Robots let ChatGPT touch the real world thanks to Microsoft

A new API allows ChatGPT to control robots through natural language commands.

A drone flying over a city.

Enlarge (credit: Microsoft)

Last week, Microsoft researchers announced an experimental framework to control robots and drones using the language abilities of ChatGPT, a popular AI language model created by OpenAI. Using natural language commands, ChatGPT can write special code that controls robot movements. A human then views the results and adjusts as necessary until the task gets completed successfully.

The research arrived in a paper titled "ChatGPT for Robotics: Design Principles and Model Abilities," authored by Sai Vemprala, Rogerio Bonatti, Arthur Bucker, and Ashish Kapoor of the Microsoft Autonomous Systems and Robotics Group.

In a demonstration video, Microsoft shows robots—apparently controlled by code written by ChatGPT while following human instructions—using a robot arm to arrange blocks into a Microsoft logo, flying a drone to inspect the contents of a shelf, or finding objects using a robot with vision capabilities.

Read 5 remaining paragraphs | Comments

FDA official behind Alzheimer’s drug scandal steps down

Investigations alleged Billy Dunn had inappropriately close relationship with drug maker.

FDA official behind Alzheimer’s drug scandal steps down

Enlarge (credit: Getty | Congressional Quarterly)

The Food and Drug Administration official who allegedly had an inappropriately cozy relationship with the maker of the controversial Alzheimer's drug, Aduhelm, is stepping down from his role, effective immediately, according to numerous media reports.

Billy Dunn, head of the FDA's neuroscience office, has been with the agency for around 18 years, during which he was involved in several high-profile drug approvals. But he gained notoriety in the wake of the shocking 2021 approval of Aduhelm, a drug has not been shown to be effective against Alzheimer's and carries risks of serious brain swelling.

In 2020, the FDA's independent advisory committee voted nearly unanimously that the agency should not approve the drug (11 of 12 committee members voted against, with one voting "uncertain"). Yet, the approval went through anyway, and Aduhelm's maker, Biogen, went on to set the exorbitant price of $56,000 for a year's worth of treatment. Outside researchers called the approval "disgraceful," three advisory members resigned in protest, and Medicare Part B premiums rose by record amounts to accommodate covering the pricey drug for Alzheimer's patients desperate for new treatments.

Read 5 remaining paragraphs | Comments

Breaks taken during psych experiments lower participants’ moods

The lowered mood could throw off the results of studies with rest breaks.

Image of a child looking extremely bored and unhappy

Enlarge (credit: Klaus Vedfelt)

An unfortunate feature of science is that two experiments that are ostensibly looking at the same thing can produce different results. Often, the different results are greeted unhelpfully as the experimenters—and sometimes even the entire field—are accused of being garbage. A more helpful response is to consider whether the experiments, while looking at the same thing, might not be identical. And, if they're not, whether the differences between them might tell us something.

A new study in Nature Human Behavior describes a subtle way some psychology experiments could differ: if they include breaks to let their participants avoid tiring out. Enforced breaks can cause people's moods to drop and continue dropping if the break drags on. And, since mood affects behavior in a variety of other psychological tests, this has the potential to have a complicating influence on a huge range of studies.

Waiting is the hardest part

The work began with an incredibly simple finding. Most studies operate under the assumption that a participant's mood remains relatively stable throughout an experiment. But the researchers here asked participants to rate their mood at the start and end of experiments—and thus at the start and end of breaks between the experiments. The researchers noticed that the mood went down pretty consistently over the course of the break. After a roughly 10-minute break, people assessed their mood as more than 20 percent lower than when the break started.

Read 9 remaining paragraphs | Comments

Lilbits: Nothing Phone (2), the next Google Pixel phone, and Amlogic’s new SN928X chip for 8K TV boxes

Now that Nothing Phone (1) knockoffs are starting to appear, how does the smartphone startup plan to make sure its next phone stands out? Well first, I doubt Unihertz has Nothing execs quaking in their boots. But second, the company is planning to mak…

Now that Nothing Phone (1) knockoffs are starting to appear, how does the smartphone startup plan to make sure its next phone stands out? Well first, I doubt Unihertz has Nothing execs quaking in their boots. But second, the company is planning to make the Nothing Phone (2) a true flagship by giving it a […]

The post Lilbits: Nothing Phone (2), the next Google Pixel phone, and Amlogic’s new SN928X chip for 8K TV boxes appeared first on Liliputing.

US Court Denies Bungie’s $2.2m Claim Against UK ‘Ring-1’ Cheat Seller

Bungie and Ubisoft’s request for default judgment in a prominent cheating lawsuit has been denied by a California federal court. The companies sought $2.2 million in copyright-related damages from a UK defendant they described as a key player in the “Ring-1″ group. The court concluded that the man was little more than a customer support staffer.

From: TF, for the latest news on copyright battles, piracy and more.

ring-1 logoOver the past several years, several videogame companies have taken cheaters to court in the United States.

In 2021, American videogame companies Bungie and Ubisoft joined forces in a lawsuit against “Ring-1”, a developer and distributor of cheat software targeting Destiny 2, Rainbox Six Seige, and other popular titles.

Bungie and Ubisoft identified four defendants who allegedly ran the cheating business. Filed at a California federal court, their complaint named Andrew Thorpe (aka ‘Krypto’), Jonathan Aguedo (aka ‘Overpowered’), Wesam Mohammed (aka ‘Grizzly’), and Ahmad Mohammed as key players.

Three Settlements

Aguedo and the two Mohammeds were tracked down in the United States and eventually admitted their wrongdoing. This resulted in settlement agreements totaling $600,000 in damages.

Thorpe, a resident of the UK, was the only defendant who failed to respond to the allegations, despite being properly informed. In the absence of a formal defense, Bungie and Ubisoft asked the court to issue a default judgment.

According to the videogame companies, Mr. Thorpe was a “prominent” member of Ring-1 who, among other things, “ran” the group’s official website. This claim was corroborated by the testimony of one of the co-defendants.

Bungie and Ubisoft requested $2.2 million in damages, a figure that includes compensation for several claims, including copyright infringement and trafficking in circumvention devices.

Unrepresented defendants run the risk of evidence weighing in favor of the plaintiffs. In this case, however, U.S. District Court Judge Edward Chen took a more balanced approach.

Court Denies Default Judgement

After reviewing the arguments and evidence presented by the videogame companies, Judge Chen denied their motion for default judgment. This means that the fourth and final defendant won’t be required to pay any damages.

The key issue at stake is whether the California federal court has personal jurisdiction over the UK resident. Specifically, Bungie and Ubisoft were required to show that he was a key player who directed his actions at the United States.

In his order, Judge Chen first considers the role of the defendant, relying on the evidence provided by the game companies. This evidence failed to show that Mr. Thorpe was a prominent Ring-1 member.

“Mr. Thorpe is not an original developer of the software or an original participant in the Ring-1 enterprise, but only joined after Ring-1 had already attracted many users,” Judge Chen writes.

“His role at the Ring-1 enterprise appears to be akin to a customer service representative. For example, customers have asked him for information related to the Cheating Software such as its features, operations, and updates.”

Jurisdiction

Bungie and Ubisoft argued that the defendant’s actions were targeted at the U.S., based on the notion that the broader actions of Ring-1 can be attributed to him. That goes too far in this case, as the court fails to see him as a leader.

Co-defendant Mr. Agueda testified that Mr. Thorpe ran the Ring-1 website. However, Judge Chen believes that the term “run” is vague and is therefore insufficient to consider the defendant an alter ego for the entire Ring-1 group.

Instead, the court decided to evaluate whether Mr. Thorpe specifically directed his personal actions at the U.S., which would be sufficient to establish personal jurisdiction. Again, Judge Chen found the evidence unconvincing.

“First, contrary to what Plaintiffs argue, the fact that the two Plaintiffs are located in the United States does not mean that Mr. Thorpe has thereby targeted the United States,” Judge Chen writes.

Second, Mr. Thorpe may have assisted customers, but there is no evidence that any of those customers were based in the United States. More fundamentally, there is nothing to indicate that Mr. Thorpe reached out and solicited those customers. Rather, the evidence indicates that the customers reached out to him.

Insufficient Evidence

Judge Chen ultimately decided that there wasn’t enough evidence to conclude that Mr. Thorpe purposefully directed his activities at the United States. As such, the court doesn’t have personal jurisdiction over the UK resident.

That left no other option than to deny Bungie and Ubisoft’s motion for default judgment and the $2.2 million in compensation they requested.

“Without the requisite minimum contacts, the Court does not have specific personal jurisdiction over Thorpe pursuant to Rule 4(k)(2), and thus a default judgment is not warranted,” the order reads.

In a footnote, Judge Chen writes that the injunctions awarded against the other defendants – prohibiting the Ring-1 members from engaging in any infringing conduct going forward – also apply to Mr. Thorpe but notes that his court has no power to enforce compliance.

Whether Mr. Thorpe is still part of the operation is unknown but, despite the previous settlements, the Ring-1 continues to be online and publicly accessible to this day.

The ruling in this lawsuit stands in sharp contrast with the criminal conviction of Gary Bowser, who was a relatively low-level ‘salesperson’ of the Team-Xecuter ‘hacking’ group. Last year, a U.S. court sentenced the Canadian to 40 months in prison for that role.

A copy of U.S. District Court Judge Edward Chan’s order, denying the motion for a default judgment, is available here (pdf)

From: TF, for the latest news on copyright battles, piracy and more.

Tesla shareholder suit says Musk and co. lied about Full Self-Driving safety

Investor lawsuit cites recall of Tesla cars that act dangerously in intersections.

Aerial view of cars parked in a Tesla factory parking lot. A Tesla logo is painted on the concrete.

Enlarge / Cars parked at the Tesla Fremont Factory in Fremont, California, on February 10, 2022. (credit: Getty Images | Josh Edelson)

A class-action complaint alleges that Tesla and CEO Elon Musk repeatedly made false statements about the capabilities and safety of the electric carmaker's Autopilot and Full Self-Driving (FSD) technology.

The complaint, filed Monday in US District Court for the Northern District of California, comes less than two weeks after a recall of 362,758 cars based on a US government finding that Tesla's "FSD Beta system may allow the vehicle to act unsafe around intersections, such as traveling straight through an intersection while in a turn-only lane, entering a stop sign-controlled intersection without coming to a complete stop, or proceeding into an intersection during a steady yellow traffic signal without due caution." The problem is slated to be fixed by an over-the-air software update.

The lawsuit was filed by investor Thomas Lamontagne and seeks to represent a proposed class of potentially thousands of people who acquired Tesla stock. Tesla, Elon Musk, and other Tesla executives "made materially false and misleading statements regarding the Company's business, operations, and prospects," the lawsuit said, continuing:

Read 13 remaining paragraphs | Comments

Windows 11 update brings AI-powered Bing to the Search Box, tabs to Notepad, Phone Link for iOS, and more

If you’ve ever wished your computer could tell you it loves you… and that you don’t love your spouse, then you may be happy to know that Microsoft is bringing its new AI-powered Bing to the Windows 11 Search Box. It’s just one …

If you’ve ever wished your computer could tell you it loves you… and that you don’t love your spouse, then you may be happy to know that Microsoft is bringing its new AI-powered Bing to the Windows 11 Search Box. It’s just one of many new features rolling out today as part of a major […]

The post Windows 11 update brings AI-powered Bing to the Search Box, tabs to Notepad, Phone Link for iOS, and more appeared first on Liliputing.

Miraxess MiraDock turns (some) smartphones into desktop computers

The company behind the Mirabook laptop dock that lets you use your smartphone as a notebook computer have launched a new, cheaper accessory called the MiraDock that lets you use your smartphone like a desktop PC. Available for pre-order for 139 € (abo…

The company behind the Mirabook laptop dock that lets you use your smartphone as a notebook computer have launched a new, cheaper accessory called the MiraDock that lets you use your smartphone like a desktop PC. Available for pre-order for 139 € (about $147), the MiraDock is expected to ship in the first quarter of 2023. […]

The post Miraxess MiraDock turns (some) smartphones into desktop computers appeared first on Liliputing.

The Pixel Watch’s promised fall detection is finally rolling out

Just like a 2018 Apple Watch, the Pixel Watch can call 911 after a hard fall.

Fall Detection on the Pixel Watch.

Enlarge / Fall Detection on the Pixel Watch. (credit: Google)

With the launch of the Pixel Watch in October, Google started slowly trying to claw its way back into smartwatch relevance. A few months later, it's launching a big feature that has been on the Apple Watch for four years: fall detection.

As Google describes the feature, "fall detection uses motion sensors built in to your watch and on-device machine learning to determine if you’ve taken a hard fall and will call emergency services if you don’t move or respond within a certain timeframe." Like on an Apple Watch, if a fall is detected, you'll get a full-screen pop-up asking if you're OK. If you don't respond, an alarm will sound for the next minute, and if you still don't respond, the watch will call 911. Fall detection has been a big reason at-risk people like the elderly have an Apple Watch, so it's good that Google is finally delivering this feature.

The Apple Watch version, along with car crash detection, had led to some false positives, with the watch dialing 911 when the users weren't hurt. Google says it has worked to stop that, saying, "Worried that doing burpees or hitting the ski slopes will trigger an accidental emergency call? Your Pixel Watch knows the difference between taking a hard fall and performing a vigorous physical activity or even quickly recovering from a small stumble—thanks to our machine learning algorithms and rigorous testing." Google calls out skiing, burpees, jumping, and swimming as activities that could cause false positives, but hopefully it has worked out the bugs through the "rigorous" testing process.

Read 1 remaining paragraphs | Comments

Beyond Good & Evil 2 studio rocked by reported gov’t labor investigation

Management shakeup comes as dozens call in sick for stress and/or burnout.

The team that brought you this <em>Beyond Good & Evil 2</em> concept art is not doing well, according to a new report.

Enlarge / The team that brought you this Beyond Good & Evil 2 concept art is not doing well, according to a new report.

If you've been waiting patiently for Beyond Good & Evil 2 since its first announcement back in 2008, we have some bad news about the studio responsible for the long-delayed project's development. Kotaku now reports that Ubisoft Montpellier has lost its managing director amid a local government investigation into labor policies that have left "an unprecedented number of developers experiencing burnout and going on sick leave."

Ubisoft Montpellier, which also developed the first Beyond Good & Evil two decades ago now, is being looked at by its local branch of the Inspection du Travail, according to "three sources familiar with the development," cited by Kotaku. In December, that office reportedly started looking into reports that dozens of developers had taken extended leave over stress or sickness over the last year. Employees have been interviewed regarding their health and well-being by a third party, according to the report.

“The health and wellness of our teams is an ongoing priority,” an Ubisoft spokesperson told Kotaku. “Given the length of the development cycle with Beyond Good & Evil 2, the Montpellier development team is undergoing well-being assessments through a third-party for preventative measures and to evaluate where additional support may be needed.”

Read 4 remaining paragraphs | Comments