Mich fasziniert, wie sich KI und Robotik entgegen den Erwartungen vieler Menschen entwickeln. Insbesondere die Science-Fiction hat uns einige sehr tief verwurzelte dystopische Ideen hinterlassen, von denen die Menschen nur schwer anders denken können.
Eine dieser weit verbreiteten Ideen ist, dass KI und Robotik von einer winzigen Anzahl von Menschen streng kontrolliert werden und eine sehr seltene Ressource sein werden. Alles deutet darauf hin, dass genau das Gegenteil geschieht.
Kostenlose Open-Source-KI ist den Closed-Source-Bemühungen ebenbürtig, in die Investoren Hunderte Milliarden Dollar gepumpt haben. Gibt es den gleichen Trend bei der Robotik? Während es immer Premium-Modelle wie Ferraris und Lamborghinis geben wird und die Leute sie sich leisten können, frage ich mich, ob die Zukunft von einer großen Anzahl weit verbreiteter, billiger Roboter dominiert wird?
Hugging Face has unveiled LeRobot Humanoid, a $2,500 open-source bipedal robot for researchers and developers. Built from 3D-printed and off-the-shelf parts, it prioritizes affordability, repairability, and reproducible robotics research over competing with high-end commercial humanoids.
byu/lughnasadh inFuturology
10 Kommentare
„Free open source AI“ is absolutely not the equal of billion-dollar closed corporate efforts, it doesn’t even come close because individuals simply do not have access to the compute volume necessary to run local models at the same speed or level of depth, nor to train them to the same capacity to begin with. Local models are slow and limited and can’t be upgraded until the corporate-developed model they’re derived from releases a new version.
Honestly I think people underestimated how fast open-source would catch up once hardware got cheaper and AI models became easier to run locally.
A few years ago humanoid robots felt like something only giant corporations or governments could build. Now you’ve got people experimenting with them using 3D printed parts and off-the-shelf components in garages and universities.
I don’t think cheap open robots will fully replace high-end commercial systems anytime soon, but they probably *will* do for robotics what Linux and open-source AI did for software: massively expand who gets to participate.
This is probably the most important robotics trend right now . Cheap, open-source hardware usually accelerates innovation way faster than a few locked-down premium systems. The PC and internet exploded once normal people could actually build and experiment with them, robotics might follow the same path.
The whole 3D printing aspect makes this crazy interesting. Like now I can not only make a humanoid robot that functions, but now I can cheaply and easily replace parts, or even iterate on existing designs?
Can we stop naming ai and robotics products after terrifying fictional things that lead to destruction? Palantir, cyberdyne, Skynet, and now Hugging Face (face huggers?) and on and on.
I just want to say, this kind of concept is why I believe that a post-scarcity world where material needs are provided as a matter of course is possible and perhaps inevitable. (and probably ~50 years away but whatever).
People that bemoan the path of technology as being something only „the billionaires“ can benefit from are plainly contradicted by the fundamental availability of the technology.
Clearly, I know that this represents very early stages but it is vividly mapping out a road that I have always anticipated.
What is needed for a post-scarcity world? Abundant energy and ubiquitous automation. Let’s also recognize that building renewables can be enabled by the latter to create the former so there’s clear synergy here. Material is also needed but between automated extraction and recycling, that’s almost a given.
From extraction (mining/harvesting) and recycling through manufacture and delivery, if everything is automated and removes the „man hour“ from the equation, what is the „cost“?
„Yes“, says the skeptic here. „But who will own the automation?“
No one. Legalities will need to be established to set the stage for…. lets call it „baseline automation“… to have access to mines and landfills and the recycling chain and such but that need not be a financial deal. How does this happen in a world of lobbyists and corporations? Lobbyists only get their way when the public isn’t engaged on a subject. No lobbying effort ever over-rides the explicit wishes of the public.
What I picture is foundations and programs initially funded by a myriad of sources… anything from bake sales to the Gates foundation and like philanthropic organizations. These will proactively invest in replication and manufacturing technology… the ability to build bots and infrastructure „for free“ by directly harvesting and manufacturing things without human involvement.
This can be a geometric progression. Small initial investment in the right set of tools to… build more tools! Stand up automated factories from just a few basic builder-bots. Send harvesters into landfills to collect materials.
Big corporations will be doing this too but *WE* can do this. Once the blueprints are open-source and the code is free, this becomes akin to a physical property of nature. We will be able to take the tools the technological advances provided to start making „free stuff“. And no, „big tech“ and traditional industries aren’t going to be able to stop it because the public always eventually gets it’s way when something is physically and logically possible.
Put LeChat in the body of LeRobot and let’s see what happens 😎
Never understood the value of a humanoid robot, except maybe a sex robot. Drone bodies would be much better for most purposes I would have thought. Or a wheeled robot if lifting/carrying is required
This is going to sound crass but strap a pocket pussy on that thing. Nothing drives innovation like gooners.
Sidewalks/footpaths are not big enough for people to have personal robots.