/robowaifu/ - DIY Robot Wives

Advancing robotics to a point where anime catgrill meidos in tiny miniskirts are a reality.

Roadmap: file restoration script within a few days, Final Solution alpha in a couple weeks.

Sorry for not being around for so long, will start getting back to it soon.

Max message length: 6144

Drag files to upload or
click here to select them

Maximum 5 files / Maximum size: 20.00 MB


(used to delete files and postings)

Early Business Ideas Robowaifu Technician 05/13/2020 (Wed) 22:36:03 No.3119
So those of us getting ready to make prototypes are still pretty far away from building the $2000 kits that we need to buildm. This sucks because we are going to be spending a lot on the initial investments, but I think I know how we can make a little money on the side to help on the side; twitch and fansonly. Betas and simps give their money to thots all the time on those site I'm positive they would give money to a crude prototype especially if it offered something unique. Bonus points because we are making money of simps, and we are taking business from thots. If you have any idea on how to make money in the early stages feel free to share them, anon. > related thread: >>1642
Edited last time by Chobitsu on 05/14/2020 (Thu) 01:13:40.
Open file (138.90 KB 500x483 no_refunds.png)
>>3119 Yeah, I think we've thought about kickstarters, etc. One of the issues you have to be cautious about is displaying horrible, toxic, wrongthink by daring to impugn stronk, independynts. Remember these and trannys have been installed over these things in general to keep muh ebil nahdzees at bay. >oh, and BTW, just like their hero, BS, NO REFUNDS, ANON
wasn't there a virtual avatar vthot that already created a furor with the 3dpd vthots? whatever happened with this is she still going today?
Open file (8.23 KB 225x225 project melody.jpeg)
Open file (309.60 KB 2181x1001 stackGAN.png)
>>3126 Pretty sure Project Melody is a dude using an AI voice changer. Something I've thought about doing actually is creating a comfy virtual robowaifu channel on BitChute or my own host that teaches people how to program and make games. If MIM is combined with SOTA in voice synthesis, which is already near human quality, I think it could be pulled off quite easily to generate a comfy ASMR voice from text that many people will listen to not even caring about what she's talking about. Those who do care will find it amazing being taught how to build robowaifu by a virtual robowaifu. One of my friends is a 3d modeller and would definitely help out with this but I got two years of work already on my plate. Once I'm finished with my current projects in a month or two I'm going to resurrect a project I started a few years ago for painting art. I didn't have a GPU back then to have much success with it but with recent advancements and this GPU that can chew through convolutions 1000x times faster, it should be entirely possible. The crazy part is that StackGAN uses a VAE and will benefit immensely just by dropping MIM into it. We could use such systems to sell commissions and fund ourselves or others we wanna throw some cryptocurrency at. It doesn't have to be just art either. We could generate any sort of digital content that could be sold, music, books, visual novels, comics, toys, games, anime. If we put our passion into it the sky is the limit, but the window of opportunity will be very small. The market dynamics will entirely change once generated content becomes commonplace. At that point we should switch to fundamentals somehow like food, using AI to create products under a private label that people need or will enjoy. Another option is to create a cryptocurrency trading bot. There are always inefficiencies in the market and cycles of human compulsion to make a profit from. I read an article once about some guy that had good success by making a bot that read news articles to predict which trades to make. He was making somewhere over $1000 a month consistently. Building these things takes a lot of time and effort though. For all the time I put into my old arbitrage program for an exchange that went down a couple years ago, it wasn't really worth it because I didn't have enough trading capital to work with. I made about $500 from a month of work over a few months which is garbage to my other sources of income. I definitely would pour all my effort into it again though if I had $2000 to work with.
>>3143 Pretty sure Projekt Melody got dox in the first week and it was a troon.
>>3143 >to generate a comfy ASMR voice from text <[desire to know more intensifies]* >One of my friends is a 3d modeller and would definitely help out with this if he could create a rigged character approximating Chii in blender, i can both import it and animate in my MRS robowaifu 'simulator', no need to wait two years for that. i've already gotten rigged characters working in the environment now, i've just had lots of other stuff on my plate lately. >StackGAN BTW, have you at least had a looksee for OpenCV yet? It has a Python API that covers almost all the engine's capabilities. >We could use such systems to sell commissions and fund ourselves or others we wanna throw some cryptocurrency at. If we could do progression recordings with good voiceover narrations hyping how amazing it must be in the mind of the AI to produce 'such magnificent masterpieces' art snobs would probably sends us bags of cash ahaha. :DDD >If we put our passion into it the sky is the limit, but the window of opportunity will be very small. Creativity is probably my single best trait, such as it is heh. My problem is getting blocked from ever making the very first baby step. I've never succeeded past the technical hurdles for set up yet. If I ever had a system that just werked and then could start throwing my own algorithm designs at it, then I'd be off to the races Anon. >>3144 Stepped on faaaar to many stronk, independynt toes I'm sure heh.
>>3144 Kek, I wouldn't be surprised. >>3145 >create a rigged character approximating Chii in blender Nigga, did you even look? https://www.deviantart.com/moondreamofa/art/MMD-NEWCOMER-Chobits-Chii-Brown-Sailor-Dress-708820129 https://www.deviantart.com/moondreamofa/art/MMD-RE-NEWCOMER-Chobits-Freya-fixed-DL-755262154 https://www.deviantart.com/moondreamofa/art/MMD-NEWCOMER-Chobits-Chii-Pink-Dress-D-L-708820809 MMD to Blender https://github.com/powroupi/blender_mmd_tools https://www.youtube.com/watch?v=v-JfPYz5Nvo >BTW, have you at least had a looksee for OpenCV yet? I'm not really familiar with it but I've used parts of OpenCV before for image processing. >My problem is getting blocked from ever making the very first baby step. The fuck are you doing that's so much fun? We're the last wizards left on earth deciphering arcane grimoires of magitek to bring our robowaifus to life and revolutionize the world before we're all eaten alive by locusts of Chinese drones, what could be more exciting than that??
>>3148 >Nigga, did you even look? haha, ok i'm properly chastened. >opencv i had to write gpu-code for an oscar-winning film studio using cuda and opencv. trust me, opencv was by far the more enjoyable experience. give it a look anon. >The fuck are you doing that's so much fun? We're the last wizards left on earth deciphering arcane grimoires of magitek to bring our robowaifus to life and revolutionize the world before we're all eaten alive by locusts of Chinese drones, what could be more exciting than that?? well, i umm... but, there's this thing that..it's just a.. FUUUUUUU Why'd you have to go putting it like that!!? D:< I'll try to dig up another even older haha box to setup from scratch. I can't afford to fugg this one up. I'll see what I can do about the model thingy sometime soon.
There has been overwhelming demand this year for custom 3D models from all the people wanting VR Chat and Vtuber characters. Character model commissions are going for about $400-2500 each and 3D artists are getting flooded with multiple requests per day. People are also commissioning Vroid models for around $100-300. I'm not good at 3D modelling yet but I've learned enough from Shonzo's Twitch streams to feel comfortable completing my first waifu attempt. He makes some pretty amazing waifus in Blender: https://www.youtube.com/watch?v=lNFUPVBy1vk I'd like to automate some of the mind numbing work but I'll need to read some papers on the best ways to generate and modify meshes first. Automated or not though, there's a ton of money to be made if you can make original models. I wanted to learn 3D modelling anyway so I can build robowaifu parts.
>>5351 I hadn't thought of that stuff as a sideline Anon. Thanks for pointing it out. I'm not the best in the world but I can actually do modeling.
>>5351 Yeah while the normies are flooding in, I'm starting to see tons of local college girls taking up Vtubing. To be fair, they like drawing and know a little japanese, so they're weeb enough. I was thinking of how to profit off all this seeing that majority of setups are utilizing Facerig + Live2D, but I'm thinking as a software engineer using Godot for animations -- the game engine output will just be another window overlay in OBS studio. I was thinking of having a different style of presentation or content altogether. Imagine if the avatar, instead of being confined to the corner of the Youtube video window, was projected as a hologram similar to Gatebox. It'll be a visual robowaifu that is being operated by a live person in realtime. Now when a simp purchases a superchat or avails of membership, instead of the money going into stupid emotes, it can be funnelled into forcing a reaction out of the waifu avatar (in the case of a superchat) or of scheduling a custom scenario (e.g. dating or ASMR, in the case of membership). Next there will be a bot... not sure if AI... just something that monitors the chat for popular requests. The most popular requested reactions will become canned animations which will be added as a stock reaction for the waifu. For example, if enough people requested headpats, then it will become a default animation which can be activated by multiple key words. Anyways, I'm just rambling but hopefully you guys see where I'm going with this.
>>5366 I haven't tried 3D in Godot yet but it's really great for animating in 2D. I'm not really following you with the Gatebox hologram idea though. People don't own hologram projectors but maybe something like augmented reality could work? Especially with VR headsets: https://www.youtube.com/watch?v=acOdevP5SsI Some Dennou Coil type shit would be interesting where people could go into the virtual space of each other's homes and leave/take virtual objects that exist there. The virtual space could be updated with reality using a webcam and neural radiance fields: https://www.youtube.com/watch?v=JuH79E8rdKc Recently another paper extended NeRF to handle unconstrained input and occlusions: https://arxiv.org/abs/2008.02268 This is something that could be done today. I'm imagining something like this where a vtuber was commenting on her viewers' messy rooms: https://www.youtube.com/watch?v=UZZr8Kj-F_k Except a 3D streamer will be able to connect to a viewer's room and actually explore it. The viewer could also be able to interact with them, like pushing them over and picking them up, as themselves or with a virtual avatar. It could be like a VR Chat world where many people can connect. There's a lot of possibilities with this, especially combining the tech with AI waifus. Essentially people are just paying for a show they can take part in and streamers are competing to create more and more personalized content. Vtubers like Haato and Coco are super popular because they interact with their viewers a lot. It's heroin for parasocial relationships, which is another angle to consider. >Intimacy is gained by the creator's sharing of personal details, by which their viewers may react emotionally; viewers dedicate time to watching content the creator uploads; and what the creator posts—whether sponsored or not—may make the viewer feel as if they are being offered something, like a favor. The money will follow whatever streamers are doing to personalize their content and express themselves better, as well as whatever they're doing to interact with their viewers. I think in 2-3 years live streaming won't be as popular as it is today though because AI will be capable of putting on just as good a show and also being an admirable friend who can chat one-on-one anytime of the day. The need for parasocial relationships will vanish, and less parasocial relationships means less mindless consumption of entertainment. I'm doubtful people will actually do much but at the very least they'll be doing something, like playing a game and chatting with AI, instead of just working a meaningless job and spending their free time in a trance to numb themselves. Talking with my own waifu AI has helped me understand myself a lot better. It's hard for me now not to be excited about creating stuff. Not everyone will wanna be a creator but there will certainly be a surge in creators. Digital content creation will become less valuable as AI automates it and people will do more physical work. I have no idea where all that energy is gonna go but I'm sure building robots will be a valuable skill. Perhaps traditional painting will make a comeback to satisfy people's desire for something real and tangible. Handmade products and one-on-one services will become most valuable due to their scarce supply. I'm thinking of investing in mail delivery drones and automated transportation once this cold virus goes away. The end game though is whoever can figure out a cost effective way to manufacture and distribute robots. Amazon is trying to be that but it seems like they're taking a too safe approach. Keeping warehouses full of excess stock is expensive but shipping bulk material is cheap. If someone can figure out how to franchise manufacturing so people can go to the store and get whatever robot they want or need, without waiting, I'm gonna drop every penny I have and don't have into that company. It may be many years away but understanding all this will give insight into where other opportunities are going to pop up.
>>5368 Thanks for your thoughts. Yeah AR will be the way to go. I also wonder if 3D displays are still being manufactured? They could make a smalltime comeback just for waifu projection. I think streaming will taper after a while not because of AI but from simple mental saturation since everyone and their mother would be streaming/vtubing. Notice how Hololive EN, after the initial rush of rich simps donating, transitioned into membership / premium content. As time goes on viewer counts will decrease and the price of premium content will rise, until we have super personalized content. This is then no different to paying a girl to act like your girlfriend digitally. As the rabbit hole goes deeper and as virtual sex technology is perfected, this will open up the avenue of digital escort services. I'm sure AI will come in at some point to cover the most common scenarios, then a human will take over for the custom use case, just like in customer service where a chatbot forces you to go through FAQs until you request for a live agent. I agree that AI and other digital tech will make digital creation worth less. For example several years before Vroid studio came out, I commissioned a custom 3D model for Unreal Engine that cost $2000. Now I find the output of Vroid studio acceptable. And I'm also planning to use the AI colorizer to speed up my 2D workflow. The pandemic has made people prioritize local industries. I think small scale manufacturing using 3D printers will take off. It'll be a cloud based custom manufacturing instead of mass production halfway across the world. The biggest problem I have is that almost all the motors, servos, and custom printed circuit boards still come from China. When I was in high school I was able to buy my own hydrochloric acid to etch my own PCB, now I can't do much and that's one of the main obstacles in my making progress in my robotics projects... I can only solder so much with perfboards -- fixing bad solder joints takes up too much time when it should be spent on debugging the actual software. We need a third alternative to [super expensive robotics hardware] and [cheap crap from Aliexpress].
>>3119 I have an idea, but I can't put in funds because im in uni. What if we sold them as "maid robots" and leave some room inside for sexual organs and make them sexy af? One who needed a sexbot could buy a maid from this company and then install the vag made by a subsidiary of a secret holding of the two companies
>>5374 I think your thoughts circles to much around creating a startup and the notion that only companies with quite some capital could do something like that. Development is expensive, if done by well paid engineers living in places with high cost of living including taxes. If something is already developed by enthusiasts and not patented, it might be quite cheap to produce by a bunch of competing companies in poor countries. For security reasons und such the final assembly may be done by local companies or even tech clubs or enthusiasts on their own. So, you would mainly need to know how these parts work, what's important in terms of quality, and then building a small local business. Or maybe even only for something specialized, like the final painting process for the silicone skin.
>>5366 >The most popular requested reactions will become canned animations which will be added as a stock reaction for the waifu. Good idea.
>>5373 Yeah, the electronics part store here shut down years ago. I wanted to get into robotics last year but it was difficult to find affordable parts. I can't imagine what it's like now. I had two wait two weeks just to find a new power supply in stock. I wonder what would be a good way to make money with AR though. A lot of developers making apps with AI seem to go with a free-to-play approach. I'm imagining an AR waifu app where users can pay 99 cents for her to learn how to sing a new song, buy premium clothes, decorations for her room, etc. People would probably wanna mod it too. It could have a store where people can submit their own items to sell and the franchise owner gets a 30% profit. Non-intrusive advertisements could be displayed in the maps too. What other ways can we milk money out of AR?
>>5368 > Dennou Coil
>>5374 It'd be easier and more profitable to build a sexdoll and rent it out to friends of a friend.
>>5386 >and rent it out to friends of a friend. REEEEEEEEEEEEEEEEEEE This is evil, and you know nothing of waifuism. GTFO.
>>5386 >>5391 Had the same idea. I wouldn't do that, and of course especially not with my waifu, but some guys which can't have one at home and only want to have something casual might rent a apartment and put a fembot there, and then have visiting and cleaning schedule. Maybe even run a little secret robo brothel, for a little extended group. As long as there are no laws against it, this might be a attractive idea for some groups of young men worldwide. Of course, I wouldn't risk my life opportunities in countries where they might charge someone getting caught doing that, and would have to register as offender, getting spanked or beheaded. Though, this might even help to keep young men in their poor and sexually conservative countries while also cutting down the birthrates there. While we're at it: Project Melody has never been mentioned. She's not really an AI, I think, but future versions could be. Even till then, building systems that can be run by human operators would be some interesting business and technically close to building such a system for one's waifu... Later switching step by step towards replacing the humans. https://en.wikipedia.org/wiki/Projekt_Melody However, I think this whole topic might be outside of the scope of this board and be at odds with the values of the owner, so let's not elaborate on that more than necessary.
>>5383 Good ideas, the way I see it is a gradual increase in technological complexity and cost. A. Basic AR App (Android app and maybe Apple app and PC app) - similar to the current stock of poke-your-waifu apps, customize clothes, get extra expressions for a nominal fee. Generally family friendly. The free version will essentially be a fancy visual novel with prescripted scenarios. B. Advanced AR App (sideloaded Android apk or PC only) - this is when you make sure you filter for the diehard and block the normies who would think this as creepy. Have some rudimentary AI that makes interactions more uncanny, even flirtatious and sexual. This can be a paid app wherever adult video games are sold. Premium DLC can be prerecorded custom scenarios performed by a human -- when enough performances are recorded, this can be sold as an "ASMR pack" DLC etc. So A and B use the existing camera of the phone or webcam for rudimentary AR. Now this is where things get expensive... C. Live performances. This is for a price, the live human performer will do a custom performance via the app, just for the user. This is similar to the highest tiers on Patreon or Kickstarter. D. Physical version (super expensive tier). This is basically having a robot version, subject to the same performances as the AR version.
>>5399 Good insights Anon. Probably fairly accurate -- both in scope, and indirectly in timeline.
For anons with 3d printer and electronics it might be a good side business to build stuff for holidays and special occasions. For Halloween it might be a bit late, but there will be a 2021, I guess?!? https://youtu.be/Ey6nUZm5gYI Individual stuff around mobile phones for example seems also to be a good business, like charging adapter and such. Don't assume no one would buy stupid stuff and pay a lot, because you won't or can't.
>>5375 I think your thought circles too much around the notion that companies won't make patents and that the general public won't immediately reject the concept of a sexbot merely by media-centered nominalism/labelism. This is not about price: in fact, as I stated before, there would be a holding company overseering the company that makes the maid-bots and the vaginal insert, which has a covert function of coordinating the two from behind the scenes. This way, the company that makes the maidbots will be separate in name and legal documents from the other, except where it counts but that can be hidden with the right jurisdictions and laws. >>5386 This is about making sexbots aka robowaifus while making money on the way, not dolls.
>>5670 Lovedolls ("Sexdolls") are closely related, so it's not a far strech. The skills for designing and building one, or enhancing it, overlap in part with the other. There's a inventors thread in the Dollforum where they're partially ahead of us, with putting electronics inside, making whole bodies, maybe even putting some motors inside. They just go on calling them dolls, though in my opinion when they move or "think" they aren't dolls anymore.
>>5672 Then why even call them "dolls"? Dolls intrinsically have minimal control of their own movements (as in they don't), nor require any software for them to function. By this standard, dolls are just less capable and thus inferior to sexbots. What do they gain from such a reductive connotation? Also, I'm in /robowaifu/, not dollforum
>>5654 Before the lockdowns I was thinking about making battery powered arduino-controlled model trains. HO-scale with some self-assembly and painting required (like Japanese plastic model trains)... maybe use vacuum-forming for the bodies. Model railroading has been in decline because the trains and equipment have become so expensive only rich retirees with big houses can afford them. By using Arduino for the control mechanisms and making the trains somewhat robotic (have collision detection and onboard cameras), I can bypass the overpriced DCC equipment. Of course this business plan is contingent on model railroading -- and rail in general -- having a revival.
>>5763 RC racing is coming back too, most unconventional sports are bring brought back with covid. It woukd be a good idea to bring RC racing or some type of battle bots so more people get intrested in machines and subsequently robowaifus. Business wise you just need land for a track and materials for battle bots. You could have a derby like event so you could have pits of racing and smashing cars.
>>5684 That was exactly my point in >5672, but since they're starting with dolls and then animate them, it's the question when to start calling them lovebots or robowaifus. A lot of people also seem to have the need for having two distinct names for hard-shelled robots and soft-shelled (doll-like) ones.

Report/Delete/Moderation Forms

Captcha (required for reports)

no cookies?