Okay so i know you guys are committed to democracy, I don't think its a meritocracy cause otherwise sophiedev would have led this project as soon as he made that robot head ff. But anyways whats in place right now is not even a democracy. Its the worst kind of unnerving anarchy that leads nowhere. In order for it to be a democracy there needs to be strawpolls with time limits and the will of the majority is set on stone from that point on. Can we please start doing strawpolls moving forward?

>>23121 Lol. Don't be 'unnerved' bro. Just state what you want to do, then do it. I'm sure others here will help you out if they're interested. Run whatever polls you'd like. Start your own board/forum. If it's legit, I'll even add it to our friends list, as I've done before. Good luck Anon! :^)

>>23123 Okay ready to give this a try, let's see https://strawpoll.com/polls/NoZr35bQay3

>>23124 This one video might not go down well but I thought it'd be funny. https://youtu.be/zBCrz9ZE0SM Anyways I think you won't be hearing any more complaining from me moving forward. >>23126 Isn't that going to be a fleshlight?

Robot expert Dr Scott Walter talking about how fast Optimus will replace jobs and why it's also economically sound: https://youtu.be/pqdouW3g1Xw - It's going to be quite fast, especially for simple and repetitive jobs which the other robots couldn't do. Many other jobs will follow. Same guy comparing humanoid robots (I'm not finished with that one yet): https://youtu.be/z9JlI2ZmqAU

>>23124 How is it going?

>>23164 Any idea what the question was Anon?

>>23163 Thanks Noidodev. Clearly this is exactly what they are targeting Optimus for. Exactly. First at Tesla, then an open market for any business willing to pay for them. Musk's comment about the price target for these humanoid worker bots was US$20K .

>>23168 >Any idea what the question was Anon? You can't read it because of Tor? >what should we start with first? ... I'm inclined to call this trivia spam. Just doing something and writing something. Maybe it's trolling, maybe it's not.

>>23124 looks like the exoskeleton won... well that one is straightforward um cause we have the stl files form the maidcom project so its just a matter of getting ordered really... if we're going with that one.

>>23185 well no not exactly. the elbows the knees and the mouth need to be moded and among other thing I might have missed I think.

>>23184 >You can't read it because of Tor? Didn't try. I said that Anon is free to run polls didn't say I myself will participate in them. We as a group already have many, many things on the 'robowaifu design table' to solve, that we already know about. Why fuss and moan to add more things when we have way too much to do already hehe? :^) >=== -minor edit

I'm just going to go ahead and say it. I'm here to do two things 1.make a robot 2.make some money what I'm not here to do 1.make a virtual girlfriend If you don't have 2 grand laying around somewhere for this project I suggest you do like the OPs pic and get a job at 711 for a month or two. But I know you do cause even I have a $2000 to spend on this.

>>23203 >thinking this only costs 2k better start saving anon

>>23210 well maybe it costs more in the us and or europe but other than the 3d printer all electronics components tend to be cheaper here in thailand since its closer to china and I don't have to wait for months. If anybody needs a cheap electronic component I can hook you up btw. I can save on shipping big time by just sending on an envelope I think.

>>23211 spoke too soon.. hm looks like the prices at amazon are lower now or is it just me...

>>23203 >1.make a robot >2.make some money Excellent. Why not make 2x that, since you're there already? >>23211 >>23212 I believe he's implying the R&D, tooling, debugging, re-R&D, re-tooling, re-debugging involved in any major prototyping effort. This phase is always much more expensive per-unit than any production run. And for good reasons, obviously. You can liken this effort a bit to some blackops X-plane program in that sense. You didn't really think they were spending US$20K per toilet seat, now did you? :^)

>>23203 A virtual girlfriend is a subset of what we need plus some additional software. Certainly easier to make it into a commercial product.

>>23216 I understand its a subset of this project but its called /robowaifu/ not /virtualwaifu/

>>23226 Excuse me, but what's wrong with working on a subset of the project? Isn't that what the million threads on this board is about? President Kennedy didn't go up to John Glenn and say, "Excuse me, you only made it around Earth, you were supposed to go to the Moon." Also, that subset is arguably the hardest subset to work on.

>>23228 time to hang a lemon

>>23228 The thing is that you're not concerned about the ai regarding movements even. You're concerned about the conversational part which I think other organizations are working out and its really not a big deal. Its really not.

>>23231 >The thing is that you're not concerned about the ai regarding movements even We're all looking forward to you solving that for everyone here Anon! Eagerly. :^)

>>23240 it can't even be solved without knowing which components are going to be used for the robot.

>>23241 Well, maybe you can read the board thoroughly through, then you'll probably know what good targets are. I'd suggest you simply abstract your efforts down to nothing more then libraries targeting Arduino control pinouts. That way when you're ready to begin tuning your AI for specific actuator torques, lever-moments, and control algorithms like Inverted Pendulums, you'll have already covered the practical basics at the microcontroller level. Good luck Anon! :^)

>>23248 Just made this now https://wokwi.com/projects/367728728858599425

Seems awfully quiet around here, I don't know if its because its the weekend or some people feelings got hurt. I happen to have a little x100 dev on one of my tabs, his first name is chat and his last name is gpt and I'm good at filling in the blanks. If you feel some kind of way don't let the door hit your ass on the way out. Pic unrelated, its just my favorite videogame character heh heh.

SInce its slowing down a bit, maybe I should ask a question. Do any of you guys feel/hope that megacorporations AI progress slows down. NOw, don't get me wrong, I'm an AI radical and want AGI as fast as possible with no regulations or brakes. However, I don't trust these corpos, especially how they wanna regulate and monopolize the AI market. I think we'd "lose" if they got to AGI before "we" did. I'm kinda hoping that the open source AI community picks up speed and hopefully one of us manages to make AGI. I think it'd be a much better outcome for everyone. Although, its probably wishful thinking since the hardware requirements for AGI is colossal and can only be afforded by the likes of Kikesoft, Don'TBeEvil and ClosedAI. I atleast wish that virtually the entire FOSS AI community wouldn't focus on LLMs atleast because thats a dead end for general intelligence and we'd see more projects on that instead.

>>23253 Thing is that there are multiple projects about that and they're opensource. Dream booth, llama, koboldai, etc... and if you felt you could contribute you'd be contributing towards those. But you're not. You're here at /robowaifu/ with a title that says robotwaifu technician.

>>23254 Well there are resources for that as well and I'm sure training existing AI has been made more accessible though such as this. https://developers.google.com/machine-learning/crash-course But I'm kind of disappointed at the lack of support and enthusiasm for a robot and everyone focusing on the freaking chatbot. I don't know about you but I would like to have something to stick my dick in that moves too. You down play it, I'm not going to down play it. Trust me there are other people working on that. Can we please focus? I don't think I'm being unreasonable.

>>23255 well here's a link with various free open source alternatives to chatgpt. https://www.kdnuggets.com/2023/04/8-opensource-alternative-chatgpt-bard.html Ugh whatever

I have yet to see a doubter proven right yet bucko

>>23279 Sadly, K Decker will have to merge yet again, elsewhere the Chikun'g maybe?, unless you can come up with better OP content bro.

>>23279 Sorry, I’m new to this board and wasn’t aware of the abundance of blade runner posts. Will use different image next time .

>>23284 no is just that this board is really cramped so you shouldn't start threads at all... I got my own chan btw, I offer it to start various boards for this project so people can start threads and stuff if you're interested chobitsu...

>>23284 Lol, I meant please make some kind of effort-post if you want to start something like a thread. Otherwise, /meta (>>20356) serves dual-purpose as our QTDDTOT which is all your simple post requires. Blade Runner is fine. Nice pic on your new post BTW. >>23286 >I got my own chan btw, I offer it to start various boards for this project so people can start threads and stuff if you're interested chobitsu... OK, let's see the chan Anon. BTW, we're only up to page 19 ATM, with 25 total possible (10 threads per page), so we still have room. It's just that we're trying very hard to always stay on-topic here. This thread simply doesn't qualify as legit, given it's simple OP. That's all.

>>23287 my chan is https://vipbiz.finance

>>23289 Well if I could have a temp mail please, I could email you an account and password for you to add whichever boards you deem necessary if you'd like...

>>23290 oh well here's a try I know nobody is going to go to it, but its there if you need it. https://vipbiz.finance/hands/index.html

>>23289 Thanks! I believe you've already introduced yourself here before haven't you Anon? >>23291 This isn't me: (>>23290), BTW. I'll generally be quite clear about things of this sort if it's me, Anon. This isn't my first rodeo in these here parts. :^) >>23290 >I could email you an account and password for you to add whichever boards you deem necessary if you'd like... What are you saying here friend? Reword pls. >=== -minor fmt, edit

>>23291 Why not start some content on your pages, Admin?

>>23316 I would but that'd require some effort from my part and I don't know if that effort would be rewarded at all really. I don't think most people around here are serious about this project tbh and I'm not seeing any willingness for anybody around here to contribute. I think I should leave really cause I'm being a buzzkill, but ideally I'd like to find people that want to do this thing. Definitely not here though.

>>23324 Fair enough. Please stop by again sometime in the future to let us know when you've discovered your tribe and your project has begun. Cheers. :^)

Here my current plan for AI hardware: - I ordered a used K80 with 2x12 GB recently, a used one of course, for 100$/95€ shipping included. It's an old GPU, only supported by older CUDA versions and might not run quantified models. Uses much energy, but it's two GPUs with 12GB each. I plan to pair this one after a while with a RTX3060 (12GB, 300€ used or 400€ new I think) in one home server. Context: https://technical.city/en/video/Tesla-K80m-vs-GeForce-RTX-3060 - That one, will then run my 12GB models. For fine tuning, or models which don't run on the K80, I would use the 3060. I don't know yet if I can somehow joint them together and use 3x12GB through the bus. It just seems to need some software support in these programs for running models at home. - I plan to use online services like Colab to learn about how to run these things, but have the K80 for more private data and learning how to do these things at home. - Then I'll get some SBCs, most likely Orange PI's, which can run small models of Whisper (speech regocnition). Also, another small home server with a Intel Arc380 (140-160€), which is fast enough to run the big and better model of Whisper at the speed of one fast human speaker. It does this quite energy efficient. These devices will not run anything else, for security reasons, and be connected to microphones which will be always on. The server will receive the audio through the SBCs from all rooms using the home network (likely on a VPN using Tinc). All of them will send the transcripts to some server in my network which can then decide how to respond. Most likely filtering first for which data is more sensitive than others. - Some small device, like a Raspi, will maybe handle responses based on AIML or using some small model. - Questions which don't contain private information might be send to OpenAI or another service. - The next step up will be getting a M40 (180€) and then a used RTX3090 (700-800€ right now I think), putting them in another home server at some point. Of course I might use this one for gaming till I get even the next GPU. These can handle the models which need 24 GB. The 3090 will do the fine tuning if I want to do that, since it has more power, while the M40 doesn't need as much energy. Context: https://technical.city/en/video/GeForce-RTX-3090-vs-Tesla-M40-24-GB - Then the next step might be getting a AMD XTX (1k-1.2k€) if it's supported well enough for AI by this time. I can use this one for gaming and then put the 3090 in a home server with the M40. If it's possible to combine cards using PCI express, then it might be interesting to think about getting another XTX later, and have 48GB vRAM. - But I hope that either Intel or AMD will come out with a prosumer or consumer card for AI at home, which is rather slow but has 48GB and is not too expensive. (If you buy K80 or M40 on Ebay make sure not to buy the 12GB versions by accident while only looking at the price. They aren't much cheaper. K80 should have 2x12GB and the M40 24GB.)

>>23346 I hope the K80 works for you. I was thinking of getting two but the support for them seems abysmal. They used to be $200 used before shipping. Colab isn't what it used to be either. The free version will boot you off in 30 minutes sometimes or a few hours into training with Pro unless you pay big for compute credits. You're much better off running your own JupyterLab notebook or renting an instance off vast.ai or runpod.io if you don't have access to a GPU.

>>23351 >support for them seems abysmal. I think you need old versions of the software, but I also remember some people taking care of that, to support old GPUs. I might need to compile some of it myself, though. I hope it works out, but the risk isn't very high. >They used to be $200 used before shipping. They're down to $60-70 before shipping now. Some recommend to go straight for the M40, which is much newer, but $120-130 before shipping. >JupyterLab notebook or renting an instance off vast.ai or runpod.io if you don't have access to a GPU. Right, I forgot about those while writing this.

>>23346 Good luck, NoidoDev!

NEW THREAD NEW THREAD NEW THREAD (>>23415) (>>23415) (>>23415) (>>23415) (>>23415) NEW THREAD NEW THREAD NEW THREAD