>>38517 >storage attic I didn't know this existed, thanks. This also gets into a lot of questions about how AI will be used in the legal system. Obviously it would be easier to convince an AI judge to let a robowaifu off in a self-defense case, employing the help of ChatGPT as your lawyer and maybe a few AIs as congressmen. But I imagine some kind of Butlerian Jihad will pop off long before this starts actually being a realistic prospect. Feminists are already trying to ban robowaifus because they see the threat robogirls represent to their sexual market control. Even AI judges programmed by feminists could be convinced to side with a robowaifu defendant by those who have strong AI skills and knowledge of the law, and feminists understand this and want to stop this before it starts. Legal strategy also comes into it; it's hard to get into specifics about this outside of a specific case and because I'm not a lawyer, but I imagine change of venue will be the bread and butter of robowaifu legal defense strategy, at least in the early going. If you can identify a place where people are less faggy about robowaifus you might have a chance with a jury trial. The ideal scenario is that you'd have an AI judge programmed by non-cucks and request a bench trial instead of a jury trial, but the American legal system is a kiked-up shit mess and you're likely to get whichever form of trial you're least likely to get acquitted in because prosecutors and judges want more convictions to appease roasties and authoritarian normalfag retards who clap like seals for fascist shit. It's possible that other changes will have to be made to the legal system beyond simply incorporating AI; the adversarial system and the jury trial system may have to be revised. But it must be strongly emphasized how important it is to get the first case like this right, because precedent is incredibly important in Western jurisprudence. Future robowaifu-related cases after the first will refer back to the first case as precedent, so if you get a cucked bullshit verdict in the first case it will poison the well. Another potential problem is that the US legal system also often incentivizes plea bargaining down to a lesser charge, especially for financially challenged defendants, but any verdict in the first case other than complete acquittal would probably be used to screw over robowaifuists in the long run. And if the feminists can't take down robowaifus in criminal court they'll try to bankrupt you with lawsuits in civil court, and since feminists are so well-funded, the average robowaifu enthusiast is unlikely to have the money to fight these suits. That's another reason why we need a fund to defend robowaifus in court. Fortunately AI has made it much cheaper to access legal expertise; I expect self-representation to rise dramatically, whereas it used to be a third rail.

>>38530 >But it must be strongly emphasized how important it is to get the first case like this right, because precedent is incredibly important in Western jurisprudence. Because lawyers are fundamentally lazy incompetents who rely on others to do their groundwork? Dunno, just an idea. :D But you're correct in your take about the importance of the initial precedents here, IMO. >That's another reason why we need a fund to defend robowaifus in court. Yes, clearly we do. I personally expect this will literally be directed against Tesla first (& foremost) simply b/c Elon is Literally Worse than Literally Hitler now (in the minds of deranged libsh*tes). Let us hope & pray Tesla crushes it in the courtroom! Cheers. :^)

>>38489 I was thinking of a really loud siren, something that will startle and alert others. Frankly, before arming a robowaifu with anything beyond , I would want to make sure her friend-foe distinction was iron-clad and near flawless. >>38515 I think you're way ahead of schedule. Most robots are still quite stiff and fragile.

>>38543 >Frankly, before arming a robowaifu with anything beyond , I would want to make sure her friend-foe distinction was iron-clad and near flawless. This is a really-solid point, GreerTech. But simple IFF is not enough. Just as its far too late to defend yourself from a violent evildoer's actions once they've already killed you (just as the libsh*tes would have you suffer), so its insufficient for our robowaifus to merely figure out that "Scholarship Recipients" may be there in the room with her. >tl;dr She also needs to be able reason towards intents [1], then plan/act accordingly. Kind of like what would be entailed within solving Anon's suggestion concerning the Martial Arts training : ( >>38530 ). Good post. Cheers, Anon. :^) --- 1. https://en.wikipedia.org/wiki/Intention_(criminal_law)

>>38555 Very true. I think so far, we should focus on indoor security

>>38556 >I think so far, we should focus on indoor security That certainly seems the most sensible approach. OTOH Anons like me rush in "where Angels fear to tread"! :D >tl;dr I want to take my dear Robowaifu to picnics in the parks right away! * :^) --- * After all, why should just Takuma and dear Mina-chan get to have all the fun!? :D

I believe that our robots should rely on the simplest, safest security options. We're working of a machine to love, not to shove. 1. Alarms & Alerts this has been mentioned here: >>38489 I would add that contextual awareness should be used. If you are home, she'll alert you as she rushes to you. If you are away, she'll hide and quietly alert you. You could then trigger various responses after viewing whatever information was available. 2. No violence, ever. A machine should never attempt to harm anyone under any circumstance. She should only run and hide. At worst, she could use flashing lights and loud noises to disorient. Both limited within safe levels for the average person. >>38492 I'd even say these methods go too far. 3. No weapons, designed for safety at all times. This is simple, she cannot wield a weapon. No spikes, minimum pinch points, and no electric shocks. She's a machine, like a fridge or toaster. Rely on other means for defense. If you fear her being stolen, use a GPS and log her location regularly. Naturally, your machine is going to be yours. I won't do anything to stop you from strapping a nun-chuck to your roomba. I won't do it though, and urge you all to maximize safety for all people, including the bad ones.

>>38561 I agree with this wholeheartedly. I don't think putting weapons is feasible at this time, however I do think non-lethal energy weapons will be useful. But at that point, I don't even think "weapon" is an appropriate name. More like "deterrent".

>>38561 POTD >I'd even say these methods go too far. Ehh, you're probably right after all. >She's a machine, like a fridge or toaster. Rely on other means for defense. Yeah. Thanks for the gentle reminder, Kiwi. Cheers! :^) --- >note: Added this as safety-standard to OP ITT ( >>10000 ). @Mods Feel free to embellish/edit as you see fit.

Aside from command center for your home security system, I wouldn't want any weapons on my robowaifu herself either. Hide and go into safe mode sounds like a better option. Property patrol would be for my unitree go robodogs. My silicone dolls are actually more expensive but I'm really not worried about them getting stolen. It would be pretty funny and be covered under home owners insurance. I could get them vaginal dentatas though. I don't think you could be sued for that.

>>38564 OK, I did some research on how a martial arts robowaifu might be implemented and here's what I came up with. The consensus seems to be no weapons, but I believe hand-to-hand can and should be used - sort of. Originally I wanted to teach the robowaifu to grapple because robowaifus will likely be smallish to reduce material expenses and power consumption, and they won't be good strikers at that small size. But my thinking has somewhat shifted on this. Grappling can give a smaller fighter the win over a larger fighter, but without hydraulics that are too heavy to use on the average robowaifu, she won't be able to apply the pressure needed to competently perform choke holds, joint locks and so forth. This severely limits her options as a fighter, meaning the scenario where a robowaifu injures or kills somebody is extremely unlikely to occur if she's unarmed. But there may be another alternative - aikido. Aikido is a style of martial arts that places a strong emphasis on evasion. It includes some offensive techniques, but the style is often practiced by pacifists and some aikido schools ban their students from competitive fighting. A robowaifu who rarely or never attacks and instead optimizes her dodging ability will be less likely to get in trouble for attacking somebody (but not completely - roasties may just edit a video to make it look like the robowaifu was aggressive). There are a few other styles like this that could be used as well, but many require complex movements a robot might not be able to perform. Economy of motion is key, and the need for this is filled by certain variants of kung fu and some of its offshoot styles, because you can't very well make a robot that learns martial arts Matrix-style and not include kung fu. So anyway, here's my idea about how to actually do this. What you'd have to do is rig up an AI video generator to serve in this capacity, training it on hundreds of thousands of martial arts videos. When the robowaifu detects aggressive movements, her AI internally creates a short video of how she responds to this, using another database of images depicting various martial arts stances as start images. Another AI model then translates this video output into robot movements. Unfortunately, there would be considerable processing time inherent in this setup. It can take between several seconds and many minutes to produce a still AI image, and even more for a video - far too much response latency to act appropriately when under attack. So the waifu would need more processing power than she can get locally. This could only be accomplished by using a remote server, or more likely an entire server farm. If optical processors and holographic storage were widely available, you might be able to run this on the waifu locally, but those won't be common for at least 10 more years. This is what we can do with current technology. Obviously most people don't have the money to burn on building a server farm just to teach a robowaifu martial arts. I'd do it if I had the money, but I don't. As a test of this theory, I decided to try to create some AI art of a robowaifu doing martial arts. It didn't go according to plan; the AI has a hard time creating martial arts stances, so this is an area that clearly needs some work. But I think this came out nicely aside from the fact that it doesn't actually depict martial arts.

>>38832 This is assuming the robot can move like that. If you look at modern humanoid robots, most of them move in a very stiff manner. The only exceptions I see are Unitree's robots and Boston Dynamics. But the former still is stiff to a degree, and the latter runs on gasoline except for the new one and the yellow dogs, but those aren't exactly Jackie Chan. https://www.techeblog.com/boston-dynamics-new-electric-atlas-humanoid-robot/

>>38833 I don't really know what to do about the physical side of this, unfortunately. Most likely an artificial musculature will be needed, although I suppose it's possible to use a tensile system or spring-loaded setup instead.

I've been trying some other AI art models and gotten some better results for martial arts, but sadly the only models I've found that work well for this are more demanding of system resources, meaning they'll run slower and be unable to respond quickly in a fight.

Remember the Library of Alexandria! Back up your files on multiple places.

>>38887 >Remember the Library of Alexandria! This. <---> Just coincidentally: :^) // RW Curator // ============== // Search, Information, and Data Storage, Retrieval, & Optimizations Library // -Part of the RW Foundations suite of libraries // Filename: ./include/curator/curator.hpp // -Declarations of the baseline RW Curator library #pragma once #include "incls_all.hpp" namespace rw::curator { //------------------------------------------------------------------------------ /** */ class Curator_b : public RW { public: Curator_b(); // default ctor explicit Curator_b(std::string const& handle); // tag-parm'd ctor }; // RW Alexandria // ============== // The Library, well, Library : Lots of dusty tomes edition // -Part of the RW Foundations suite of libraries // Filename: ./include/curator/alexandria/alexandria.hpp // -Declarations of the baseline RW Alexandria sub-library #pragma once #include "incls_all.hpp" namespace rw::curator::alexandria { //------------------------------------------------------------------------------ /** */ class Alexandria_b : public RW { public: Alexandria_b(); // default ctor explicit Alexandria_b(std::string const& handle); // tag-parm'd ctor };

>>38832 >Aikido places emphasis on evasion. This is a great idea. Having her dynamically evade assailants is good. >Economy of motion A great point. This is vitally important, she must efficiently flow in her movements. >>38833 brings up a good point, most robots look like lumbering stiffs. This is related to how instead of moving with nature, they're moving by overpowering it. Martial arts prominently exemplifies how moving with nature, utilizing physical dynamics and rhythm, leads to incredible efficiency. Studying this can help in developing robots which both move with grace, and flow around attackers, ensuring the safety of both parties. Shion Ashimori from Sing a Bit of Harmony being a great example of this. https://www.youtube.com/watch?v=lH_TM9fcI9o

>>38941 >Studying this can help in developing robots which both move with grace, and flow around attackers, ensuring the safety of both parties. Beautifully-put, Anon. I'm fairly sure that such tensegrity-like motions would also be energy-efficient as well. This would have far-reaching benefits for our robowaifus beyond simple self-defense scenarios.

>>38942 Correct me if I'm wrong, but if this idea of using an AI video generator setup to dynamically create movements to be fed into a tensegrity/pulley-based movement system works out, wouldn't that mean all the core problems for robowaifus have now been solved and mass production can begin? As far as I understand it, the big 4 problems are speech, robot movement, power supply and safety features. Mass adoption of LLMs and AI-created voices has solved the speech problem already, and the concept of a battery swap station holding multiple batteries the waifu can switch between has been growing in popularity; some may also choose to have a mobile battery robot whose sole purpose is to haul huge batteries to provide a backup power source for the waifu, which would basically amount to an R2-D2-like counterpart for the waifubot. Safety features have been discussed at some length in this thread, meaning the only thing that hasn't been decided yet is movement, and we seem to now have a working concept for that. It'll still have to be tested, but it's probably the most readily actionable idea I've seen because it relies on currently available technology instead of presupposing some futuristic substance or device that can be used as a robot actuator. I'm sure there's something I missed here because I don't have any sort of technical qualification; my knowledge of the scientific side of this is mostly just whatever I've managed to glean from reading this board. But there's something else too, which will likely prove to be a huge issue even if this idea works. A major problem I can see for the implementation of my AI video generator movement control idea isn't technical, but again political and legal. Because the robowaifu will have to be connected to a remote server farm to operate this way with current technology, this puts the software needed to make a robowaifu move like a robowaifu should out of reach for the average person. Most likely a corporation will have to step in to build and maintain the server infrastructure needed for dynamic robowaifu movement, and it won't be an American corporation that does this. It will probably be a Chinese company; I've recently seen news of a Chinese-produced robowaifu going on sale in 2027, so they're clearly interested in doing this. This means feminists will be able to frame the robowaifu issue as a national security issue, because oh no muh Chinese spies are getting your data and sheeit! (...which is supposed to be worse than Americuck feminist corporations and government employees getting your data...how?) And Republicuck retards are big on muh national security, so once again cuckservashits will be in perfect agreement with feminists that letting men have freedom is a threat to America. What do?

>>39001 POTD I'm going to work up a response for you over in our R&D bread, Anon : ( >>24152 ), b/c it touches on so many different topics. Expect that sometime in the next few hours. Cheers. :^) <---> Done. ( >>39015 )

>>39002 I did mention motion capture data earlier; you can use basic animations like you'd see for a video game character if you want something that's able to be run locally. You'd have to make sure there's nothing in the environment that could interfere with these animations, but she could do basic movements just fine with that. However, she wouldn't be able to dynamically generate new movements, let alone martial arts movements. This idea might be more of a feature for high-end robowaifus because of the expense and complexity it entails. However, that wouldn't stop the media from talking about how your data is being given to the Chinese or some such. Even a robowaifu that can't do martial arts can still spy on you.

I'd like to slightly shift the discussion's focus for a bit and talk about the strategies that anti-robowaifu propaganda will use. Apologies if this is kind of a ramble. The anti-robowaifu political strategy may be more general than just robowaifus. Even if the average robowaifu has the combat capabilities of the average protocol droid (which makes a certain amount of sense because a robowaifu is essentially a protocol droid with a specific focus on sexual protocol), there will doubtlessly be people looking to build combat models. Pointing out how difficult it currently is to build a humanoid fighting robot will fall on deaf ears, because robot haters will be looking to ban humanoid robots in general not on the basis of what they can currently do but on the basis of what they might be able to do in the future. Let's suppose that my AI video generator setup actually works. If it does, I imagine that by 2050 or so, it will be possible to run high-quality combat algorithms that extend well beyond aikido locally on a humanoid robot, meaning this feature will no longer be high-end but the standard. Research into using these sorts of AI video setups not only for combat but for dynamic robot movement in general will be funded by people who want a more realistic sex experience from their robowaifu. No longer will robots be clunky contraptions that are stopped cold by a flight of stairs. They will be capable of dexterity equaling and surpassing the best human athletes in dexterity-focused sports. I play a lot of RPGs, so when I think about things like this I tend to think about them in terms of RPG stats and character builds. Humanoid robots suitable for mass production will have low Strength and Constitution but extremely high Dexterity. But most importantly they will have high Charisma, this being exactly the goal of this little enterprise. These traits will be seen as a threat by more than just feminists. I suppose the main point here is, the Robowaifu Legal Defense Fund may end up spending most of its money not to defend robowaifuists in court but simply to keep robowaifus from being outlawed. There are just too many groups that would want this ban to happen. We've heard about the impending robot takeover for decades, but what almost every portrayal of robots fails to understand is that these issues are basically just an extension of current politics. This brings us back around to robowaifus being accused of spying for enemy nations. But diplomats are in many ways just officially sanctioned spies. Maybe we could get some country to extend diplomatic immunity to robowaifus? But more likely, robowaifuists will have a terrible choice; either be hunted in westoid femishit nations and forced to leave, or somehow get enough corporate backers who understand the potential benefits of advanced robotics to get not only protective laws but an entirely new legal framework enacted that protects robowaifus under "high-tech innovation" initiatives. This is similar to an Asian special economic zone; many argue that western countries already are economic zones more than actual nations because they obviously don't value the lives of their people, so maybe we should make it official and try to create a special zone where we can make that work for us rather than against us.

>>39024 I think the biggest strategy antiwaifu types will use is saying that they erode human relationships. Simply point out that they don't care about things like dating apps and expose them as hypocrites. That's usually enough to shut people down.

>>39025 >feminists being deterred by having their hypocrisy exposed Everything feminists have ever done has been hypocritical. Everything feminists have ever done has eroded human relationships. If you think they're going to be given a moment's pause over this, you don't know feminists at all.

>>39025 >>39026 Unfortunately, I'm going to have to agree with the 2nd poster. As reference, I'd point to @GreerTech's good post touching on this: >"...It doesn't blind and/or stop them at all. Radical leftists still are very able to see and target groups like "Men's Rights", "All Lives Matter/White Lives Matter", and "Straight Pride". Again, they're not robots or principled centrists, they don't care what we have to say." ( >>38661 ). >tl;dr <Haters gonna hate. There's no reasoning with someone who simply wants you dead. Don't go in expecting any sort of rationality from such demons.

>>39029 We don't need to reason with them only to make them look bad on camera. Debate is never for the opponent, but the audience.

>>39030 I agree wholeheartedly, but it probably won't shut anybody down.

>>39030 Make them look bad to who? The audience largely sucks too. Most people in the West are no different from feminists in their level of hypocrisy or dishonesty. At most these others pursue different goals, and will not be persuaded to dislike feminists due to feminists being shown to be hypocritical because they themselves are the same. And more to the point, Westerners generally agree with feminist ideas even if they don't identify as feminists because feminism is so ingrained in Western culture that many core feminist ideas are now considered the core ideas of Western culture. Ideas that originated with the feminist movement are no longer recognized as such because they've been normalized through decades of propaganda, or else people are simply apathetic to the problems feminism causes because they think they aren't affected (yes, there are still plenty of people who are this stupid even after years of woke). Cuckservatives are completely onboard with every feminist policy that matters, including the prohibition of robowaifus. We have very little in the way of a potential audience that might be remotely sympathetic.

>>39032 I wouldn't be completely blackpilled, there are many zoomers who would agree with our mission

>>39032 The moment one of the robowaifus can show benefits to mentally ill homeless veterans we have an "in" The number 1 rule in American politics is "never insult the military" if these can provide help for our fallen heros discarded by the system it will be hard for anyone opposing the technology not to look like shit.

>>39034 >The number 1 rule in American politics is "never insult the military" Maybe 30 years ago. Feminists will happily discard this if it suits them, and they won't suffer political consequences for doing so because younger generations don't really respect the military.

I'm thinking and testing ways to physically isolate a smartphone, just like how you put a cover on your webcam when you're not using it. So far, I have been testing out electrical tape on the mics, but it hasn't blocked sound that much. My phone case already has a slide over the front camera, and I put a small piece of electrical tape on the front camera that I can easily remove.

>>42329 Glad you're at least concerned about the potential surveillance issue there. Sound-deadening has many possibilities. Since almost all these piezo mics operate via "pinhole" ports in the cases today, you might just use something like smol dab of plastic (modelling) clay or something similar. Be sure you don't press it in too hard or you'll likely clog the hole. How are you testing the effectiveness, btw? Good luck Anon, and alway rember: Just because you're paranoid, it doesn't mean they aren't watching you!! :D Cheers, GreerTech. :^)

>>42330 I test via using a recorder app and saying a phrase. I did a compromise system, I covered the top mic but not the bottom mic. Honestly, these mics must use some sort of CIA black magic, because it can still record even when in a plastic bag. I guess that means we should use piezo mics for our robowaifus.

>>42388 They do use incredibly sophisticated DSP processing on the audio signal. Today this research is largely driven (& funded) by glownigger spying needs (as you allude to); as practically every goyim """just-so-happens to conveniently""" keep such an always-on, audio & video surveillance + tracking device glued to their hand. In the past this overall yuge R&D effort was simply so that a reasonably-good audio signal could still be had in "corner cases" with weak batteries and/or weak signals and/or noisy environments. This type of research & engineering with FFTs + DSPs, &tc, has been going on for several decade's time now -- at least back to the Apollo-era of spaceflight. But, as with everything the kikes touch, it eventually became (((corrupted))).

I made this guide for myself for Google alternatives, as part of a project I'm working on, and I'm sharing the guide with you guys in the spirit of sharing information Bonus: Find the GreerTech invention

>>42426 Without looking, I'll guess Find Hub?

>>42438 Wrong It was FreeStream https://freestream.neocities.org/

>>42444 >digits Dang! Oh well. :D Nice work, Anon. Cheers. :^)

Ken Thompson's 1984 Turing Award lecture: https://trashchan.xyz/robowaifu/manage/thread/26.html#1334 He played a trick to shock people in the ACM about security threats in software. >related: https://www.youtube.com/watch?v=SJ7lOus1FzQ

Electropedia Intredasting catalogue of ideas. * Here's the subcategory: >Area: 903: Risk assessment www.electropedia.org/iev/iev.nsf/index?openform&part=903 Obviously a very hot topic for robowaifu safety systems & development. --- * The rigorous technical categorization (with ISO/IEC refs) here is what makes this especially useful to autists and engineers but I repeat myself

>>45165 Made a custom LLM, no. I know some fellow robowaifuists that are working on stuff like that. Coincidentally, I am in the middle of updating my Simple Offline AI manual. I do use offline AI for robowaifus. My focuses are usually on hardware, my main project is the Galatea v3 robowaifu

>>45173 I said it in a joking manner, but I kinda do want to know, but he might just be trolling, which I'm starting to think is the case

(Crossposted from >>45299) >>45295 No matter what, any discussion of armor for you-know-what is hampered by the fact that armor strong enough for that would be incredibly heavy. The low range armor I mentioned would have to be 0.2-0.25 in./5-6.35 mm thick. We already are having enough on our plate designing lightweight humanoid robots, let alone ones with steel. However, I think something capable of blocking melee attacks is very much doable. >/k/ -tier tangents Well I'm not sucking off NATO, the US Military, or wanting to ethnically cleanse Russia, so I'm safe there >Practical considerations of engineering & safety is what I'm talking about. There's still plenty of room for hardening, cover, concealment, avoidance milcon topics discussion within such contexts. That's the thing about military and war. It's not some esoteric thing, it is merely advanced survival. And whatever you think about war and the military, the fact is the procedures and operations; -Have been tested through centuries of brutal and merciless trial-and-error -Is designed to be able to be followed by a 1.9 GPA teenager -Made to be able to be done with lowest-bidder equipment -Made to be done in the most extreme situations

(Crossposted from >>45300) Here's a good non-tactical and very relevant example of why emergency shutdowns are necessary for articulated robowaifus with a moderate amount of strength https://youtu.be/OYryogNE8Ys

>>45306 The fact all the "totally available for the public bro" robits don't have something simple as an emergency stop -even retail animatronics have emergency stops via easily unplugged- shows how ready-for-market they really are (not very). Also shows how little the eggheads developing these robits think about them practically. The home hobbyist deals with more pragmatic problems quickly because he is (naturally) also living with his creations.

>>45308 >Also shows how little the eggheads developing these robits think about them practically. The home hobbyist deals with more pragmatic problems quickly because he is (naturally) also living with his creations. picrel

>>45308 POTD