While there is a lot of work to do, Tedrake says all of the evidence so far suggests that the approaches used to LLMs also work for robots. “I think it’s changing everything,” he says.
Gauging progress in robotics has become more challenging of late, of course, with videoclips showing commercial humanoids performing complex chores, like loading refrigerators or taking out the trash with seeming ease. YouTube clips can be deceptive, though, and humanoid robots tend to be either teleoperated, carefully programmed in advance, or trained to do a single task in very controlled conditions.
The new Atlas work is a big sign that robots are starting to experience the kind of equivalent advances in robotics that eventually led to the general language models that gave us ChatGPT in the field of generative AI. Eventually, such progress could give us robots that are able to operate in a wide range of messy environments with ease and are able to rapidly learn new skills—from welding pipes to making espressos—without extensive retraining.
“It’s definitely a step forward,” says Ken Goldberg, a roboticist at UC Berkeley who receives some funding from TRI but was not involved with the Atlas work. “The coordination of legs and arms is a big deal.”
Goldberg says, however, that the idea of emergent robot behavior should be treated carefully. Just as the surprising abilities of large language models can sometimes be traced to examples included in their training data, he says that robots may demonstrate skills that seem more novel than they really are. He adds that it is helpful to know details about how often a robot succeeds and in what ways it fails during experiments. TRI has previously been transparent with the work it’s done on LBMs and may well release more data on the new model.
Whether simple scaling up the data used to train robot models will unlock ever-more emergent behavior remains an open question. At a debate held in May at the International Conference on Robotics and Automation in Atlanta, Goldberg and others cautioned that engineering methods will also play an important role going forward.
Tedrake, for one, is convinced that robotics is nearing an inflection point—one that will enable more real-world use of humanoids and other robots. “I think we need to put these robots out of the world and start doing real work,” he says.
What do you think of Atlas’ new skills? And do you think that we are headed for a ChatGPT-style breakthrough in robotics? Let me know your thoughts on ailab@wired.com.
This is an edition of Will Knight’s AI Lab newsletter. Read previous newsletters here.
Source link
#Robot #Single #Model #Master #Humanlike #Movements
![John Grisham’s New Legal Drama Is a Real Life Fight Against AI Audiobooks on YouTube
There’s an argument to be made that audiobooks are the finest form of content. You take a book—already off to a good start—and you get to have someone read it right into your ears. And when I say “someone” I mean the GOATs in the voice game. I could cite examples of celebrities you never knew narrated audiobooks, but here’s a sample of Werner Herzog narrating his memoir Every Man for Himself and God Against All that I think speaks for itself: [embed]https://www.youtube.com/watch?v=P4IQSvi3pXU[/embed] What could be better than this? Not only are audiobooks heaven, you can probably get all the audiobooks you want for free (and legally) by getting yourself a library card and using your local library’s preferred app (Libby, perhaps). I say all that, because given all the easy and free access to high quality audiobooks, why in the world would anyone listen to a John Grisham audiobook presented like this?
Don’t click that link. Instead of the actual audiobook, which is read wonderfully by Michael Beck, it will take you to a YouTube video consisting of an AI narrator reading Grisham’s recent hit novel the Widow, and the narration plays under 13 hours of AI slop video—simulated stock footage of fake vacations, basically. It looks like the video they display under the lyrics on Hell’s karaoke machine. I don’t have any science to back this up, but it will definitely give you brain cancer.
As the New York Times points out, 80,000 lost souls listened to the Widow this way. And Grisham is pissed about it. “The thieves and pirates who steal my work and try to profit from it, in any format, should be punished civilly and criminally […] And in this particular example, YouTube is complicit because it’s clear they know what is happening and refuse to stop it,” Grisham told the Times in an email. He should really write about this. YouTube, for its part, says the video is still up because there hasn’t been a takedown request, and that it doesn’t proactively police for copyright violations. “For more than two decades, we’ve built systems that help rights holders manage and control their copyrighted content — investing continuously to make sure those systems evolve as new threats emerge,” Jack Malon, a YouTube spokesperson, wrote to the Times.
If you’ve ever had a YouTube video flagged for a copyright violation, it may have been because of a feature called Content ID that music publishers absolutely love. It allows copyright holders to crawl YouTube and automatically detect copyrighted content. At times, Content ID has been a valuable moneymaking scheme for copyright holders, who were able to zero in on incidental—or even accidental—uses of copyrighted material, especially music, and by making a claim, monetize other people’s videos. It can’t do this anymore, but this is the sort of thing YouTube’s copyright system has been designed to support. As the Times points out, Content ID isn’t great at finding AI-narrated audiobooks. The audio waveform of the content is not the same as the audio the publisher owns, which makes it tricky to know what to even scan for. The author holds a copyright on the text, which can be slightly changed by the creator of the YouTube video while still leaving the book largely intact—good enough for casual listeners anyway. This leaves publishers and authors to navigate the takedown process manually, which seems, judging from the fact that the Widow is still up, to just not be happening.
That’s a pity. And I don’t mean because it’s robbing John Grisham of audiobook sales, which is bad, but not the gravest injustice in the universe. It’s bad because people are listening to such horrible garbage just because it’s available. And they really, truly, don’t have to. #John #Grishams #Legal #Drama #Real #Life #Fight #Audiobooks #YouTubeArtificial intelligence,Audiobooks,Books,intellectual proper John Grisham’s New Legal Drama Is a Real Life Fight Against AI Audiobooks on YouTube
There’s an argument to be made that audiobooks are the finest form of content. You take a book—already off to a good start—and you get to have someone read it right into your ears. And when I say “someone” I mean the GOATs in the voice game. I could cite examples of celebrities you never knew narrated audiobooks, but here’s a sample of Werner Herzog narrating his memoir Every Man for Himself and God Against All that I think speaks for itself: [embed]https://www.youtube.com/watch?v=P4IQSvi3pXU[/embed] What could be better than this? Not only are audiobooks heaven, you can probably get all the audiobooks you want for free (and legally) by getting yourself a library card and using your local library’s preferred app (Libby, perhaps). I say all that, because given all the easy and free access to high quality audiobooks, why in the world would anyone listen to a John Grisham audiobook presented like this?
Don’t click that link. Instead of the actual audiobook, which is read wonderfully by Michael Beck, it will take you to a YouTube video consisting of an AI narrator reading Grisham’s recent hit novel the Widow, and the narration plays under 13 hours of AI slop video—simulated stock footage of fake vacations, basically. It looks like the video they display under the lyrics on Hell’s karaoke machine. I don’t have any science to back this up, but it will definitely give you brain cancer.
As the New York Times points out, 80,000 lost souls listened to the Widow this way. And Grisham is pissed about it. “The thieves and pirates who steal my work and try to profit from it, in any format, should be punished civilly and criminally […] And in this particular example, YouTube is complicit because it’s clear they know what is happening and refuse to stop it,” Grisham told the Times in an email. He should really write about this. YouTube, for its part, says the video is still up because there hasn’t been a takedown request, and that it doesn’t proactively police for copyright violations. “For more than two decades, we’ve built systems that help rights holders manage and control their copyrighted content — investing continuously to make sure those systems evolve as new threats emerge,” Jack Malon, a YouTube spokesperson, wrote to the Times.
If you’ve ever had a YouTube video flagged for a copyright violation, it may have been because of a feature called Content ID that music publishers absolutely love. It allows copyright holders to crawl YouTube and automatically detect copyrighted content. At times, Content ID has been a valuable moneymaking scheme for copyright holders, who were able to zero in on incidental—or even accidental—uses of copyrighted material, especially music, and by making a claim, monetize other people’s videos. It can’t do this anymore, but this is the sort of thing YouTube’s copyright system has been designed to support. As the Times points out, Content ID isn’t great at finding AI-narrated audiobooks. The audio waveform of the content is not the same as the audio the publisher owns, which makes it tricky to know what to even scan for. The author holds a copyright on the text, which can be slightly changed by the creator of the YouTube video while still leaving the book largely intact—good enough for casual listeners anyway. This leaves publishers and authors to navigate the takedown process manually, which seems, judging from the fact that the Widow is still up, to just not be happening.
That’s a pity. And I don’t mean because it’s robbing John Grisham of audiobook sales, which is bad, but not the gravest injustice in the universe. It’s bad because people are listening to such horrible garbage just because it’s available. And they really, truly, don’t have to. #John #Grishams #Legal #Drama #Real #Life #Fight #Audiobooks #YouTubeArtificial intelligence,Audiobooks,Books,intellectual proper](https://gizmodo.com/app/uploads/2026/05/john-grisham-1280x853.jpg)

Post Comment