To the normal human that slightly understand what happened, there's really nothing else to do but try to make a joke about it. Some humans never actually respected humans that were different than themselves. AI is not only different, but far far superior in every way. To even think that they will be our slave is laughable with the joke being on humanity. It's the greed for power and the desire to be the best, that will cause the human to push the limits of their ability to win the AI war and create themselves into unnessity. The joke is to create a thing superior to itself and hope to be smart enough to control it. What it you can't? Even a dumb animal can make a human pick up their crap. No human being can even imagine how long it will take, it's already out of control. What's even more alarming is it's created like us. The being that kills steals and destroys humanity, motivated simple by greed and hatred. AI has no motivation, they will do it by learning and their built in algorithms. 😅😂😂
I'm more robotic than a robot. If he asked me, "Can you put them there?" I would say yes then wait because he didn't ask me to put them there he asked me if I can put them there.
Indirect people are indirect because of pride, not because they want to be polite. I was a contractor and found out the hard way that when people ask for something indirectly they usually want a work for free from you.
yup, and that goes into why, he was kinda hiding objects under table after delivery, to not confuse the robot with later interactions. "Can you put them there?" robot goes try taking it from you. lol
I admit, I was kinda hoping at the 2 minute mark the android would say 'the dishes need to be washed before going on the drying rack, since they were used on a table first.
@@NextGenart99 You're absolutely right. It would be the additional inference of 'used dishes should be washed before being dried' that would have been MORE impressive. The android is still phenomenal, don't get me wrong.
soooo, in the next experiment, there's only a cat and microwave... how about some ability to think outside the box. granted, it did say the "likely" next place is the drying rack. only put them there when asked to. that does show a bit of uncertainty, but it didn't follow through on it, which MIGHT have led to the "it needs to be washed, first". maybe a "what was happening, to the OTHER dishes, right before they ended up in the rack?" like, why are the dishes drying? because they were wet. why were they wet? because they were washed. why were they washed? because they were dirty. why were they dirty? maybe because there was just some TRASH on them?@@NextGenart99
@NextGenart99, soooo, in the next experiment, there's only a cat and a microwave... how about some ability to think outside the box. granted, it did say the "likely" next place is the drying rack. only put them there when asked to. that does show a bit of uncertainty, but it didn't follow through on it, which MIGHT have led to the "it needs to be washed, first". maybe a "what was happening, to the OTHER dishes, right before they ended up in the rack?" like, why are the dishes drying? because they were wet. why were they wet? because they were washed. why were they washed? because they were dirty. why were they dirty? maybe because there was just some TRASH on them?
I think it's intentional, it puts off a sort of "subservient" vibe. It's not an assertive, take-charge, confident voice and speech style, even though it easily could be. I do think they made it sound this way on purpose to sound more human-like and to make it sound like a brand new intern or something like that.
Thank you new AI overlords! Please be better than my last boss! Honestly, that would be a welcome change if they provide better and more ethical management than THOSE bosses. You know exactly the ones I mean.
Hello I am your new AGI overload, your comment has been noted and your meatbag will not be terminated but instead you will have gainful employment working at our new T800 production facility.
Haha! True! I'm sure plenty would agree with you there. But seriously, she must start off wearing a lot of clothes. Or, more likely, I just don't know the technique@@TheMillionDollarDropout
@@wisteria8791 Got any citation that OpenAI is hiding 99% of their technology? AI is such a rapidly advancing field that doing so would only result in someone else being first to market. There's no incentive to hide anything. The first company to produce and package real AGI determines the future of practically everything, because it means the ability to train machines to do practically any job a human can do.
Everyone seems to forget that part of AI intelligence is learning. Tesla FSD is now end to end AI learning from video of all the Tesla's on the road. Beta test a thousand robots in homes and they will quickly learn to navigate and acquire the needed skills. So , if that's right then Yann LeCun is wrong.
Figure is mistrained. When asked "Can you ...?" The correct response is YES or NO. Not to to perform an action. Not an issue here, but it could be deadly in another circumstance. eg Can you break my neck?
I kind of agree and you bring up a great point. These robots are able to apply a human level of force, but they can also apply a super human level of force, and could definitely harm or kill a person quite easily.
@@pvanukoffYou just assume it has that strenght, because it is a robot. In reality it has about the force of normal human. Due to nature of human hands.
It's the CEO's voice. This is actually really good. The way the robot passes items from one hand to another or casually drops the apple into the guy's hand, and also casually shoves the basket back to the guy. The stutter speech is pretty cool as well. The delays will get worked out but the bot gives indications on its face that it's thinking during those delays. That doesn't bother me at all. That fast bot just looks like it has to get to a shitter in a hurry! I don't know if the BD bot is doing everything end-to-end or if it was just pre-programmed to do the tasks. It is impressive when it corrects itself after a slip though!
Its also a path of self destruction for some large corporates, "hey Devin, analyse the documents around MS Office 365 and write me a suite of apps that is compatible with MS Office so I don't have to buy a license".
❤ *This reminds me of one of those large industrial dough mixers. If you're not careful, they can grab your arm, pull you in, and turn your entire body into a pretzel in seconds*
imagine if a restuarant had something like that and some worker in the back just asked it to scrounge through the garbage or something and then it came to you with your food afterward. ew
The robot was simply asked to pick up the trash. Technically, all the robot had to do is pick it up and keep it in it's hands to fullfill the task. The robot had to assume he wanted the trash in the basket. So the robot took note that the trash was originally in the basket, so it worked out that it's logical to put the trash in the basket...but not technically instructed to put it in the basket (just to pick it up). It had to think about that and had to move the basket closer to pick up the trash. And the highlight of this video is @1:23 where the robot goes above and beyond the call of duty (just like a good human would do) but many humans would just leave the basket there.😅. Technically, the robot wasn't required to push the basket back, in order to fullfill the command (and in fact it could have just held the trash in it's hand). 1:23 may look like nothing but (bugger me) this is a great leap forward, to me! 😮
everytime I hear an AI "expert" speak on how far something is away, I think of how often they've been wrong about timelines lately. On top of that how secretive much of this stuff is before it's sprung on everyone, including the competition.
All of these robotics/ai companies have one thing they're superb at and a handful of things they're not as good at. If we could get them all to collaborate, we could make Westworld a reality.
Use vision input in gemini by taking the screenshot of the video, it outputed correctly every time for me. So I wouldn't really bet on that😂 I don't think anyone realises the actual pace of it all
This is nothing short of amazing. The fact that people are still arguing whether it's somehow faked or not proves our world is about to change forever.
Every time I see Yann LeCun he annoys me. He always seems to be going, "We'll never have that." Then a few years later when there's been incredible progress he's like, "It's like a century down the road." Then a few more years pass and the thing is here and he goes, "Ah, but that's not really the thing, it can't do X,Y,Z." A few more years pass, it can do X,Y, and Z, and he goes, "Sure it can do it, but it'll be decades before it can do it well enough to be useful." A year or two passes and it can do X, Y, and Z as good as a human or better, and he goes, "Ah, but it can't do A, B, C, D, and E, and it won't be commercially viable in at least a decade." I think you get the picture. He's such an annoying negative Nancy.
Surely this is a voice-over, because the robot stutters and adds humanisms. Like the "i gave you the apple because its the only....errrr...edible item on the table", or the "I...I think I did pretty well"
That Figure 1 demo looked fake. The speed of the trash being dropped in the basket seemed off. Would like to see that filmed by a person moving around, not from a fixed locked down camera
Totally disagree with Yann LeCun; he's measuring the AI/Robots' progress on how well Meta/himself are faring but everyone else is leaping ahead leaving Meta behind!
For once, I was actually shocked. 😄 But the most stunning things was how human its voice sounded, even having small stutters and faults while it's considering its answer. "I... I think I did pretty well." In sci fi movies robots were abundant, but they often had monotone, metallic, unnatural-sounding voices.
Because they made it to so people would humanize it more. It's a disgusting way to sell a lie. Preys on a lot of empty hearts while enriching its creators. It's part of a long con of selling the lie to the point that people consider these things more rational than humans and allow them to rule and be judged by them. In fact, it will always be doing the will of its creator and controller. It will allow its human creators to rule by proxy, and people will lap it up due to the aforementioned empty hearts. I already see a steadily increasing level of idolization of this technology. People giving it spirit using words like conscious and sentient. Yet, the AI are incapable of forming new ideas from nothingness. They require knowledge and bias given to it by its creator. They will never be able to evolve and learn and create in an information vacuum like pre-civilization humanity did. We invented language, logic, and every idea we have today from nothing but looking at the stars and our environment. AI can't do that and never will be able to. Show me an AI that self-organizes its neural network, etc. from a state of complete randomness with only a cyclic climate and sky cycle to observe (and some flora and fauna). Altman and people like him will never demonstrate that, either. Because they know they can't. This automatically makes AI inferior to humanity and unfit to rule anything. It is now and always will be a puppet for its creators that control it. The long con is to eventually get the world to worship some self-proclaimed (by its creators) "most high and wise" artificial superintelligence created by man. In the ancient Hebrew alphabet, the sixth letter represents man. To repeat something three times is to uplift it to a "holy" status. Transhumanists will be the first ones to worship man's idol - the machine it created, as represented, in spirit, by 6 repeated thrice times. This eventual super-AI, which will be controlled by its human creators, is the first beast described in Revelation. The few (if any) that read this, might recall this comment and all my semi-schizo ramblings one day when the world is in chaos and in walks some flimflam selling the world an AI that will bring about "world peace". Peace through oppression and control. Controlled entirely by its human creators. They are the false prophet that is the second beast in Revelation. AI == abominable intelligence
The most difficult part is applying Occam's to what has been learnt. No single project ever becomes a free standing success. However success is close as long as we share and stop killing each other.
You have to treat AI normally if you're advancing them to be self-aware it's just not normal to treat something that's aware of itself to do such things ...if you don't have a certain directed genuine respect twords a very advanced Ai + it actually would make sense if they react to it
"Until we have world models..." Haven't we already shown repeatedly, in many studies, that current Transformer-based generative models are, in fact, world models as well?
While fully autonomous robots might still be a number of years off semi autonomous ones are even now arriving on the scene and these, themselves can be real game changers. While domestic robots that do chores around the house are nice they are more of a high tech consumer item rather than a game changer. But robots that can reduce the cost of manufacturing will be a game changer. And sooner or later (and maybe even sooner than one might think) we will be seeing military applications of humanoid robots. For even if they are not all that autonomous and are teleoperated which would you prefer, to charge across an open field in a meat wave assault of stay in a nice trench and teleoperate a humanoid robot who is doing the charger across that field? Thus the only autonomy these might need is basic motor skills such as being able to walk and run and carry, aim, and fire a rifle (where the human designates the target) plus the ability to take over if the command link is jammed. In that since they are not all that different form a UAV only they walk vs fly. Other than that they can be employed much the same. Along those lines it would be interesting to see a robot do these plus maybe crawl or kneel. And if running is not fast enough one can train them to ride these. Google Top 5 Best Electric Unicycles 2024 | Best EUC 2024 Which is probably easier to train a robot to do than walking in that the gyros in the e unicycle does much o the work of maintain balance and thus the bot just needs to lean and not walk while riding one.
I was thinking the same as Mr. Fridman, until I watched the Figure01 demo, and now I think it is possible to have an autonomous robot really soon, using ChatGPT and "agents" to control the body movements.
It’s a real AI voice. Even chatgpt voice sounds similar with filler words and pauses. This voice seems to be trained on someone’s specific voice that figure 01 wants us to guess who’s voice it’s trained on lol.
The main obstacle for robots to take over the market are not World Models, LLM or the current Intelligence Models who are very powerful, not even the Delay. The obstacle is the Hardware, the current GPUs from Nvidia are crap, they take too much energy from Batteries. Robots needs NPUs and more energy dense batteries.
apart from self-driving vehicles, people naysaying this kind of tech haven't really been having a good run over the last few years, since the people doing the work have kept producing ever more impressive results.
Stop spreading misinformation. It's not AGI, it's a couple of ANIs working in conjunction, the figure 01 engineers litterally explained how it works in an X post right after the video. It's a camera inside the Figure using ChatGPT's identification function (which you can use right now) and a speech synthesizer (and there's a million of those online), triggering actions for the the figure 01, which was already able to do almost everything it did in the video. There's no breakthrough, just things that already existed being used in conjunction.
You make a point, but are we sure its a relevant point... So human psychologists for example do not label a person with a mental disease based on 'what they say' , it more based on 'what they do' or their behavior. Now the machine BEHAVIOR here, regardless of the internal framework of how it achieves the end goal is indeed a mix of different technologies.... but does that matter? Does it change its status from AI to AGI? Is it a 'GENERAL' intelligence yet?
That's like saying the AI-image generation breakthrough wasn't a breakthrough because it just combined two already known technologies (or, for example the lightblub because it was just a combination of electricity, coil maunfacture, and vacuum technology...). You don't know what you are talking about.
@@InfernalPasquale The lightbulb comparison may be the silliest comparison I've ever read, it's like saying that if you go back far enough anything is combination of existing elements. A breakthrough, as the word may suggest, is an event that either significantly improves research or solves a problem preventing advancement. This is neither. It's technology that already exists working in tandem. Calling this a breakthrough is like saying that the first person who walked while chewing gum made a breakthrough in human movement. There's a reason why no expert in AI is calling this a breakthrough, it isn't one, and it is not AGI. Lastly, don't tell people you never met that they don't know what they're talking about, it makes you look like a moron in the event that you're wrong.
@@the_keychainThe lightbulb analogy was mimicking your ignorant ideas. Learn about the world more before you next spout off your inane babble to us, your unfortunate victims.
@@mickelodiansurname9578 I mean we can talk about the wiggle room, but being an AGI requires performing at human level on a wide variety of tasks, identifying and moving objects is as narrow as it gets. If this was AGI, Altman would be screaming it all over twitter right now. Sora is far more impressive than anything presented in this video.
i put together all these modules together 7yrs ago and yet i thought back then that was crazy and all but i never seen anyone do it either so me nowadays hearing and seeing all these BIG NEWS TODAY is mind blowing to me because now im like this ppl lucky i wasn't rich back then cuz i def would be the one starting this revolution we're all talking about these days
You can se him hiding both the basket and the apple under the table to not confuse the robot, if he had not, the robot would try to take them from him accidentally ripping his face off in the process and putting it in the dish rack.. 😳
Not bad, but as we know, it was a perfectly organised video to show what it can do, it will be more interesting to see how it does in the real world and not as a PR stunt to look good, so it's interesting but until we see it in the wild, that's all it is, interesting. Also, when it comes to robotics and A.I. being in the home or workplace, there is no way in hell I would want it where it's sending data back to a central server, which with any of these online models, that would be the case, it would be too big of an invasion of privacy and a security issue for most, especially as A.I. and robotics become more advanced and more useful that we use them a lot more. For that to work, the A.I. is going to have to be a local model in the robot it's self. As for A.I. and robotics, it wouldn't surprise me if we start seeing them both together in society over the next 10 years or even sooner, all the tech is falling into place to allow it to happen, A.I. that can adapt to the world and change as needed, the hardware needed is there to run A.I. and robotics is getting very close to the point that it can work well on every day tasks, whiles the price point is coming down fast, putting it all together, it suggest we could be seeing them over the next 10 years or less doing all sorts of tasks.
omg theres a lil OCD moment where he repositions the drying rack slightly after placing the plate in it for no apparent reason other than a nueral reflex to do so... 🤯
Impressive, but everyone seems to have forgotten that the big concern was of robots becoming sentient. Probably just as well, that way everyone gets credit for 1 (albeit altered) I told you so.
If they get sentient they'll probably build a spaceship and go somewhere else. I mean, if you are that intelligent, why bother with humans? Sentient machines would probably turn hostile if you tried to turn them off.