Glitching robots are funny but AI experts say they signal Terminator-level hazard: Early warning signs | Latest Tech News

They might appear to be scenes out of a slapstick comedy, but viral robot mishaps are no laughing matter — experts even warn they might foreshadow a “Terminator”-level apocalypse.

A humanoid dance bot was lately performing for patrons at the Haidilao hotpot restaurant in San Jose, California, only to end up actually tearing up the dance ground, knocking over tableware, smashing plates, and sending chopsticks flying like Mr. Magoo’s machine doppelgänger.

The spectacle ended with human staffers dragging the flailing droid out the door as bemused clients seemed on.

While the malfunction elicited guffaws online, techsperts warn that these cybernetic pratfalls might signal one thing more alarming as bots develop into more and more embedded in on a regular basis life, as these consumer-grade machines can come aside when the rubber hits the highway.

A dance bot goes haywire at the Haidilao hotpot restaurant in San Jose, California. Xiao Hong Shu

“I think these incidents are often treated as funny only because the immediate harm was limited and the context was theatrical,” Dr. Roman Yampolskiy, a tenured affiliate professor and laptop scientist at the University of Louisville, told The Post. “People laugh at low-stakes failure.”

But, the AI specialist added, “From a safety perspective, they should also be taken seriously, because they reveal something important: systems that appear polished and entertaining can still behave unpredictably in the physical world.”

In the past couple of months alone, a handler in China was kicked in the groin by an superior Unitree robot he was controlling, and a droid shockingly slapped a youngster during a dance demo gone awry.

A person will get kicked in the groin by a Unitree humanoid robot during a demo gone awry. X/wmorrill3

What if, Yampolskiy inquired, related malfunctions occurred around a child, a hospital affected person, or a member of the public during a police interplay?”

“The event would be viewed not as comic relief but as a dangerous systems failure,” said the researcher, who has authored over 100 papers on the existential risk of AI. “A glitch in a dancing robot is mostly embarrassing. A glitch in a security robot, delivery system, self-driving platform, medical assistant or industrial machine can injure people, damage property or trigger cascading failures.”

While seemingly far-fetched, Yampolskiy said that these minor hiccups scale into bigger issues as the automation offensive makes AI more ubiquitous in sectors like security, healthcare and even romance.

More than 60 bomb squads across the US and Canada are already utilizing Spot, Boston Dynamics’ hyper-advanced, 75-pound robo-dog for roles ranging from armed standoffs to hostage rescues, Bloomberg reported.

Multiple companies are also working on developing hyperrealistic helper bots for in-home use, such as Clone Robotics’ “Protoclone” — touted as the “world’s first bipedal, musculoskeletal android” — which might allegedly stroll, speak and full chores.

“As AI moves from screens into bodies and institutions, the cost of error rises dramatically,” Yampolskiy declared.

The cutting-edge automatons are changing into stronger and quicker, too.

Researchers in South Korea have developed a chemical construction for an synthetic muscle that can doubtlessly enable humanoid robots to carry 4,000 occasions their weight, while China’s Bolt humanoid robot can run up to 22 m.p.h.

A Unitree robot by chance slaps a younger boy in China during a botched dance demonstration. JamPress

The bots can kick our butt as effectively.

Viewers had been understandably involved over a video of Unitree’s next-gen humanoid robot, the H2 — touted as having both business and personal use — that was noticed lifting a smaller droid off the ground with a knee strike, sending its breastplate flying.

What would occur if a human had been on the receiving end of such a hit? Such mishaps aren’t just theoretical. In February, a Unitree G1 robot by chance struck a man in the nostril, inflicting him to bleed, while it was making an attempt to proper itself after falling during a efficiency in China.

“With existing reinforcement learning policies, their robot is trained to do whatever it takes to stand up after a fall,” Eren Chen, who claims to work for the robotics firm Booster Robotics, wrote on X. “During that recovery attempt, it kicked somebody in the nostril, inflicting heavy bleeding and a attainable fracture.

“This should be treated as a high-priority safety issue for Unitree to fix.”

A robot is escorted away by cops in China after seemingly stalking a girl. @CyberRobooo/X

However, Yampolskiy identified, “human safeguards can reduce risk, but they do not eliminate it.”

“Better testing, physical constraints, geofencing, kill switches, supervision, and strict deployment standards all help,” he declared. “But no complex system is perfectly reliable, especially when it operates in open-ended real environments.”

It comes with the technological territory.

“So, yes, some accidents are to be expected, just as with cars or other powerful tools,” he said. “The difference is that society must decide what level of failure is acceptable, and that threshold should be very low for systems operating around people.”

These humanoid battle bots compete in a robot combat membership recognized as Rek. X/REK

Unfortunately, tech bigwigs haven’t always been forthcoming about potential pitfalls — or at least haven’t used them as signs they ought to slow their roll.

Robert Gruendel, a former engineer for human robotics firm Figure AI, sued the company because, he claimed, they fired him for warning that their robots “were powerful enough to fracture a human skull,” CNBC reported.

However, Figure has denied the allegations, with reps saying that Gruendel was “terminated for poor performance,” and that his “allegations are false.”

Who is culpable ought to one of these cybernetic Frankensteins go haywire?

Yampolskiy believes that the “primary responsibility lies with the companies that design, deploy and profit from these systems.”

“Developers, operators and users may each bear some share, depending on the facts, but the burden should fall most heavily on those who choose to release insufficiently reliable systems into real-world settings,” he said. “More broadly, these episodes are early warning signs.

“Today, they go viral as odd or amusing clips. Tomorrow, with more capable and more widely deployed systems, the same class of failure may be discussed in terms of injury, liability, and public safety.”

Stay informed with the latest in tech! Our web site is your trusted source for breakthroughs in artificial intelligence, gadget launches, software program updates, cybersecurity, and digital innovation.

For recent insights, knowledgeable coverage, and trending tech updates, go to us often by clicking right here.

Student accused of trying to murder Sam Altman AI doomer | Latest Tech News

In January, podcaster Andy Mills interviewed an AI doomer who had advocated on Discord for killing tech execs. Still, when news broke last week that a 20-year-old had been arrested for making an attempt to murder OpenAI CEO Sam Altman, Mills was shocked.

“When I saw that they had released the name of this guy,” Mills told The Post, “I was like, ‘Holy s–t. It’s Dan.’”

On Monday, Daniel Moreno-Gama was arrested for throwing a Molotov cocktail at Altman’s San Francisco home on April 10, then making an attempt to burn down OpenAI’s headquarters some 4 miles away. Investigators allege he was carrying an anti-AI manifesto that read, “If I am going to advocate for others to kill and commit crimes, then I must lead by example … .” The DOJ has charged him with tried murder and arson.

Daniel Moreno-Gama is accused of tried murder and arson in assaults on Sam Altman and OpenAI’s San Francisco headquarters.

The interview, “Sam Altman’s Attacker, In His Own Words,” debuted Thursday.

Mills, host of the podcast “The Last Invention,” which explores different faculties of thought about artificial intelligence, discovered Moreno-Gama on a Discord channel, Pause AI, devoted to speaking about the hazards of AI.

Hiding behind the username Butlerian Jihadist — the title of a novel in the “Dune” collection by Brian Herbert and Kevin J. Anderson — the Spring, Texas, faculty scholar was anonymously flirting with utilizing violence against tech executives.

“Will speaking about violence get me banned?” he requested moderators.

“[I] reached out and said, ‘Hey, man, what did you have in mind when you talk about violence?’” Mills told The Post. “And he said, ‘How about Luigi-ing some tech CEOs?’” — a reference to Luigi Mangione, who is accused of murdering United Healthcare CEO Brian Thompson in December 2024.

During the interview, when the host requested if the Lone Star College scholar actually thought violence against AI executives was a good concept, Moreno-Gama softened a bit.

Moreno-Gama beforehand made violent feedback about tech execs on Discord. Department of Justice

“I didn’t really mean that as a threat or anything,” Moreno-Gama said. “I think before we even think about violence, we need to exhaust all our peaceful means first. I think protesting, I think sharing information — I think that needs to come way before we even consider [violence].”

Mills pressed: “Do you think that if we continue to see the industry move in the direction it’s moving now, that by whatever means necessary, we have to stop the extinction of the human race?”

Moreno-Gama paused for a number of seconds before replying, “I’ll say no comment.”

“He seemed earnest and intelligent, and very informed,” Mills recalled. “He was incredibly well informed on the AI doomer position.”

Sam Altman’s home was pelted with a Molotov cocktail in an obvious assault. AP

In their dialog, Moreno-Gama distanced himself from calls to violence and instead advocated for insurance policies that would regulate the scale and quantity of data facilities.

“I don’t believe I’m a violent person,” Moreno-Gama said. “I would normally only advocate for violence as the absolute final, like, I don’t want to say final solution, but you know, the final … final … OK, you get what I’m saying.”

Moreno-Gama recalled first encountering AI as a high-schooler.

“When ChatGPT came out, at first I thought it was the greatest thing on Earth. I thought, this is awesome. I get to basically cheat on everything,” he said. “I wasn’t thinking about the repercussions that might have on learning at the time. I was just a sophomore [in high school].”

Investigators say that when Daniel Moreno-Gama was arrested for the tried murder of Sam Altman (above), he was carrying a manifesto that read: “If I am going to advocate for others to kill and commit crimes, then I must lead by example … .” REUTERS

But soon, he took to YouTube and encountered the warnings of people like Eliezer Yudkowsky — a researcher and writer who has said synthetic basic intelligence might threaten human life: “If anyone builds it, everyone dies.”

“If most people were as educated as me on this topic, if they knew the amount of information, the amount of statistics I knew, they would probably lean towards my position pretty heavily, I’m guessing,” Moreno-Gama told Mills. “Like, if there was a bridge where the engineer said there’s a 25% chance that it collapses, most people probably wouldn’t take that.”

He also admitted that he felt compelled to “do something to… get [AI doomerism] to be part of the public conscious a little bit more than it is currently.”

“[I] reached out and said, ‘Hey, man, what did you have in mind when you talk about violence?’” Mills told The Post of Daniel Moreno-Gama. “And he said, ‘How about Luigi-ing some tech CEOs?’” — a reference to Luigi Mangione (above), who is accused of murdering United Healthcare CEO Brian Thompson in December 2024. Steven Hirsch

Still, Mills walked away from the interview with no sense that Moreno-Gama would truly act on the fantasy of “Luigi-ing” CEOs.

“I’ve talked to violent offenders plenty of times in my career,” the host told The Post. “He did not strike me as someone on the edge of violence… People say s–t on the internet all the time, like, ‘F–k ICE, let’s kill them.’ But how many people would actually do it?”

But he did detect a need for notoriety. 

“The guy was not only down to do [an interview], he was, like, pumped about it,” Mills recalled. “He was so eager to come on our podcast, he even followed up with us afterwards, being like, ‘When’s it gonna publish? When’s it gonna publish?’”

Daniel Moreno-Gama encountered the YouTube teachings of people like Eliezer Yudkowsky (pictured) — a researcher and writer who has said synthetic basic intelligence might threaten human life: “If anyone builds it, everyone dies.” Twitter / Eliezer Yudkowsky

He also believes Moreno-Gama carried a trace of martyrdom.

“I think that there’s elements in his behavior that resemble some of what we see out of school shooters — this sort of desire to ‘Make sure people know it was me and I was a warrior.’”

Stay informed with the latest in tech! Our web site is your trusted source for breakthroughs in artificial intelligence, gadget launches, software program updates, cybersecurity, and digital innovation.

For contemporary insights, professional coverage, and trending tech updates, go to us frequently by clicking right here.