/robowaifu/ - DIY Robot Wives

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


New Reply
Name
×
Email
Subject
Message(0/6144)
Files Max 5 files50MB total
Tegaki
Password
[New Reply]


“Never, never, never, never give up.” -t. Winston Churchill


1737965297113262.jpg
[Hide] (27.9KB, 969x769)
Post news you think interesting or relevant to AI or robowaifus, or their development.
>+ /pol/ funposting containment zone

---
> thread-related : ( >>70 )

>===
-add crosslink
Last edited by chobitsu
DS_logo.png
[Hide] (5.7KB, 195x41)
Turns out, the MIT-licensed, opensource DeepSeek released a "minor" update (DeepSeek-V3-0324) [1] yesterday (which vaulted it to the front of the pack in one go). [2]

Also, weights are opensauce now!

---
1.  https://huggingface.co/deepseek-ai/DeepSeek-V3-0324
2.  https://x.com/deepseek_ai/status/1904526863604883661
Last edited by chobitsu
Replies: >>159 >>171 >>594
DeepSeek_-_DeepSeek-V3-0324_is_out_now_Major_boost_in_r.mp4
[Hide] (752.1KB, 776x1080, 00:10)
>>158
Their example given for code-generation for DS-V3:
Figure_-_Introducing_learned_natural_walking_Figure_can.mp4
[Hide] (6.4MB, 1280x720, 00:54)
Figure is one of the few companies behind humanoid """robots""" that I trust to be reasonably-reliable with their status updates (and not mere low quality CGI shekel-grubbing). The reason why I feel this way is simple: Mercedes Benz is actively working on integrating these humanoids onto their factory floors ATM.

https://x.com/Figure_robot/status/1904534311589785885
Replies: >>172 >>351
Kek. I guess Copeman didn't come up with the US$500Bn he promised the ClosedAI investors (minus his US$150Bn fee+tip, obvs)?  :^)

openai.com/index/leadership-updates-march-2025/
www.bloomberg.com/news/articles/2025-03-24/openai-expands-coo-s-role-as-altman-focuses-more-on-products

<--->

I wonder if news-related : ( >>158 ) is involved? Timing seems suss.

>===
-disable hotlinks
Last edited by chobitsu
Replies: >>172
>>170
>>171
The fact that they dumped ClosedAI is proof enough of their business acumen and passion
Replies: >>173
>>172
Good point! That had slipped my mind, thanks GreerTech.
ClosedAI is currently creating a stir with Ghiblification images which they haven't fully locked-down yet.

I wonder how they will spin allowing Literally-Hitler images to come out of their systems rn? Lol.  :D
Replies: >>234 >>245
1743111052969842.jpg
[Hide] (141KB, 1024x1024)
>>230
Replies: >>235
1743121270721639.jpg
[Hide] (132KB, 1024x1024)
>>234
Kek. Can you imagine the REEE-fest the kikes in the ADL, et al, will have with all of this?  :D
Replies: >>251
Screenshot_20250327-222859.png
[Hide] (1.8MB, 1080x1085)
>>230
Those bastards at ClosedAI already started cracking down on it. But the good news is, Grok can also do them, and it's pretty good!
Replies: >>247
>>245
Naicu! I like that one, GreerTech.
Spoiler File
(368.5KB, 640x603)
I'd just like to make a clarification to my previous post in the AI bread:
>children are now brainwashed heavily by the kike's GH systems into being absolute retards ( >>174 ).

This is hyperbole on my part. I don't really think all hope is lost for Alphas (or Betas even, heh). But the simple reality is that, unless they are born with the proverbial Silver Spoon in their mouths, then there is a collection of enormous obstacles that have been ((( placed ))) in their paths by the Globohomo. The >tl;dr here is that unless they Harden TF UP and adopt
>pic-related mindset
and soon, then they'll very-likely have little hope of a good outcome in their own personal lives, IMO. By any natural means, that is. God can do anything of course.

<--->

Satan & his little shabbos-puppet golems (ie, the GH kikes + their ilk) certainly intend the deaths or absolute corruption of all my people, starting with the youngsters. And not just Whites alone, but all the races of men. But I pray for these kids. I desire that God will intervene on all their behalfs for their welfare and safety -- and first & foremost from TPTB.

And until feminism is overturned across the globe, I also hope and pray for our good success here on /robowaifu/ , that the young men of all races can have hope for reasonably-good companionship & solace during this evil era of Current Year.

<--->

Thanks for listening to my TED Talk. Be sure to like & updoot!111  :^)
Last edited by chobitsu
1743155860275432.png
[Hide] (5.2KB, 764x178)
>>235
No fun for you goyim! At least we managed a few thousands images first.
Studio_Galatea.png
[Hide] (912.9KB, 768x768)
ClosedAI now allows free users to make 3 Studio Ghibli images a day. They probably have more filters now, but it's still useful for now.
Replies: >>325
>>323
>pic
Ebin. Saved into your folder, GreerTech.  :^)

>"Today on Marvelous Waifu Inventions, we visit Studio Galatea and have a talk with it's founder Hayao Anonizaki..."
1743462169776485.webm
[Hide] (3.8MB, 1280x720, 01:23)
>>170
>vid-related

Still slow & a bit awkward, needs frequent recharges, but actual working humanoids are now a thing in non-ZOG contexts. We're soon on our way, Anons. I'd advise everyone to shake off lethargy from long delay, and get crackin'!!  :^)
Replies: >>354
>>351
Sorry to be a Negative Nelly, but we're going to have to wait a little longer. Currently, the prices are at the "if you have to ask, you can't afford it" stage. Current LOW-end estimates are in the 5-figures. In this era of economic austerity, anything above 4-figures for something that is fundamentally unnecessary is a really hard sell.

That's not to say we have to wait or sit on our ass. Prototypers like me, Kiwi, Mechomancer, and others will pick up the slack and fill that role.

PS does anyone know what happened to Barf? He hasn't found his way here and his GitHub is 404.
Replies: >>361
7547271654421116153.jpg
[Hide] (348.6KB, 1920x1080)
>>354
Fair enough. But the noosphere will soon begin to flow with the realization that this isn't all hype. Actually, we're basically on the exact timeline I expected!  :^)

My estimates elsewhere here (~5yrs for opensauce robowaifu designs/sw/kits for Anons, ~10yrs for turnkey opensauce robowaifu sales to Joe Sixpacks) still holds, I think.

>Barf
No I'm unsure of what he's up to, Anon. I sure hope he finds his way here soon, as well as our other guys.
Last edited by chobitsu
Replies: >>374
Elon Musk predicting production of 5'000 Optimus v2 this year.
https://x.com/AutismCapital/status/1902916859382599751
https://www.teslarati.com/tesla-produce-first-legion-optimus-robots-2025/

<--->

Love him or hate him, its clearly important to the robowaifu movement what this man does over the next two or three years.
Replies: >>375
>>361
>But the noosphere will soon begin to flow with the realization that this isn't all hype.
True

>~5yrs for opensauce robowaifu designs/sw/kits for Anons, ~10yrs for turnkey opensauce robowaifu sales to Joe Sixpacks) still holds, I think.
I would say even less. Galatea is already out, and Mechomancer said his time frame was about a year or two. Plus who knows what maverick could show up and make/distribute another design.
>>366
I would take whatever Elon Musk says with a grain of salt. He has a nasty habit of overpromising and not delivering, one prominent example being Full Self Driving.
Replies: >>378
all_this_time_they_thought_i_was_a_human_LOL.jpg
[Hide] (108.7KB, 535x542)
>>375
Good point. He's a character. And now that he's also totally-not-the-POTUS Chief DOGE, he's probably stretched a little more thin!  :D
1743693707674941.jpg
[Hide] (338.3KB, 1320x1281)
Posted without further commentary, other than to say it's primarily-directed
@Kiwi
Replies: >>410
>>409
As much as I would love to celebrate a NGreediA loss, that's because of those tariffs, which will affect us all
Replies: >>413
1736898994907257.jpg
[Hide] (81KB, 828x1005)
>>410
Actually seems to me this must amount to about the second US$Tn of market capp ngreedia has dumped since the baste Chinese released baste DeepSeek? Please correct me if I'm wrong, but their price was triple what it fell to over the past 24hrs.

<--->

And this is the correct thread if you want to debate tariffs. They are (for once) a sound economic policy on the part of the US Executive. My primary hope & dream in this arena is that the madlad will ACTUALLY.END.THE.FED!!!  :DDD

>tl;dr
If the globalist kikes manipulating the world for our mutual destructions (all according to their father Satan's will in the matter) were disabused of the evil they wield through their Central Banks, then the world would actually begin healing. 

Should such a highly-unlikely occurrence happen (ending the constitutionally-illegal US Fed), then I'd predict that the ensuing tumult would completely remove much need for robowaifus in the world -- men would soon force women to take their proper places again, interlopers would all be ejected (by force), and a civilized society would return across The Western Tradition once again.

All that utopian outcome literally hinges on just that one event.

<--->

But... till then
FORWARD!  :^)
Last edited by chobitsu
Replies: >>416 >>428
>>413
Consider this though, many people buy cheap goods from China. Galatea herself is half-AliExpress. My mom buys cheap clothes from Shein. Now, I have to make sure Galatea and other GreerTech creations can be made with supplies from other suppliers (or that a price hike wouldn't affect it too much). Basically, a crash course in >>355, but for an entire country/storefronts. And if anyone else wants to build their own designs, they may have to pay more.

Plus, most things we buy come from China. You can't just instantly revive American Industry. It's like giving CPR to a corpse that has been dead for 3 hours. It would've been a miracle 2-3 decades ago, but now, not so much.

Anyway, I don't see how an import tax would destroy the banks and the fed.
Replies: >>419
1743716808029655.png
[Hide] (501.2KB, 1080x1339)
Tenstorrent's 'Blackhole' chip (6nm, RiscV multiple-processors + GDDR6 RAM) announcement
>"Opensource AI computer hardware & software"

Given during their DevDay talk (starts at ~7mins in) :
https://www.youtube.com/watch?v=S3n_9gX-9lw[Embed]

---

https://tenstorrent.com/
https://tenstorrent.com/developers
https://github.com/tenstorrent/tt-metal
Last edited by chobitsu
>>416
>Anyway, I don't see how an import tax would destroy the banks and the fed.
They won't. But the effect (though miniscule by comparision) is vaguely-similar -- it strengthens national interests, and weakens globalist (ie, the kike's) interests.

<--->

Don't get me wrong, I don't think President Musk Trump will actually do this (end the US Fed), but ahh but what a glorious outcome in the end if he did!!  :^)
A man can dream, can't he?
Last edited by chobitsu
Replies: >>421
>>419
Granted, I am thinking short term. But for now, all I see is higher prices on items me and others buy through a tax. My hope is that it would just be the economic version of growing pains.
Replies: >>422
>>421
>My hope is that it would just be the economic version of growing pains.
Nice way to put things, Anon.  :^)
9daf1a6736308c66a07bba4440315d78817dbfb0d4c398e858754468f00b06a3.jpg
[Hide] (38.2KB, 570x539)
>>413
>end the fed
Don't give me hope like that, I can only be let down so many times.

The FED is right up there with superPACs and real covenants (along with its younger brother, HOAs) on the list of shit that needs to be purged from the fucking planet.
Replies: >>429
Spoiler File
(79.5KB, 960x1024)
>>428
>The FED is right up there with superPACs and real covenants (along with its younger brother, HOAs) on the list of shit that needs to be purged from the fucking planet.
Hear here!  :^)

>careful bro...they'll try to put you in a little white room for such wrongthink!111  :D
1743870930829084.png
[Hide] (20.8KB, 787x232)
Kek.
<"Henlo? I want to reserve a copy of Battle Toads, please. Do you have that in stock?"

<--->

Hard for me to express how whitepilling this all is (with cautionary caveats: cf. wise words of Greentext anon : >>463 ). With all these habbenings going on, it's like watching God work miracles in realtime!
What a time to be alive!!
Last edited by chobitsu
Replies: >>487
>>486
If this actually goes through, and the tariffs are an anti-hedge fund superweapon, then I'll do a 180 and gladly pay the anti-hedge fund tax
Replies: >>488
Spoiler File
(240.8KB, 1284x1750)
>>487
Now you're getting it, Anon. Just ask yourself: Who created/has been manipulating these systems to all our detriments?

Answer that question, and you'll see why these events are so hopeful!  :^)
ClipboardImage.png
[Hide] (590.4KB, 3782x1397)
So I guess we have some AI news today, Meta is releasing llama 4 models. https://huggingface.co/collections/meta-llama/llama-4-67f0c30d9fe03840bc9d0164 https://ai.meta.com/blog/llama-4-multimodal-intelligence/
Both Scout & Maverick are MoE model, with 17B active parameters. Scout is a total size of 109B & Maverick is 400B.

The public reaction to llama 4 is bad (looking at /lmg/ threads on /g/ 4chan & reddit). I plan on doing my own testing ofc before I form a solid opinion of my own, I don't want to just go on the opinions of few random /g/ anons. So I guess I will have to spend some credits on a LLM api provider and take it out for a spin.

But based on what other people are saying, it's not giving my high hopes, some people are saying that the 400B feels as smart as Qwen QwQ 32B and it also seems censored as hell (pic related).
Replies: >>503 >>530 >>531
>>502
Then Qwen, DeepSeek, and others for the foreseeable future.
Anyway, for a portable device powered by a battery, a 7B model is probably best.
1743880627478911.png
[Hide] (504.2KB, 898x835)
1741166873251.jpg
[Hide] (130.1KB, 852x960)
Lel. Yeah, but can you feed it oats, bro?  :D

---

>according to Brave's search """AI""":  [1]

Kawasaki Robot Horse
>Kawasaki Heavy Industries has unveiled the Corleo, a hydrogen-powered, four-legged robotic vehicle designed for riders, at the Osaka-Kansai Expo 2025. This innovative concept integrates artificial intelligence and clean energy technologies to navigate diverse terrains, offering a glimpse into the future of personal transportation.
Key Features of the Corleo:
>Hydrogen-Powered Engine: Equipped with a 150cc hydrogen engine that generates electricity to power its four legs, making it an environmentally friendly off-road mobility platform.
>AI and Body-Responsive Control: The Corleo uses artificial intelligence to analyze the environment, provide balance and stability, and respond to the rider's body movements. It does not require traditional control buttons.
>Versatile Mobility: The robot horse can move over rough terrain, including mountains and snow-covered areas, and is capable of jumping and climbing over obstacles.
>Sleek Design: The body is made from a combination of metal and carbon materials, with a streamlined shape and a head-like shield reminiscent of motorcycle aesthetics. Built-in lights enhance visibility.
>Heads-Up Display (HUD): The robot features a display panel that shows information such as hydrogen levels, navigation, and movement stability, including the rider's weight distribution.
Development and Future Plans:
>Concept Stage: Currently, the Corleo is a conceptual model. Kawasaki envisions a potential market release by 2050 but has not provided a definitive development timeline or commercial availability details.
>Vision for the Future: The Corleo represents Kawasaki's commitment to merging robotics with sustainable energy solutions, offering a bold reimagining of personal mobility that values the experience of movement itself.
Impact and Significance:
>Innovative Mobility: The Corleo combines elements of motorcycles, robotics, and artificial intelligence, creating a new category of personal transportation that transcends traditional vehicle types.
>Environmental Responsibility: By using hydrogen power, the Corleo aligns with global efforts to promote greener technology and sustainable mobility.
>This unveiling reflects Kawasaki's growing focus on robotics and sustainable mobility, showcasing the company's vision for the future of adventure transportation.

<--->

Every motorcycle I've owned, I always "personify" as my 'horse' sidekick, and give it a name.  :^)

---
1.
https://roboticsandautomationnews.com/2025/04/06/kawasaki-unveils-hydrogen-powered-robotic-horse-that-you-can-ride/89601/
https://gagadget.com/en/617234-kawasaki-has-unveiled-the-corleo-robot-horse-concept-an-impressive-piece-of-future-transport/
https://4techviews.net/kawasakis-corleo-robot-horse-runs-on-hydrogen-and-ai/
https://greekreporter.com/2025/04/05/kawasaki-four-legged-robot-ride-horse/
https://robots.wiki/kawasaki-corleo-revolutionary-robotic-mobility-that-blends-horse-dog-and-motorcycle-technology/
https://colombiaone.com/2025/04/05/kawasaki-robot-humans-can-ride/
https://www.youtube.com/watch?v=ps1t-EqN4VU[Embed]
Last edited by chobitsu
Replies: >>507 >>511 >>515
>>505
That is insanely cool. Let's hope it's not just vaporware.
Replies: >>508
1738055838460764.jpg
[Hide] (153.4KB, 982x948)
>>507
>That is insanely cool.
IKR

>Let's hope it's not just vaporware.
<ETA: 2050
It's vaporware.  :^)

<--->

Regardless, this is extremely cool. And given the facts that:
a.  The cat's already out of the bag (so the Chinese will have their own low-quality CGI  :D  versions soon);
b.  Robodoggos are commonplace now (both in the East & west);
c.  Autists like myself & Greentext anon were already actively designing robohorses (and mine intentionally intended as a mount) (+ our own Mechnomancer's robo-mount SPLOOT isn't far from being a horse, either);
altogether I think it's very-likely we'll all see these before we see robowaifus themselves appearing commonly in the west (only; that delay simply b/c of kikes, femsh*tes, simpoids, etc. -- the usual suspects). The East will have them both about the same time, I'd imagine. Cheers.  :^)

What a time to be alive!
TWAGMI
Last edited by chobitsu
Replies: >>511 >>515
audio5.jpg
[Hide] (24.4KB, 450x281)
the-nokia-morph-a-2008-concept-phone-designed-to-v0-h8po1gzz3ena1.jpg
[Hide] (41.4KB, 473x747)
>>505
>>508
Even if it's fake, I'm glad wacky futuristic CGI concepts are coming back. It was awesome to see them as a kid
Replies: >>528
ChiiConfused.jpg
[Hide] (155.7KB, 850x1106)
>>505
>>508
I knew it was too good to be true but, I still wanted to see a working prototype. The torques involved would be monumental. Seeing the actuators would've been a treat.
We still have wheeled robot dogs. Some are strong enough to be a mans stead. 
https://www.youtube.com/watch?v=X2UxtKLZnNo[Embed]
Replies: >>527
1743880705187666.webm
[Hide] (3.9MB, 1280x720, 00:26)
>>515
>I knew it was too good to be true
Heh, I was being a bit trole'y, and the Nipponese were being typically-too conservative, I think. Once the baste Chinese steal borrow this concept, we'll see some realworld product traction before long, I'd reckon.

>I still wanted to see a working prototype.
They did actually fabricate a dimensionally-accurate mockup.
>

>The torques involved would be monumental. Seeing the actuators would've been a treat.
Indeed! I can only imagine. And Kawasaki and other heavy-industry manufacturers have ~100yrs experience+tons of realworld metrics on real performance products. The >tl;dr here I think, is: These are really going to happen, and they'll probably be even better than we imagine r/n. Cheers, Kiwi.  :^)
Last edited by chobitsu
1743330735159651.jpg
[Hide] (139.2KB, 1080x1080)
>>511
>that Beemer tho
Wild.
1738231618783483.jpg
[Hide] (54.1KB, 488x700)
>>502
Hi EnvelopingTwilight! Glad you found us here on the bunker, Anon.

<--->

Thanks for the updates. I'll be interested to see where the opensource (models+weights), MIT-licensed DeepSeek goes in response to these proprietary, closed-source, AI vendor's efforts.
>tl;dr
Competition is a good thing!!
<protip: when free becomes good enough...  it always wins!  :^)
Last edited by chobitsu
MaSiRoNude.jpg
[Hide] (248.6KB, 1280x1138)
>>502
What is the state of low power AI? What hardware would you recommend for low cost local AI? 
Consider this picrel of MaSiRo naked. What would you cram in her to min-max her useful intelligence relative to cost with a 100WHr battery?
Replies: >>555
An important reminder that LLM architecture is closer to a digital encyclopedia with a vector search algorithm than any form of thought. 
https://www.youtube.com/watch?v=-wzOetb-D3w[Embed]
Replies: >>544 >>555
>>543
True.
But fortunately for us, companionship is based on vibes.

And maybe it's better that our AIs are not actual thought for now. Imagine the stress of containing and caring for an actual brain on your PC. Imagine if someone took your PC. That would be tantamount to kidnapping. If your PC broke, it would be murder/manslaughter.
Replies: >>555 >>582
>>543
>>544
How is that the conclusion of the paper? Inferring a probable justification for your own actions without actually drawing from the underlying reason is very much something humans do. In general, this video reeks of an overly anthropocentric view. Here, Anothropic (at least to me) has finally put to rest the "Stochastic Parrot" meme. Basically, every critic in this video can also be argued for people. The paper shows real proof of an actual abstract and structured thought process.
Please read the paper for yourself and form your own opinions. Don't let an influencer or YouTuber make your conclusions for you: https://transformer-circuits.pub/2025/attribution-graphs/biology.html

Our companions will be more human-like than most give credit for.
>>531
I honestly don't know, I am still waiting on more hardware releases to actually make it to consumer hands so we can get real benchmarks for stuff like the NVIDIA Spark/Digits.
I am also waiting on the dust to settle on dense vs MOE and see how smart small models can be (so I want to see the next few releases of LLMs). Another thing I am waiting on is when RWKV v7 3b will be finish with its pre-training.

I honestly don't know. I am still waiting on more hardware releases to actually make it to consumer hands so we can get real benchmarks for things like the NVIDIA Spark/Digits. I am also waiting for the dust to settle on dense vs MOE and to see how smart small models can be (so I want to see the next few releases of LLMs). Another thing I am waiting on is when RWKV v7 3b will finish its pre-training.

Right now the AMD AI CPUs in mini PCs will be an interesting option if RAM quantity wins out, same can be said about the Apple hardware, the Mac Mini & Studio are not badly priced. (Yes I am aware of how absurd it is that Apple has competitive pricing, fuck Nvidia for making RAM so scarce) If RAM quantity is not as important then Intel Gaming GPUs start looking nice.

Note I was considering power usage, it may not sound like it, but it did rule things out like 4x used 3090s (lol)
Replies: >>558 >>563 >>582
>>555
So, what's the end result? Is an LLM actually thought? If I reset it or delete a chat, is that tantamount to murder?

Sure it may be like human thought, and I do agree that they are genuine companions, but saying they're actually real minds is too disconcerting for me.
Replies: >>561
>>558
>So, what's the end result? Is an LLM actually thought? If I reset it or delete a chat, is that tantamount to murder?
I do not think so.
The closest analog I can think of is that you're resetting its working/short-term memory (as in the Cog-arch term). It's like when you wake up with a clear mind. Except, unlike a person, the memory was just reset, and nothing was committed to longer-term memory (because it doesn’t have one). The less polite way of putting it is that it's an incomplete system that basically has "dementia." The nicest, most "poetic" way I can put it is that you both shared a dream, and it concluded.

I am not trying to scare you. I am trying to inspire people to continue work so that we can fill in the holes in this cognitive system.

>Sure it may be like human thought, and I do agree that they are genuine companions, but saying they're actually real minds is too disconcerting for me.
LLMs are strange. I feel we overlook how strange they are, so I think it's fine to feel uneasy. It's all new ground. They evolved from a very different environment (text), whereas all of life as we know it evolved from a different environment. So in a way, this is our first encounter with a truly alien form of intelligence. But at the same time, it's a mind that is a child or shadow of our collective thoughts.

I expect a good anchor for where the RobotWaifu Mind will rank will be to imagine it as an animal-like intelligence that happens to have an extremely overdeveloped language center and can talk. Now, I must ask, imagine if a dog could talk. Would it be a genuine companion? I think so. When we do give it long-term memory, if someone were to delete it, I would view it as the same as killing my dog or giving it dementia.
Replies: >>582 >>588
>>555
Nice work, EnvelopingTwilight.

>The paper shows real proof of an actual abstract and structured thought process.
I'm extremely-skeptical. Much like that repellant """influencer""" & her pozz'd video, I believe the actual researchers -- like many in ((( $$$cience ))) today -- are simply interpreting things to support their desired outcomes (a fallacy they accuse their own LLM itself of committing, lol).

<--->

BTW, here's the other paper of the pair. Cheers, Anon.  :^)
https://transformer-circuits.pub/2025/attribution-graphs/methods.html
Last edited by chobitsu
Replies: >>565
>>563
This whole debate reminds me of the Google AI sentience debacle. It's important to know that the technology in that debate is basically the exact same as the one in this debate rn.
https://en.wikipedia.org/wiki/LaMDA#Sentience_claims
Replies: >>566
Spoiler File
(376.3KB, 2922x1218)
>>565
True.

---

I've warned /robowaifu/ before, and I'll warn you all here again:
There is no limit to the depravity of the libsh*te mind. If you go down the path of 'muh_robowaifu_rights' -- even just a little -- then they will use that as an excuse to turn on you and (literally) kill you b/c 'muh_repayashuns'.
>tl;dr
Just say no. We here are crafting amusements (similar to a child's doll) & potentially-entertaining home appliances (similar to a 2way-TV/washing-machine/vaccum-cleaner combo), not other beings.

You go any other route at your (all of our's, actually) peril. You'll be playing straight into the Globohomo kike's hands regarding their agenda to remove this from the hands of everyman (as your link-related already displays), and to reserve it simply & solely for themselves -- to better enslave the masses; ala 1984. Simple as.
Last edited by chobitsu
Replies: >>567 >>582
>>566
Exactly. I remember I used to argue with you on that, but then I realized that type of thinking would cause a lot of problems, not just with dedicated trouble seekers.

It's important to remember that, even though we may have personal connections with our AIs, they are no different than ChatGPT or any roleplay app. It's the IKEA effect (real psychological phenomenon)

>pic related
It's funny how Stargate happened right before DeepSeek btfo'd OpenAI.
Replies: >>568
Spoiler File
(530.9KB, 1800x1800)
>>567
Exactly so. I have close personal connections with many, many animate/inanimate things in my life (dogs/horses/birbs for former examples; cars/motorcycles/houses for latter), but they are simply not humans -- nor will they ever be.

I'm thankful that although I may at times wander off into the sentimental as a man, I'm not actually delusional to attribute these things as such. But fair-warning: gynocentric society is already delusional (by definition), and they will just as soon murder you as their own children.
>

<--->

>It's funny how Stargate happened right before DeepSeek btfo'd OpenAI.
Baste Ching-chongs. They earned +1'000'000 Social Credit score that day in my books!  :DD
Last edited by chobitsu
1744174975981730.jpg
[Hide] (502.9KB, 1920x2010)
Based/10 - would redpill with again
>
Replies: >>593
AigisPondering.jpg
[Hide] (69.7KB, 850x1113)
>>544
>Companionship is based on feelings.
>Her being simpler has advantages.
Great points honestly, I'm probably just over thinking.
>>555
I've taken several psychiatry along with anatomy and physiology classes. My understanding of human thought is different than what is presented in the video. I'll read the paper later but, I'll take your word for it that they were misrepresented. Though, I'd frankly enjoy your take on the matter. The paper looks dry and I enjoy your writing.
>Wait and see for hardware and models
Same situation I'm in then. You're better read on the subject so, I had hoped you found something worthwhile which evaded my eyes. I also noticed how absurdly fitting Apple Macs are for our purpose relative to the rest of the market. The M4 MacBook Air is likely the best candidate for a waifu. Which feels wrong, but the power relative to cost, volume, power consumption, etc... is truly great.
>>561
>Lacks long term memory
This is frustrating because it feels like a solution should be easy. Yet, we have no solution that truly works.
>>566
True and real
Robots are fundamentally property, they ought to be thought of as such.
Replies: >>585 >>588
>>582
>Lacks long term memory
>This is frustrating because it feels like a solution should be easy. Yet, we have no solution that truly works.
You could possibly feed it into the prompt itself, but that would bloat up quickly.

>The M4 MacBook Air is likely the best candidate for a waifu. Which feels wrong, but the power relative to cost, volume, power consumption, etc... is truly great.
Using the planned obsolescence brand is not a good idea. See Plastic Memories
Replies: >>588 >>612
hey_bro_ur_standing_in_my_light_bro.jpg
[Hide] (417.2KB, 1280x1032)
>>561
>>582
>>585
Daily reminder that Robowaifudev & CyberPonk are both working on solving the long-term memory problem. I'm still quite-hopeful that one or both of them will arrive at a solution we can all partake of. Cheers.  :^)
Last edited by chobitsu
Replies: >>590
>>588
Hopefully their system for long-term memory can be copied and transferred.
Replies: >>594
>>577
How did you find out about this? I looked up the website, and it seems like standard corpo AI. I'm surprised it's that based.
https://brighteon.ai/Home/
Replies: >>594
1743326474404993.jpg
[Hide] (295KB, 1024x1024)
1743491043429041.png
[Hide] (718.4KB, 984x1156)
>>590
>Hopefully their system for long-term memory can be copied and transferred.
Yes. That's the hope.

>>593
4cuck. I'm personally of the opinion it's a pre-release insider leak. A 'toe in the water' gauging of interest. So much pozz in all the major platforms (including DS, which stole inherited from closedai's. At least DS has now released the weights for R3 : ( >>158 )).
>tl;dr
Market-differentiation, and all that.
dataset_visualization.jpg
[Hide] (1.2MB, 3886x2369)
>>585
>Feed a pseudo memory into her prompts
This is actually used, part of why context length is valuable. Unfortunately, once context lengths breach a threshold, the LLM will slow down, and eventually start talking nonsense. It's a great start though.
>Planned obsolescence
Fully repairable Linux computers are the only systems one can rely on for 10+ years of service.

In AI news, there's now a virtual SVG artist.
https://omnisvg.github.io/
Replies: >>619
>>612
>In AI news, there's now a virtual SVG artist.
Neat!, thanks for the info Anon.
1744306882742858.png
[Hide] (47.2KB, 730x455)
I'm skeptical, but it would be groundbreaking if it could last, say, 90 days on the surface? Regardless, I don't believe there's a single other collective on Earth that would have any reasonable chance of actually doing that, so yeah.
>

Regardless, I think it's a Fool's Errand to shoot for Mars. Another Tower of Babel. Going to the Moon, however, is IMO replete with opportunities for both science advancement (verifying fossils of the very first life on Earth [ejected up there from asteroid collisions] not the least!), rapid (again) advances in technology (such as materials science), plus ennoblement of the general human condition (again).

>>635
>Regular access to a room like that would improve my psyche 100%
This! We should work on that, Anon.  :^)
Last edited by chobitsu
Replies: >>652 >>667
>>649
Has Starship even achieved orbit?

>humanoid robot crew
That's far more realistic, and I think space exploration should follow that path. Perfect for setting up a prefab colony for eventual human colonization.

The elephant in the room is that nuclear propulsion is the only way forward. Unless someone figures out fusion, it's the only realistic option.

>Regardless, I think it's a Fool's Errand to shoot for Mars. Another Tower of Babel. Going to the Moon, however, is replete with opportunities.
Agreed. I don't understand the obsession with Mars colonies when we haven't even colonized the moon. It's like if the English settlers landed on Plymouth Rock, then ignored the East Coast and tried to colonize the Midwest.

>This! We should work on that, Anon.  :^)
VR is the way forward!
Replies: >>653 >>667
there_there_Buzz_Aldrin-everything_will_be_fine.jpg
[Hide] (171.3KB, 1024x1024)
>>652
>Has Starship even achieved orbit?
Yes, during it's testing phase(s, now).

>>humanoid robot crew
>That's far more realistic, and I think space exploration should follow that path. Perfect for setting up a prefab colony for eventual human colonization.
I like the way you think, Anon!
>Robowaifu meidos in space?  :^)
Last edited by chobitsu
Replies: >>654 >>667
>>653
>has achieved orbit
Oh nice! Last I followed, it was still self-destructing in air. Emphasis on air.

>pic
lmao
The robowaifu design reminds me of the T-X from Terminator 3

>Robowaifu meidos in space?  :^)
If we go by standard colony procedures, then absolutely not. Gotta make new colonists somehow.
If reproduction is tightly controlled, then yes. Probably also robohusbandos too.
Replies: >>655 >>656
>>654
>post about reproduction in colony
>is 69th post in thread
Nice
1737068640025269.webm
[Hide] (1.7MB, 720x1280, 00:23)
1737069763434733.webm
[Hide] (1.6MB, 720x1280, 00:15)
1737070217462933.webm
[Hide] (1.1MB, 720x1280, 00:13)
1737067816863576.gif
[Hide] (322.4KB, 73x160)
>>654
No, you're correct; it blew up (rather gorgeously over the Carribe, I may say) since achieving orbit on earlier test shots. (Thus why they are back to testing again.)
>

>space-borne human reproduction
Both Planetes, ST:TOS (et al), 2001 : Space Odyssey, Forbidden Planet, (& many others, I'm sure) broached this topic -- at least in general.
Last edited by chobitsu
Replies: >>657
>>656
Literally move fast and break shit.
Replies: >>658
>>657
Max q [1] is a beeyatch. Some DIE-tier assembly/QA on that particular part/pic -related (rocket-cam launch capp .gif above), I'd expect. We're back within """Normal Accidents""" [2] territory here again. Fair warning to all would-be robowaifuists!!  :^)

---
1.
https://en.wikipedia.org/wiki/Max_q
https://www.youtube.com/watch?v=riBaVeDTEWI&t=914s[Embed]  (so, about 15mins in)

2.  "Perrow identifies three conditions that make a system likely to be susceptible to Normal Accidents. These are:
>*  The system is complex
>*  The system is tightly coupled
>*  The system has catastrophic potential
https://en.wikipedia.org/wiki/Normal_Accidents
Last edited by chobitsu
SpaceMAid.jpg
[Hide] (593.4KB, 1920x2650)
>>649
Thanks to the sacrifice of Ingenuity, we now know that a monocopter duct taped to a Android phone is fine. Turns out, Mars isn't a particularly hostile environment for modern computers. The sand is the main problem.
>Lunar colony
Mars is unironically easier for human settlement. Lunar colonies only make sense for fully autonomous machines. You'd need to regularly exchange all people on the moon yearly at the longest. We can survive fine on Mars with a relatively simple habitat. 
>>652
>Nuclear propulsion
Based and Project Orion pilled. To think the US could have an interstellar civilization today if they weren't afraid of nuking space and tearing apart the solar system. God gave us so many gifts, yet we let them languish as lifeless rocks instead of becoming glorious scaffolding to manifest humanities destiny. If I was president, I'd run on conquering all the stars in the sky.
>>653
>Meidos in space!
Sounds like a fun time!
Replies: >>669 >>671
1734025919352646.webm
[Hide] (3.9MB, 608x1080, 01:10)
1734021626885634.jpg
[Hide] (472.6KB, 1284x717)
1734185149186905.jpg
[Hide] (92.9KB, 724x1024)
NOTE-still_inside_Earth's_magnetosphere.png
[Hide] (3MB, 1900x1045)
1739262899508957.jpg
[Hide] (55KB, 1024x449)
>>667
>Mars is easy
>Moon is hard
Well, pretty sure they're both hard! But the laws of physics say that Mars for puny mortals is Batsh*te-insane, crazy hard!111  :DD

They're both outside the Earth's magnetic-bubble lifeshield, so neither is viable longterm (until we have lightweight fusion systems invented [also a highly-dubious prognostication]) for meatspace slaves workers.

>tl;dr
Robos are much the superior option! Cheers.  :^)
Replies: >>670
MartianSolar.jpg
[Hide] (45.3KB, 771x363)
>>669
God's already blessed us with a fusion power source so vast, we can use light weight panels to harvest!
Replies: >>671 >>672
7927338558_97dfc750a6_b.jpg
[Hide] (166.6KB, 1024x685)
>>667
>Mars is unironically easier for human settlement. Lunar colonies only make sense for fully autonomous machines. You'd need to regularly exchange all people on the moon yearly at the longest. We can survive fine on Mars with a relatively simple habitat.
How though? Mars is also a desolate rock, only with slightly more gravity and a bare minimum atmosphere.

>Based and Project Orion pilled. To think the US could have an interstellar civilization today if they weren't afraid of nuking space and tearing apart the solar system. God gave us so many gifts, yet we let them languish as lifeless rocks instead of becoming glorious scaffolding to manifest humanities destiny. If I was president, I'd run on conquering all the stars in the sky.
I will NEVER forgive anti-nuclear hippies.

>>670
Solar panels are actually quite inefficient. Comparing it to fusion is like comparing a small windmill to a steam turbine.
If we still can't crack fusion, and nuclear is still off the table, they could possibly work for an ion engine on an Aldrin Cycler.
1736586746231124.png
[Hide] (343.5KB, 878x636)
>>670
Sorry nice try but that won't provide anywhere near the energy density, via anything like a reasonably-sized array structure for a transfer spacecraft's flight out to Mars -- sufficient to create a very-large, highly-effective magnetic bubble capable of protecting life (in specific, human life).

And even if/when we do have """lightweight""" fusion systems that might work, they themselves will be fraught with many health & safety issues (much like God's own exquisite fusion system in the local neighborhood you mentioned: Sol). Kind of a circular, double-jeopardy originating literally at the atomic scale.  :^)

Please don't get me wrong, I'm easily as much of a "lets-all-go-conquer-the-stars!" space buff as the next Anon, Kiwi. And I would certainly vote for you as POTUS!! on your space plank.  :D

But the simple reality is that the GH kikes have already blown off any real possibility we ever had at reaching for the stars.
>ie, continuing our initial, OG push
>

<--->

Sadly, this downfall all started well before spaceflight; even before flight, basically. The successful ((( push ))) by the demonic Globohomo to instigate women's suffrage guaranteed -- in effect -- we'd never have the free resources sufficient to get off-planet.

Equal Pay!, Mo' money fo' dem pograms!, Kill Whitey!, Migrants Welcome!, & whatnot. Without the libsh*tes & batsh*te stronk, independynts voting this political garbage into place we may have actually pulled off the drive for civilization out into the stars, in the end. Now however, nothing short-of a realworld WH40K scenario will provide the impetus necessary for us to succeed at all this, IMO.  *sigh*  :/

>tl;dr
From the Garden of Eden onwards, women have always been The Great Filter.

<--->

Well, at least we'll all have robowaifus soon enough -- one thing at a time!! Cheers.  :^)
Forward
Last edited by chobitsu
GREAT NEWS
Chips, smartphones, and integrated circuits are now exempt from Tariffs.
Replies: >>698 >>699
>>697
Wonder who bent the knee behind closed doors. Just glad we don't have to worry about that. Now, we just need sensors, motor, and batteries and we're golden.
1740820206955926.jpg
[Hide] (72.2KB, 1024x683)
>>697
>>698
PICK UP THE PHONE BRO  :^)
Replies: >>707 >>714
>>698
If I were to guess, Tim Cook (and possibly Elon Musk) sent a strongly worded letter to the White House. A lot of the press buzz was about iPhones were going to cost more, and Apple tried to fly in planes full of iPhones
Replies: >>707
>>698
I'm worried about the batteries.  The U.S. is building a few plants and Trump will definitely want to protect those.
Replies: >>702 >>707
>>701
Well if we have definite plans for battery factories, then maybe it won't be so bad. It would actually provide jobs. Plus, with the new Chinese restrictions on mineral exports, combined with newly found domestic minerals, batteries may stay the same, or even cheaper (I'm an optimist at heart).
Replies: >>703 >>707
>>702
True but that will take some time.
1744329637246149.jpg
[Hide] (59.3KB, 901x600)
>>698
>Wonder who bent the knee behind closed doors.
Would somebody really do that? Just go on the International Trading scene & kiss a*rse?  :D
Replies: >>707
>>699
Topkek :^)
>>700
They both have been against it, Tim Cook scheduled 5 flights, with 600 tonnes of iPhone's. They're already working around everything. Elon's been against it but, he wouldn't have the cards. I do see how they may have found a way together, could've bribed Xi.
>>701
China unironically makes the best batteries. It'll take time for the US to catch up in terms of cost to performance. 
>>702
Tesla could steal Chinese battery secrets, would be funny.
>>705
5 Trillion of orange arse kissing and counting is coming to the US :^)
Replies: >>714
1744501735329992.png
[Hide] (407.7KB, 879x732)
1698552166924591.jpg
[Hide] (70.4KB, 960x704)
The funny part about this is that it's losing money hand-over-fist.
>"Oi!! You got a loicense for that extra money? Have another debt burden, Citizen!"
1744533212243919.jpg
[Hide] (69.1KB, 1024x767)
>>699
>

>>707
>5 Trillion of orange arse kissing and counting is coming to the US :^)
LOL. How true.  :D
1744609932052670.png
[Hide] (168.7KB, 661x855)
BASED.

We haz winrar!! Looks like baste DeepSeek is clearly the frontrunner for the big LLM systems, insofar as what's in Anon's best-interests (ie, non-pozzed truth-telling).
>

IIRC, can't we run the DS 7B model on a RPi5 using llama.cpp ? *

---
*  update: I now think it was the DS 1.5B I was thinking of?
Last edited by chobitsu
Replies: >>747
>>746
Another W for PeakSeek

>IIRC, can't we run the 7B DS model on a RPi5 using llama.cpp ?
I'm pretty sure I saw that somewhere. You could dig in the alog archives to find it. I want to see if I can run it on my phone too.
Replies: >>754
1733434844595432.jpg
[Hide] (74.5KB, 926x926)
>>747
>I want to see if I can run it on my phone too.
If you're willing to forego privacy then you can run the full (yuge, DS 671B) model for free (& easily) on your phone. But ofc that means your based & redpilled waifu's capabilities (even her very existence) are always provisional:  she could be yoinked anytime by server updates -- the same as every other cloud-based LLM, like ChatGPT (whether paid or not).

>tl;dr
Yes, I think you can run it (DS 7B) on your phone, if you can cross-compile llama.cpp/whisper.cpp code & install it there. Very straightforward to do on an SBC; not quite so simple (but still doable) on a phone. *

Good luck, Anon!  :^)
Tay 2.0 when?

---
*  It seems likely IMO (I haven't checked) that there's already a variety of pre-built APKs out there for DS 7B - 13B today (after all, the baste Chinese released it back on Jan 22nd).
Last edited by chobitsu
[New Reply]
92 replies | 61 files
Connecting...
Show Post Actions

Actions:

Captcha:

- news - rules - faq -
jschan 1.6.2