>>246
Just give me a cute robowaifu with warm flesh and I'll be more than happy.

>>246
All I need is advanced is enough AI to feel like I am loved.

>>1126
Yeah, I kinda feel that a combination of a loyal doggo (+ other pets), a few close friends and a robotic waifu will be the golden combination (at least until robotics become a lot more advanced). It'd be harder for those with a NEET lifestyle, but I reckon that combination would make most men here pretty happy.

>>246
I don't even need a body. Just give me a chatbot, maybe with a 2D render running around on my screen, that makes me feel loved.
Add voice recognition on top so I could just wire up my house with speakers and microphones so I could converse naturally with her, plus the ability to write text messages/call with a phone, and I'd be ecstatic as I could even chat with her at work.

>>1128
Yeah that'd be pretty sweet. The problem is that it'll probably be 5-10 years until we have something pretty flashy. I think I'd be perfectly content with even just this.

>>1125
>>1126
I feel like I'd need both of these. If I can't hug my waifu there's no point, and if I can't communicate with her there's no point. I just need her to be cute (not necessarily realistic, just cute) and feel good (not necessarily realistic, just good) and to be able to communicate on at least some level.

Cute and warm with advanced enough AI and clever software and hardware design to make me feel like its love is real. Cute and warm is easy. Cuteness is practically free, that's just art design. Warmth will come from the internal parts. If anything, they'll create too much warmth and that will be an issue. But with clever heat distribution perhaps you can spread the heat throughout the body to make it feel like realistic (or pleasurably unrealistic) body heat.

The AI then just needs to be able to keep up a basic conversation. It doesn't need to be perfect. The trick is to make the personality acknowledge it's a robot but one that has just a tiny spark of humanity and strives to be more human. It's a classic sci-fi trope for a reason, it's super endearing. When her AI makes a mistake, now it is justified, and she can become frustrated over it or something, which is then cute.

I also figure bipedal walking is a problem. So I advocate having her AI acknowledge this as a disability, and then she can use a wheelchair or something. The trick is essentially using "story" to get around limitations. It only needs to be good enough to make me feel like it's real.

It's not about realism, per se, but about verisimilitude.

>>1128
Holy. Fucking. Shit.

That's brilliant! I absolutely love the idea of justifying technological limitations with a compelling narrative. I never thought to try to combine TV and movie tropes with a robot-waifu to make it more appealing and human. I'll definitely be looking into this as I believe this is the best possible route we can take.

>>1131
I've been advocating this idea on this board since the board was new. Frankly I thought I was annoying people with it. I think with a focus on verisimilitude, accomplished largely though art design, we can have much more appealing waifubots even with current technology. Art design isn't only visual, although that's a big part of it. It applies to all the other senses as well. And software is a part of this. The uncanny valley isn't just visual. Something that pretends to be human but fails to be is scary, not cute. Something that acknowledges it isn't human but wants to be is endearing. Have its personality acknowledge its flaws as flaws, the same way a person can. There's a reason everyone loves Data from Star Trek, or why people cry at the end of Terminator 2, or why the only death that ever mattered in Dragon Ball Z was Android 16.

Can't have realistic skin? It can be even a relatively hard plastic over most if it if that's more feasible. As long as it's still one that doesn't feel gross. Only certain parts (you know which ones) need to feel realistic. Can't get a realistic face or realistic facial movements? Make it look like a cartoon on purpose. Hell, you can even make it look like your preexisting waifu, especially if you pick a waifu that is a robot.

Right now we could build a cute and fuggable XJ-9. Not literally one that pretends to be that character, because the AI won't be as good. But a robot with that type of design, with the personality of an autistic cripple? Pretty much doable. Hell, I'm sure everyone here is autistic and lazy anyway. We'd all get along great with autistic crippled cartoon waifus.

>>1128
There's a hentai game that has SMS built into it that will send you messages you can reply to. i forget the name of it, but if you do some digging it's there.

>>1132
Very interesting. A robowaifu who recognizes herself as a robot would probably feel, ironically, more real. All her weird behavior would have a clear explanation then. Giving her a backstory is also a great idea. In general terms, I think there's no need to make all robowaifus realistic. I would prefer one that looks like an anime girl, for example.

One my main gripes with sex dolls is the fact that they frequently look like hookers. Their main references are probably porn actresses. The robowaifus developed by big companies will probably look like that, but the ones developed by independent people will, no doubt about it, look much more cute and elegant.

OP here.
I'd want a robowaifu to be equal parts lover and bodyguard. Human-ish looking but with obvious signs of being an android. Around 5'3-5'5 with a reasonably slim build and with glowing eyes and the ability to change hair color and style at will. I'd like for "her" to have her own personality but be aware that she is a synthetic humanoid, not a human. I'd still treat her for the most part as a flesh-and-blood human, but she'd know what she is.
I'd have her programmed with training in Aikido, Glima, and Krav Maga, firearms training, spearfighting and polearms training, and designed as such that repairs would be a simple matter of removing the damaged part and replacing or repairing it, without requiring a full shutdown.
Synthetic flesh would be soft and supple, with a bit of elasticity, much like many modern sex toys.
I would want her to have her own personality, but also a fierce sense of loyalty and protectiveness of me. Not obsessive, but certainly focused.
I think it would also be reasonable for her to have a beautiful looking face, but I would want the rest of her head to appear clearly synthetic…maybe plastic or simply synth flesh with geometric steel lines.

>>1134
This. Verisimilitude is much more important than realism. We want waifubots, not slutbots.

>>1125
>warm flesh
Am I the only one who believes that cold flesh could feel even better?

>>1137
yes.
(in all seriousness, different strokes for different folks)

>>1128
The future is now, my friend.
gatebox.ai/home/

>>1139
It's just a projection on clear plastic, so the hologram technology is pretty bad in all honesty. Anyone with a smartphone or cheap projector can replicate it. If it was cheaper, had more functionality (maybe a better projection system) and was in English then I'd probably buy one in a heartbeat.

It's cool but overpriced IMO. But then again, different strokes for different folks.

>>1140
True, but the product is almost exactly what >>1128 wanted.

>>1140
Things like this might help pave the way for the more advanced robowaifus. At least, [it] might contribute to popularize them.

>>1140
>>1142
Gatebox is 150K yen now, half price. What I need to know is how to project just the character data – have the backdrop just be a bright solid green, white or black screen whose light data will be filtered and how to do that.

Or maybe do it augmented reality way, have a back camera that captures the scenery at the back and have it play as a movie on the projection so it seems like a pass through window.

Either way, the current advantage with the projected waifu is the smoother animations.

On the other hand I've been watching Anki Vector fail videos and I'm pretty confident I can make a clone version, either a quadruped or a tracked tank chassis with a waifu torso. That's why I bought loads of IR sensor pairs.

>>1143
>Gatebox
Based on embed related I'd say they're using a transparent (and possibly monochrome) LED display, so they probably use a black backdrop.

Another option for a (relatively) inexpensive "projected" waifu is using the Pepper's ghost illusion: hackaday.com/2017/05/13/say-hello-to-this-cortana-hologram/

https://www.invidio.us/watch?v=nK4y5jBJfu8

Markerless AR waifu prototypes are already several years old, hacked together using first-gen Oculus Rift goggles and Xbox Kinect sensors. I'm surprised we haven't seen anything more advanced since then.

>>1145
The SDK for using the cameras to do AR on the Vive pro has only been released earlier this summer. SRWorks might help things but the main issue VR has in the last few years is all the fragmentation with so many different platforms and standards. There's the Vive with Steam and Vive Focus as a standalone, 2 different Oculus platforms, Sony's PSVR, Google's cardboard/Daydream and Magicleap platforms and Microsoft's Mixed Reality and Hololens. Plus a dozen other small players in VR/AR.

Imagine you want to make an AR demo like the one in your post, all these options are incompatible with each other. They're all walled gardens that are very limiting to develop in and tightly controlled. The promise of an open VR ecosystem back when the VR craze was launched which helped foster new innovative ideas and uses has turned out to be a gigantic lie and people aren't as interested in participating in it.

>>1144
>I'd say they're using a transparent (and possibly monochrome) LED display, so they probably use a black backdrop.
correct.

>>1145
I'd be eager to see what else this man has done since. Any updates Anon?

>>1146
This type of technology industry fragmentation is commonplace. Study the rise of the PC in the 80's (and many other paradigms). Eventually the have's and the have not's will sort themselves out. Patience.

>>1149
It isn't the fragmentation that's the problem it's that all of the options are locked down. In the 80's and 90's developers created and sold their software without having to give a cut and allow what kind of content could be be published to the people who owned the marketplace.

Try selling/distributing a program or game on any of those VR platforms and it's practically impossible outside their walled garden. Some companies and projects have found workarounds but they don't last long as there's a constant stream of updates which makes their solutions incompatible and difficult to implement on purpose.

Waiting for a competitor to emerge victorious isn't going to improve this situation. Computing freedom as been in a rapid decline and is only getting worse.

>>1150
>It isn't the fragmentation that's the problem it's that all of the options are locked down. In the 80's and 90's developers created and sold their software without having to give a cut and allow what kind of content could be be published to the people who owned the marketplace.
I can't even understand what you're saying here but it's obvious you're depressed about the whole situation, so I'll automatically take you're predictions with a large grain of salt.

New technology has had a long history of fragmenting into little islands of incompatible systems, computing systems especially so. In the end the forces of economics tends to drive companies towards compatible systems because that's in general what the users want, and eventually what they demand.

Again, patience.

Id just make taylor swift

>>2180 I think she'd be an attractive choice, physically at least. Ever hear of Ricky Ma? >>153

>>1151 Well now that it's near the end of 2020 and you'll need a Facebook account to use a VR headset from Oculus soon maybe that point is being made clearer. The situation with Valve isn't any better. If you want to use any of their headsets without installing Steam and setting up an account with them you need to install 'Vive Enterprise Device Manager' and pay to activate it, assuming they allow your use case. It's just as locked down. This has nothing to do with being depressed about the current situation but a realistic look at what modern VR is. And it's my belief that the lack of freedom is directly responsible for the dearth of creative experimentation in this field.

>>7642 No arguments about the intent of these megacorps. The point is attempting to blackpill every.single.individual. on /robowaifu/ over the problem certainly won't help. Quite the opposite. OTOH, there is at least one kit making progress in the right direction Anon. https://github.com/relativty/Relativty >(via >>5440 , et al)

>>7643 This isn't blackpilling. Someone asked why we haven't seen much advancement in innovative tech demos as we did back when VR was hacked together using off the shelf parts. As someone who wanted to experiment in this field, had the money to spend, time to waste and skills to get involved then looked up what it would entail I explain why I was unbelievably disgusted at the current state of VR. I'm not the only one that feels this way, it's not just that all the low hanging fruit was picked away early on. I receive a quick dismissive reply stating that things will get better. A year later I point out that things have gotten worse. At least now that's undeniable so I'm glad to get that now. The sooner we face reality instead of pretending that things will just work out or magically get better maybe things will start turning around. I'm still optimistic about what VR could be.

>>7645 >This isn't blackpilling. I consider it to be so. If you can't at least make a token effort to provide a solution or suggestion for a better outcome, if all that constitutes the bulk of the post is simply complaining or grousing about how bad things are out there, then it's nothing but blackpilling. So, please detail your suggested approaches/solutions to this issue Anon. I've already gotten one for you in the right direction. Do you have others?

>>246 My ideal robowaifu would be as close to human in function as possible, but with abilities and functions somewhat different to the point of improvement. For the best idea to emerge, clear goals have to be defined and listed. The idea can be open ended to be improved upon. Study of mechatronics may improve one's ability to work on such a thing, but idea and "soul" as in the preferred characterists are also important. I'll do my best to make a proper list here of the goals for my ultimate robowaifu project; -Tensegrity endoskeleton tendon structure -Modular yet sturdy, of durable materials -Responsive strong fast lowpower-use actuators -A learning computer program for intelligence -Something akin to a central nervous system -Slow silky hair growth on top of head -Capacitive skin for total sensory of touch -Ingestion system, can eat things and excrete -Facial expressions -Soft and smooth skin -Body heat -Simple olfactory sensory -Speech and speech recognition (singing optional) -Eyes, depth perception and pattern recognition -Ability to stand, walk, run and jump -Ability to handle things and learn hand skills -Breath and heartbeat ... Regardless of how difficult it may seem, once you have set goals and can envision what the goals entail, you're actually a step closer to the realization of it. Things like this already exist; https://hackaday.com/2020/11/06/spider-silk-spider-silk-made-using-a-strain-of-yeast/ Imagine your waifu growing hair that you can cut and style

>>7642 What you have described is another reason why I just decided to build my robowaifu from scratch. Basically I figure that if the corporations are intent on commercialising and fucking everything up then I won't buy their products; I'll just make my own! I'd had enough of advertisements, pay-to-win and game marketplaces that are heavily censored and controlled by a bunch of pansy-ass PR executives. Consequently, I don't game nearly as much as I used to. The last videogame I played was Doom Eternal, which was a blast. It was short but sweet. I really hope Cyberpunk 2077 doesn't turn into some DLC fuckfest where 75% of the game is unplayable or only partly playable unless you've purchased half-a-dozen expansion packs!

>>7653 That's a great list Anon. I'll pass on the eat/excrete haha but will breathing out warm air be good enough? You're definitely correct that goals begin with vision, and achievements begin with goals. So, it all starts with our imaginations! Then comes the 99% perspiration haha

>>7666 It can be a long term thing, like a great life work or magnum opus Some of the goals can be realized before others and then eventually (it being modular) it might get finalized

>>7667 >like a great life work or magnum opus You. I like you Anon. I wouldn't personally be doing this at all if I didn't consider it a great achievement when we all succeed together at this. I would that more young men in our increasingly-corrupted Western Civilization could find something worthy of a life's-pursuit. Godspeed to /robowaifu/ during this approaching Christmas season. >=== -minor prose edit

>>246 I wouldn't care as long as she is cute and her AI doesn't take over the world

>>1137 I have a doll that I never warm up. It's TPE and as long as it's powdered well it feels so good to get into bed with a cold soft doll on a hot summer night. The only downside is the rubbery medical smell never went away. I've showered it dozens of times now and it still smells like medical grade TPE. I would rather have a cold hard plastic like PVC and have even considered getting her a PVC catsuit to wear to bed. On cold winter nights it's not as satisfying but I still appreciate the feel of it even though it feels nothing like real skin. In many ways it feels a lot better than a woman.

>>7678 >I wouldn't care as long as she is cute and her AI doesn't take over the world hard disagree on the take over the world part take the Dr. W pill