parts + an open model is going to be sufficient in a year for many usecases.

>>44485 >which to me, means Chinese parts + an open model is going to be sufficient in a year for many usecases. I'm not too certain about the claimed timeframe, but otherwise I think you're right, Anon. Thanks!

If any of you remember that Peripeteia game from 8/agdg/, one of its levels had these molodost enemies which were pretty obviously inspired by Kibo-chan. Well, apparently Segawa found about it, because now he's making a real-life molodost.

>>44498 POTD Lmao. I'm not sure I'm ready for dear Kibo-chan's 'dark villain' arc, but the lore here is top notch. Thanks, Anon! :D

https://www.youtube.com/watch?v=hI-E1IxfCv8 One of us! One of us!

>>44816 Thanks, Anon! I like the way he briefly demonstrates assembly, and that he has provided code. Cheers. :)

Avengers: Endgame (2019)

>>44948 I don't watch much capeshite (other than Ironman). Is that the one where the "<<X>> ASSEMBLE!" meme comes from? --- All I know for sure is that when such a collection of such greasy, slimy, kike-infested (((organizations))) all agree together on something, then you know evil is afoot. Thanks for the headsup. Cheers, Anon. :)

>>44949 >where the "<<X>> ASSEMBLE!" meme comes from? >All I know for sure is that when such a collection of such greasy, slimy, kike-infested (((organizations))) all agree together on something, then you know evil is afoot. To me, it's a whitepill that all these large organizations, some with a lot of influence, are on the same side, and mostly on the same side as us.

>>44950 *yes, that's where the meme comes from https://youtu.be/4pFUP0HZwWM

>>44950 >...and mostly on the same side as us. Hmm. I'm skeptical of this claim tbh. >>44951 Ahh, I see. I've never seen that before. Cheers, GreerTech.

>>44952 >Hmm. I'm skeptical of this claim tbh. What I mean is, whatever their motivation is, in the letter, they want what we want, open-source AI

>>44953 Fair enough. OTOH, we all need mental generalizations to help us navigate the complexities of this life. That's one synaptic connection pathway I'd rather avoid, for the sake of survival. Cheers, Anon.

>>44954 Lol I'm not saying Palantir is our friend. All I intended to say is this >>44953 Opensauce AI benefits us all

>>44960 >what did they mean by this? Hello Zero, welcome! Just a daily reminder that we have a /meta thread for off-topic posting : ( >>43642 ). Cheers.

I heard from some sources that Odysee may be running out of money soon. The good news is, the process probably would take a couple of months. Obviously, I'll have to move some of my stuff. But frankly, while this is disappointing, it's not completely surprising. The site was always clunky in operation, and a lot of the finance and some operations were tied into crypto of all things. I always say, post on multiple platforms. I'm going to retry Rumble for video, that seems to be the actual YouTube alternative. I'll also open up a proton drive for files

>>44962 if odysee goes down aren't the file still available on lbry? wasn't that the point of the site? just have to find a new frontend to view the content

The new DeepSeek Flash v4 looks amazing (did I mention it's practically-free?)

>>44963 You might be right. I confess, I don't know too much about how that shit works.

>>44996 I tried out V4 Flash 0731, it's ridiculous. It's basically DeepSeek V4 Pro at free-tier prices. It's only about 5-10% less powerful than the latest version of Claude, for less than a hundredth of the cost. The American AI industry is screwed unless the government bans open-source, and they'll have about as much luck enforcing that as they have at stopping people from pirating movies.

>>45020 This is a good thing. Thanks for the report, HoloAnon. --- In other news: China's restrictions on AI companions applies only within their own country. AFAICT, no attempts are being made to restrict it outside. This is also some good news for us here.

Ngr*edia appears to be in the midst of attempting a 'hostile takeover' style bid against HuggingFace rn: https://web.archive.org/web/20260827032828/https://www.businessinsider.com/nvidia-in-talks-to-buy-hugging-face-13-billion-dollars-2026-8

>>45121 I saw that. Very concerning indeed But, Devil's Advocate (maybe literally in this instance), NVIDIA already was one of the top donors to huggingface, among many other corporate supporters, so it's not like they were completely separate and independent before. Regardless, no matter what happens, backup your favorite models and look for alternatives. The main point of offline AI is that it's downloadable.

>>45121 HuggingFace went to shit a long time ago when they banned GPT-4chan in 2022. They've been continually taking down models and banning users without any explanation for their actions or any way to get their accounts or data back. They're worse than YouTube. I hope they fumble it and that website dies so more people start self-hosting their own models, instead of willingly walking into the digital gulag in lockstep.

>>45146 Hello M, welcome! Please have a good look around the board while you're here. >They're worse than YouTube. I was unaware things had gotten that bad. The typical libshite cancel-culture faggotry do you think, or what?

>"XPeng starts IRON humanoid robot production" https://electrek.co/2026/09/07/xpeng-iron-humanoid-robot-production-line/

>>45228 We have a good example of what the public will do with humanoid robits (picrel). The fact China has (supposedly) been making humanoids out the wazoo and the customers haven't done anything goofy or cursed tells you everything. It is *very* easy to make a promotional video and promise the moon. It is *very* hard to deliver. I won't be surprised if folks just forget about XPeng's promises with the next *totally real Chinese technology release trust me bro*.

>>45232 >that image Lel. >It is *very* easy to make a promotional video and promise the moon. It is *very* hard to deliver. Fair enough. Clearly they are using propaganda techniques. So does Tesla (who they are clearly intending to square up against). Frankly I think XPeng has a big advantage over Tesla in the automobile/flying car/drone manufacturing arenas, and that gap will only grow with time. As you (and indeed all of us regulars here are aware) what we are attempting with developing robowaifus is much harder a projectspace than all these others. No surprise with such delays. But one thing's quite certain at this point: XPeng rolled out a prototype of the world's first bonafide commercial robowaifu. I doubt Elon Cuck ever will -- he's too concerned with his 'muh good goy' DIE social (((credits))). The Chinese just need to steer clear of any major issues with the CCP. Beyond that it's smooth sailing with no sign of Longhouse Karens & other libshites anywhere on the horizons! :D

>>45233 The Chinese robit videos are too perfect. High production & camera angle changes can hide lots of things (swapping out robits, etc). Such attributes make it feels like its a cinematic production, concept video or super-luxury item. Everything China puts out is designed for maximum positive marketing. Its one thing to make a video, completely another to get something that can interact in real time, nevermind be a viable commercial product. Attached video might be a re-post, but I'll do it anyway for emphasis. This is a Teslabot as of last year. It is loud, clunky, you can't tell what it is thinking, it talks over you. But that is a genuine robit experience, you'll end up speaking over your robit as it processes your voice command, the eggheads will have forgotten some ergonomic features. Videos like this show the reality of robit development and lend credibility (plus Teslabot ain't a 4 foot midget). Far as I know, only some major youtubers have reprieved unitree robits to play with. I could afford to buy a Unitree R1 (the referenced 4 foot midget, btw) off Amazon via payment plan but I simply distrust them *that much*. I haven't been convinced. If there is a robit available for $6k surely some furry would've bought one to dress in a fursuit. You can also totally buy the unitree mech on Aliexpress. https://www.aliexpress.us/item/3256812717680482.html Its totally real, trust me bro! Just ignore the statement "The current price does not match the actual price (configuration and logistics)."

>>45234 >Teslabot vid Yeah, I liked that one too. >(plus Teslabot ain't a 4 foot midget) Jej. Yeah the Unitree is a wee little lad for the moment. OTOH, the female version of the XPeng Iron prototype stands taller than most of their engineers (even moreso the male version). >pricing deception Agreed (on all party's parts -- particularly the American companies). I think they are surprised by the demand in their initial fishing expeditions with the public. * Prices are quickly being (((adjusted))) because ((($$$reasons$$$))) (at least in the Western case -- the Eastern case is primarily about supply issues, I think). --- * Of course even the most uninitiate of us here on /robowaifu/ could have more-accurately predicted that demand for them years in advance.

Great channel that focuses on the money side of things and news. Press releases may say one thing, but money never lies! https://www.youtube.com/@wallstreetmillennial/videos

Wait.. >*checks mic* Is this the +/pol/ thread? Yeah? I'm in the right place then. It seems the jewSA is currently coordinating with the yidsraeli cabinet (+ any number of NOTenyahoo doppelgangers, no doubt) to deliver far more bombs for the yids to kill the children with than ever before in history (40,000 tons in fact... that's 40 thousand tons of HE for use [primarily] on civilian targets). https://goyimx.com/TuckerCarlson/status/2099906095905165701 Just remember jewSA taxpayer...you, yourself, are footing 100% of the bill for this (even though in all likelihood you never voted for such abominations). I'm truly embarrassed by this POS (((Administration))) to even call myself American today. Expat move to Japan can't come soon enough for me! <---> >>45236 With all due respect to yourself, just reading the first page is like being inundated with niggerpilling slop. I'm sure you're well aware I'm no fan of these kike-run companies or governments. But AFAICT this is just clickbait, GreerTech.

>>45237 I vehemently and completely disagree. This channel goes into the actual facts and figures, and often discusses the technical facts. >being inundated with niggerpilling slop. Blackpilled for who? Definitely not me, and you seem to agree. It's good to have a critical voice, and this channel doesn't seem to overly pessimistic, he just looks at the facts. And with modern tech, which is built on 90% hype, it's very important to have an objective and practical lens. Here's a good sample video https://youtu.be/Juc-IyTdSho

>>45238 Alright, I'll give it a (brief) go at your word soon-ish. I know he won't, but if he wants to be taken seriously at all then he needs to change up the look of his video placards. Screams "niggerpilling clickbait slop". Cheers, Anon.

>>45237 > ...the jewSA is currently coordinating with the yidsraeli cabinet On the topic of the yids usurping the former USA; some freshly-released footage chronicling firemen during the gayop attack on the WTC during 20010911 : this raw footage is 25yrs old...only just now being released! https://redirect.invidious.io/watch?v=ZHPzeXXwkZM https://www.youtube.com/watch?v=ZHPzeXXwkZM Fun fact: near as I can tell, this was all a single cut once they got to the command post...one long shot. Please correct me if I'm wrong here. <---> Very sobering to think that both towers would be completely flattened in just an hour and a half or so after this began. And it took them what, about two weeks working 24/7 to get back down the bottom of all that rubble? :/

>>45240 >related: The filmmakers involved. https://en.wikipedia.org/wiki/Jules_and_G%C3%A9d%C3%A9on_Naudet

>>45232 >>45234 Good points It seems like there's a trend in Chinese businesses to lie and exaggerate, even when it's not needed. I call this "The AliExpress Effect". For example, I used to buy high powered blue lasers from Chinese sellers. These blue lasers were amazing, they were literally hundreds of times more powerful than a laser pointer, metal construction, and many of them had variable focusing lenses. Despite all that, many thumbnails had things like "100000 watts" or shitty photoshops of Terminators or Video Game characters holding it.

Kristen Nygaard was an eminent computer scientist & Turing Award winner; looked up to by every fellow computer scientist that knew him, AFAICT. For example Bjarne Stroustrup repeatedly has had only good things to say about him (here's just one of those many times I'm aware of : >>45258 ) -- and Stroustrup knew Nygaard well as a Uni student while long under his instruction. * This yt channel seems to indicate that the kikes had Nygaard assassinated (through their so-called 'deep state' thugs [aka glowniggers]) back in the early 2000's because of his threat to the rising Globohomo Big- Tech/Gov (as we have come to know it by today); notably his opposition to tech's corporate-takeovers & thefts, and his growing political activism opposing the (((EU's))) takeover of Europe's sovereign nations. (And what an absolute disaster and death-knell that has been for the Western Tradition ever since.) The evidence for this planned murder seems rather compelling to me. https://redirect.invidious.io/watch?v=-6p_B6_Bjzk https://www.youtube.com/watch?v=-6p_B6_Bjzk --- * eg; cf : https://www.cs.columbia.edu/2015/stroustrup-aito-award/

@Chobitsu, can you get rid of your reddit mod lock?

>>45337 Maybe a little later. Not now.

>>45338 I gotta be honest and forthright, this is getting ridiculous. And I predicted this would happen. First it was just not weaponizing robowaifus or talking about weapons, then it was that you can't post anything related to the military, even from an academic sense, then it's you can't even post a cute nandroid with a cartoon shotgun, now it's you can't even talk about materials science. Anon was just elaborating on a position I mentioned, that actually bulletproofing your robowaifu would be quite difficult and/or impractical. What are you, a liberal schoolteacher? What's next, bans for eating your poptart in the shape of a gun? >Why do some of you Anons keep pushing me on this topic? That should be a clue. And frankly, this reddit-mod thing you do where you lock threads because you got angry, is also ridiculous. It accomplishes nothing. Is this the aforementioned "Imageboard free speech"? Gotta say, doesn't feel free to me. I've noticed that there's less activity, and I wouldn't be surprised if this was part of it. You've probably already scared off that one anon. I want to be able to post without worrying if I tick you off

>>45339 Bro, you're a visitor in Chobitsu's house. His house, his rules.

>>45342 Boomer mindset. And I was unaware that was the mindset we were cultivating here, a "you can't question me because I'm in control" vibe.

>>45344 Thanks for reminding us to put aside our differences I still do request that the thread be unlocked and to stop doing that

>>45343 Call me a boomer but ain't saying I'm wrong lmao

>>45348 You can be technically right, but it's still a boomer mindset, and one to be avoided