Author name: Shannon Garcia

chatgpt-unexpectedly-began-speaking-in-a-user’s-cloned-voice-during-testing

ChatGPT unexpectedly began speaking in a user’s cloned voice during testing

An illustration of a computer synthesizer spewing out letters.

On Thursday, OpenAI released the “system card” for ChatGPT’s new GPT-4o AI model that details model limitations and safety testing procedures. Among other examples, the document reveals that in rare occurrences during testing, the model’s Advanced Voice Mode unintentionally imitated users’ voices without permission. Currently, OpenAI has safeguards in place that prevent this from happening, but the instance reflects the growing complexity of safely architecting with an AI chatbot that could potentially imitate any voice from a small clip.

Advanced Voice Mode is a feature of ChatGPT that allows users to have spoken conversations with the AI assistant.

In a section of the GPT-4o system card titled “Unauthorized voice generation,” OpenAI details an episode where a noisy input somehow prompted the model to suddenly imitate the user’s voice. “Voice generation can also occur in non-adversarial situations, such as our use of that ability to generate voices for ChatGPT’s advanced voice mode,” OpenAI writes. “During testing, we also observed rare instances where the model would unintentionally generate an output emulating the user’s voice.”

In this example of unintentional voice generation provided by OpenAI, the AI model outbursts “No!” and continues the sentence in a voice that sounds similar to the “red teamer” heard in the beginning of the clip. (A red teamer is a person hired by a company to do adversarial testing.)

It would certainly be creepy to be talking to a machine and then have it unexpectedly begin talking to you in your own voice. Ordinarily, OpenAI has safeguards to prevent this, which is why the company says this occurrence was rare even before it developed ways to prevent it completely. But the example prompted BuzzFeed data scientist Max Woolf to tweet, “OpenAI just leaked the plot of Black Mirror’s next season.”

Audio prompt injections

How could voice imitation happen with OpenAI’s new model? The primary clue lies elsewhere in the GPT-4o system card. To create voices, GPT-4o can apparently synthesize almost any type of sound found in its training data, including sound effects and music (though OpenAI discourages that behavior with special instructions).

As noted in the system card, the model can fundamentally imitate any voice based on a short audio clip. OpenAI guides this capability safely by providing an authorized voice sample (of a hired voice actor) that it is instructed to imitate. It provides the sample in the AI model’s system prompt (what OpenAI calls the “system message”) at the beginning of a conversation. “We supervise ideal completions using the voice sample in the system message as the base voice,” writes OpenAI.

In text-only LLMs, the system message is a hidden set of text instructions that guides behavior of the chatbot that gets added to the conversation history silently just before the chat session begins. Successive interactions are appended to the same chat history, and the entire context (often called a “context window”) is fed back into the AI model each time the user provides a new input.

(It’s probably time to update this diagram created in early 2023 below, but it shows how the context window works in an AI chat. Just imagine that the first prompt is a system message that says things like “You are a helpful chatbot. You do not talk about violent acts, etc.”)

A diagram showing how GPT conversational language model prompting works.

Enlarge / A diagram showing how GPT conversational language model prompting works.

Benj Edwards / Ars Technica

Since GPT-4o is multimodal and can process tokenized audio, OpenAI can also use audio inputs as part of the model’s system prompt, and that’s what it does when OpenAI provides an authorized voice sample for the model to imitate. The company also uses another system to detect if the model is generating unauthorized audio. “We only allow the model to use certain pre-selected voices,” writes OpenAI, “and use an output classifier to detect if the model deviates from that.”

ChatGPT unexpectedly began speaking in a user’s cloned voice during testing Read More »

it’s-not-worth-paying-to-be-removed-from-people-finder-sites,-study-says

It’s not worth paying to be removed from people-finder sites, study says

Better than nothing but not by enough —

The best removal rate was less than 70%, and that didn’t beat manual opt-outs.

Folks in suits hiding behind trees that do not really obscure them

Enlarge / For a true representation of the people-search industry, a couple of these folks should have lanyards that connect them by the pockets.

Getty Images

If you’ve searched your name online in the last few years, you know what’s out there, and it’s bad. Alternately, you’ve seen the lowest-common-denominator ads begging you to search out people from your past to see what crimes are on their record. People-search sites are a gross loophole in the public records system, and it doesn’t feel like there’s much you can do about it.

Not that some firms haven’t promised to try. Do they work? Not really, Consumer Reports (CR) suggests in a recent study.

“[O]ur study shows that many of these services fall short of providing the kind of help and performance you’d expect, especially at the price levels some of them are charging,” said Yael Grauer, program manager for CR, in a statement.

Consumer Reports’ study asked 32 volunteers for permission to try to delete their personal data from 13 people-search sites, using seven services over four months. The services, including DeleteMe, Reputation Defender from Norton, and Confidently, were also compared to “Manual opt-outs,” i.e. following the tucked-away links to pull down that data on each people-search site. CR took volunteers from California, in which the California Consumer Privacy Act should theoretically make it mandatory for brokers to respond to opt-out requests, and in New York, with no such law, to compare results.

Table from Consumer Reports' study of people-search removal services, showing effective removal rates over time for each service.

Table from Consumer Reports’ study of people-search removal services, showing effective removal rates over time for each service.

Finding a total of 332 instances of identifying information profiles on those sites, Consumer Reports found that only 117 profiles were removed within four months using all the services, or 35 percent. The services varied in efficacy, with EasyOptOuts notably performing the second-best at a 65 percent removal rate after four months. But if your goal is to remove entirely others’ ability to find out about you, no service Consumer Reports tested truly gets you there.

Manual opt-outs were the most effective removal method, at 70 percent removed within one week, which is both a higher elimination rate and quicker turn-around than all the automated services.

The study noted close ties between the people-search sites and the services that purport to clean them. Removing one volunteer’s data from ClustrMaps resulted in a page with a suggested “Next step”: signing up for privacy protection service OneRep. Firefox-maker Mozilla dropped OneRep as a service provider for its Mozilla Monitor Plus privacy bundle after reporting by Brian Krebs found that OneRep’s CEO had notable ties to the people-search industry.

In releasing this study, CR also advocates for laws at the federal and state level, like California’s Delete Act, that would make people-search removal far easier than manually scouring the web or paying for incomplete monitoring.

CR’s study cites CheckPeople, PublicDataUSA, and Intelius as the least responsive businesses in one of the least responsive industries, while noting that PeopleFinders, ClustrMaps, and ThatsThem deserve some very tiny, nearly inaudible recognition for complying with opt-out requests (our words, not theirs).

It’s not worth paying to be removed from people-finder sites, study says Read More »

all-the-possible-ways-to-destroy-google’s-monopoly-in-search

All the possible ways to destroy Google’s monopoly in search

All the possible ways to destroy Google’s monopoly in search

Aurich Lawson

After US District Judge Amit Mehta ruled that Google has a monopoly in two markets—general search services and general text advertising—everybody is wondering how Google might be forced to change its search business.

Specifically, the judge ruled that Google’s exclusive deals with browser and device developers secured Google’s monopoly. These so-called default agreements funneled the majority of online searches to Google search engine result pages (SERPs), where results could be found among text ads that have long generated the bulk of Google’s revenue.

At trial, Mehta’s ruling noted, it was estimated that if Google lost its most important default deal with Apple, Google “would lose around 65 percent of its revenue, even assuming that it could retain some users without the Safari default.”

Experts told Ars that disrupting these default deals is the most obvious remedy that the US Department of Justice will seek to restore competition in online search. Other remedies that may be sought range from least painful for Google (mandating choice screens in browsers and devices) to most painful (requiring Google to divest from either Chrome or Android, where it was found to be self-preferencing).

But the remedies phase of litigation may have to wait until after Google’s appeal, which experts said could take years to litigate before any remedies are ever proposed in court. Whether Google could be successful in appealing the ruling is currently being debated, with anti-monopoly advocates backing Mehta’s ruling as “rock solid” and critics suggesting that the ruling’s fresh takes on antitrust law are open to attack.

Google declined Ars’ request to comment on appropriate remedies or its plan to appeal.

Previously, Google’s president of global affairs, Kent Walker, confirmed in a statement that the tech giant would be appealing the ruling because the court found that “Google is ‘the industry’s highest quality search engine, which has earned Google the trust of hundreds of millions of daily users,’ that Google ‘has long been the best search engine, particularly on mobile devices,’ ‘has continued to innovate in search,’ and that ‘Apple and Mozilla occasionally assess Google’s search quality relative to its rivals and find Google’s to be superior.'”

“Given this, and that people are increasingly looking for information in more and more ways, we plan to appeal,” Walker said. “As this process continues, we will remain focused on making products that people find helpful and easy to use.”

But Mehta found that Google was wielding its outsize influence in the search industry to block rivals from competing by locking browsers and devices into agreements ensuring that all searches went to Google SERPs. None of the pro-competitive benefits that Google claimed justified the exclusive deals persuaded Mehta, who ruled that “importantly,” Google “exercised its monopoly power by charging supra-competitive prices for general search text ads”—and thus earned “monopoly profits.”

While experts think the appeal process will delay litigation on remedies, Google seems to think that Mehta may rule on potential remedies before Google can proceed with its appeal. Walker told Google employees that a ruling on remedies may arrive in the next few months, The Wall Street Journal reported. Ars will continue monitoring for updates on this timeline.

As the DOJ’s case against Google’s search business has dragged on, reports have long suggested that a loss for Google could change the way that nearly the entire world searches the Internet.

Adam Epstein—the president and co-CEO of adMarketplace, which bills itself as “the largest consumer search technology company outside of Google and Bing”—told Ars that innovations in search could result in a broader landscape of more dynamic search experiences that draw from sources beyond Google and allow searchers to skip Google’s SERPs entirely. If that happens, the coming years could make Google’s ubiquitous search experience today a distant memory.

“By the end of this decade, going to a search engine results page will seem quaint,” Epstein predicted. “The court’s decision sets the stage for a remedy that will dramatically improve the search experience for everyone connected to the web. The era of innovation in search is just around the corner.”

The DOJ has not meaningfully discussed potential remedies it will seek, but Jonathan Kanter, assistant attorney general of the Justice Department’s antitrust division, celebrated the ruling.

“This landmark decision holds Google accountable,” Kanter said. “It paves the path for innovation for generations to come and protects access to information for all Americans.”

All the possible ways to destroy Google’s monopoly in search Read More »

macos-15-sequoia-makes-you-jump-through-more-hoops-to-disable-gatekeeper-app-checks

macOS 15 Sequoia makes you jump through more hoops to disable Gatekeeper app checks

gate-kept —

But nothing is changing about the kinds of software you can run on your Mac.

The Mac's Gatekeeper feature has been pushing developers to digitally sign their apps since it was introduced in 2012.

Enlarge / The Mac’s Gatekeeper feature has been pushing developers to digitally sign their apps since it was introduced in 2012.

Apple/Andrew Cunningham

It has always been easier to run third-party software on a Mac than on an iPhone or iPad. Despite the introduction of the Mac App Store a couple of years after the iPhone’s App Store opened, it has always been possible to download and run third-party scripts and software on your Mac from anywhere. It’s one reason why the iPhone and iPad are subject to new European Union regulations about software sideloading and third-party app stores, while the Mac isn’t.

That’s not changing in macOS 15 Sequoia, the new version of macOS that’s due to be released to the public this fall. But it is about to get more annoying for some apps, according to a note added to Apple’s developer site yesterday.

“In macOS Sequoia, users will no longer be able to Control-click to override Gatekeeper when opening software that isn’t signed correctly or notarized,” the brief note reads. “They’ll need to visit System Settings > Privacy & Security to review security information for software before allowing it to run.”

Users (including me) had noticed this behavior in early macOS Sequoia betas, but this note confirms that the change was made on purpose and that the software is working as intended.

What’s changing and what isn’t

To understand what’s changing, it’s helpful to understand how macOS handles third-party apps. Though software can be downloaded and run in macOS from everywhere, Apple encourages developers to digitally sign their software and send it to Apple for notarization, which Apple describes as “an automated system that scans your software for malicious content, checks for code-signing issues, and returns the results to you quickly.” Notably, it is not the same as the app review process in Apple’s App Stores, where humans check submitted apps and can refuse to distribute them if they run afoul of Apple’s rules.

Notarization does come with benefits for users—users can be sure that the apps haven’t been tampered with and can run them with minimal hassle from Gatekeeper, macOS’ app-screening security feature. But it creates an extra step for developers and requires the use of a $100-a-year paid Apple Developer account, something that may not be worth the cost for hobby projects or open source projects that don’t generate much (or any) income for their contributors.

Unsigned, non-notarized software will refuse to run in current macOS versions, but it has always been possible to right-click or control-click the app or script you want to run and then click Open, which exposes an “open anyway” option in a dialog box that lets you launch the software. Once you’ve made an exception for an app, you can run it like you would any other app unless the software is updated or changed in some way.

The section of the Settings app where you'll need to go in macOS Sequoia to allow unsigned apps to run.

Enlarge / The section of the Settings app where you’ll need to go in macOS Sequoia to allow unsigned apps to run.

Andrew Cunningham

Which gets us to what Sequoia changes. The right-click/control-click option for easily opening unsigned apps is no longer available. Users who want to open unsigned software will now need to go the long way around to do it: first, try to launch the app and dismiss the dialog box telling you that it can’t be opened. Then, open Settings, go to the Privacy & Security screen, scroll all the way to the bottom to get to the Security section, and click the Open Anyway button that appears for the last unsigned app you tried to run.

This has always been an option for skirting around Gatekeeper, going all the way back to the days when Settings was still System Preferences (and when Apple would let you disable Gatekeeper’s checks entirely, something it removed in 2016). But it takes so much more time that I never actually did it that way once I discovered the right-click trick. Now, doing it the long way is mandatory.

I don’t want to oversell how disruptive this is—generally once you allow an app to run the first time, you don’t have to think about it again unless the app is updated or otherwise modified or tampered with. Apple isn’t allowing or disallowing any new behavior in macOS. Popular apps from major developers do tend to be notarized, rendering this change irrelevant. And if this change pushes more developers to sign and notarize their apps, that is arguably a win for user security and convenience.

But for most people most of the time, it’s just going to make a minor annoyance into a medium-size annoyance. And among the conspiratorially minded, it’s going to reignite 12-year-old anxieties about Apple locking macOS down to the same degree that it already locks down iOS and iPadOS.

The macOS 15 Sequoia update is currently available to developers and the general public as a beta if you’ve signed up for either of Apple’s beta programs. An early iteration of the 15.1 update with some Apple Intelligence generative AI features enabled is also available to developers with Apple Silicon Macs.

macOS 15 Sequoia makes you jump through more hoops to disable Gatekeeper app checks Read More »

report:-apple’s-external-dvd-drive-is-up-burning-discs-in-dongle-heaven

Report: Apple’s external DVD drive is up burning discs in dongle heaven

ashes to ashes, disc to disc —

Other DVD drives are cheap and plentiful, but Apple’s slot-loader was unique.

Apple's external DVD-burning SuperDrive may be fading away.

Enlarge / Apple’s external DVD-burning SuperDrive may be fading away.

Apple

Apple has always been eager to dump technologies when the company feels they have outlived their usefulness. The original iMac came without a floppy drive. The iPhone 7 came without a headphone jack. Mid-2010s MacBooks and MacBook Pros came with USB-A ports. And the original 2008 MacBook Air came without a built-in optical drive for CDs and DVDs. By the time 2012 and 2013 Macs rolled around, products from the iMac to the MacBook Pro followed suit.

These exclusions have often made Apple’s devices thinner, lighter, sleeker, or some combination of all three. But they’ve also meant that people who still needed those technologies also needed to deal with dongles, adapters, or clunky external accessories hanging off their devices. For the MacBook Air and other modern Macs that needed to read or burn optical discs, that clunky accessory was Apple’s SuperDrive, an external DVD burner that connected via USB.

After 16 years of availability, it looks like the SuperDrive’s run could be coming to an end. As noticed by MacRumors, the drive’s status has shifted to “sold out” in Apple’s online store, a more definitive and permanent-sounding label than the “currently unavailable” status assigned to some other out-of-stock products.

Though it’s been more than a decade since Apple introduced a new Mac with an optical drive built in, modern versions of macOS still have roughly the same level of support for CD and DVD drives that they did back when optical drives were standard-issue equipment. Plug an optical drive into a modern Mac—whether it’s a SuperDrive or a third-party model—and you’ll still be able to burn and rip audio CDs with the Music app, rip or burn CD and DVD image files with Finder or Disk Utility, or burn files to a disc for archiving with the Finder. Even the venerable DVD Player app is still included, though macOS relies mostly on third-party software to handle Blu-ray discs.

Third-party external DVD drives can be had for as little as $20, and external Blu-ray drives start around $50, making the $79 DVD-only SuperDrive an iffy financial proposition. It was also never updated with a USB-C connector, so connecting it to any modern MacBook requires yet another dongle. But Apple’s drive was unique, as it was a metal, slot-loading optical drive from a major manufacturer; SuperDrive clones on Amazon go for $30 or $40 but come from no-name companies and have mixed customer reviews. For now, if the news of its potential demise suddenly makes you want one, the genuine SuperDrive is still in stock at Amazon and Best Buy, among a few other third-party retailers.

We’ve contacted Apple to check on the status of the SuperDrive and will update this article if we receive a definitive response.

Report: Apple’s external DVD drive is up burning discs in dongle heaven Read More »

31%-of-republicans-say-vaccines-are-more-dangerous-than-diseases-they-prevent

31% of Republicans say vaccines are more dangerous than diseases they prevent

Vaccines save lives —

The partisan divide on vaccine falsehoods threatens the health of children nationwide.

Polio victim Larry Montoya is at the airport for the arrival of cases of vaccine, which were distributed as part of the KO Polio campaign, September 5, 1962.

Enlarge / Polio victim Larry Montoya is at the airport for the arrival of cases of vaccine, which were distributed as part of the KO Polio campaign, September 5, 1962.

Public sentiment on the importance of safe, lifesaving childhood vaccines has significantly declined in the US since the pandemic—which appears to be solely due to a nosedive in support from people who are Republican or those who lean Republican, according to new polling data from Gallup.

In 2019, 52 percent of Republican-aligned Americans said it was “extremely important” for parents to get their children vaccinated. Now, that figure is 26 percent, falling by half in just five years. In comparison, 63 percent of Democrats and Democratic leaners said it was “extremely important” this year, down slightly from 67 percent in 2019.

Overall, only 40 percent of Americans now say it is extremely important for parents to vaccinate their children, down from 58 percent in 2019 and 64 percent in 2001.

More broadly, 93 percent of the Democratic group said it was “extremely” or “very” important for parents to vaccinate their children this year, while only 52 percent of the Republican group said the same.

On the other end of the spectrum, 11 percent of the Republican group said vaccinating children was “not important at all,” and an additional 8 percent said it was “not very important.” For the Democratic group, only 1 percent was reported in each of those categories.

Dangerous disinformation

Perhaps most concerning, the data indicated that a growing number of Americans view vaccines as more dangerous than the diseases they prevent—including polio, measles, tetanus, rotavirus, diphtheria, whooping cough, meningitis, and RSV, among others. Now, 20 percent of Americans overall think vaccines are more of a threat than the dangerous diseases they effectively prevent.

The partisan divide is most stark on this sentiment. In 2019, the two parties were about the same. Twelve percent of the Republican group and 10 percent of the Democratic group held this erroneous belief. But now, a whopping 31 percent of the Republican group say vaccines are a more significant threat than dangerous diseases, while the percentage among the Democratic group fell to 5 percent.

Republicans and Republican leaners are much more likely than Democrats and Democratic leaners to believe the false and debunked claim that vaccines are linked to autism—19 percent of the Republican group believe this falsehood compared to 4 percent of the Democratic group.

The polling data aligns with national vaccination trends tracked by the Centers for Disease Control and Prevention. During the pandemic, rates of routine vaccination among kindergartners slipped from the protective target of 95 percent—which prevents infectious diseases from spreading widely—to 93 percent. Additionally, nonmedical exemptions from vaccinations have reached an all-time high of 3 percent nationally. At least 10 states have exemption rates at or above 5 percent, preventing them from reaching the protective target of 95 percent vaccination coverage.

31% of Republicans say vaccines are more dangerous than diseases they prevent Read More »

microsoft-says-delta’s-ancient-it-explains-long-outage-after-crowdstrike-snafu

Microsoft says Delta’s ancient IT explains long outage after CrowdStrike snafu

Your bad —

“Delta, unlike its competitors… has not modernized its IT infrastructure.”

Delta Air Lines customers looking for missing bags wait in line in an airport baggage claim area.

Enlarge / Delta Air Lines customers looking for missing bags wait in line at Los Angeles International Airport (LAX) on July 24, 2024.

Getty Images | Patrick T. Fallon

Microsoft says that Delta Air Lines’ ancient IT infrastructure is to blame for the airline’s inability to quickly recover from last month’s CrowdStrike debacle.

With Delta threatening to sue Microsoft and CrowdStrike, both companies issued responses saying that Delta refused repeated calls for help. A Microsoft letter to Delta yesterday said the Windows maker is starting to figure out why Delta took longer than other airlines to recover.

“Microsoft continues to investigate the circumstances surrounding the CrowdStrike incident to understand why other airlines were able to fully restore business operations so much faster than Delta, including American Airlines and United Airlines,” the letter from Microsoft attorney Mark Cheffo said. “Our preliminary review suggests that Delta, unlike its competitors, apparently has not modernized its IT infrastructure, either for the benefit of its customers or for its pilots and flight attendants.”

On July 19, a faulty update from security firm CrowdStrike crashed millions of Windows PCs. In a July 29 letter, Delta attorney David Boies said the airline has “reason to believe Microsoft has failed to comply with contractual requirements and otherwise acted in a grossly negligent, indeed willful, manner in connection with the Faulty Update,” according to CNBC.

Cheffo’s response to Boies said that “Microsoft empathizes with Delta and its customers regarding the impact of the CrowdStrike incident.” But the Boies letter “and Delta’s public comments are incomplete, false, misleading, and damaging to Microsoft and its reputation,” Cheffo wrote.

“Given Delta’s false and damaging public statements, Microsoft will vigorously defend itself in any litigation if Delta chooses to pursue that path,” Cheffo wrote. The letter demanded that Delta preserve documents related to the outage.

Delta allegedly refused offers to help

CrowdStrike previously wrote to Delta on Sunday. “CrowdStrike’s CEO personally reached out to Delta’s CEO to offer onsite assistance, but received no response. CrowdStrike followed up with Delta on the offer for onsite support and was told that the onsite resources were not needed,” the letter said.

Microsoft’s letter on Tuesday provided a similar description. “Even though Microsoft’s software had not caused the CrowdStrike incident, Microsoft immediately jumped in and offered to assist Delta at no charge following the July 19 outage,” the letter said. “Each day that followed from July 19 through July 23, Microsoft employees repeated their offers to help Delta. Each time, Delta turned down Microsoft’s offers to help, even though Microsoft would not have charged Delta for this assistance.”

The letter said that after one Microsoft outreach on July 22, a “Delta employee replied, saying ‘all good. Cool will let you know and thank you.’ Despite this assessment that things were ‘all good,’ public reports indicate that Delta canceled more than 1,100 flights on July 22 and more than 500 flights on July 23.”

Senior executives repeatedly reached out to Delta executives “with similar results,” the letter said. Microsoft CEO Satya Nadella emailed Delta CEO Ed Bastian on July 24, but Bastian never replied, according to the letter.

Microsoft says it has a pretty good idea of why Delta refused its help. “In fact, it is rapidly becoming apparent that Delta likely refused Microsoft’s help because the IT system it was most having trouble restoring—its crew-tracking and scheduling system—was being serviced by other technology providers, such as IBM, because it runs on those providers’ systems, and not Microsoft Windows or Azure,” the letter said.

We contacted Delta today and will update this article if the company provides a response.

Microsoft says Delta’s ancient IT explains long outage after CrowdStrike snafu Read More »

china-begins-launching-a-megaconstellation,-and-it-sounds-a-lot-like-starlink

China begins launching a megaconstellation, and it sounds a lot like Starlink

Sailing in LEO —

Like Starlink, China’s Qianfan satellites have an easy-to-pack flat-panel design.

A Long March 6A rocket launches the first 18 Internet satellites for China's Qianfan, or Thousand Sails, broadband network.

Enlarge / A Long March 6A rocket launches the first 18 Internet satellites for China’s Qianfan, or Thousand Sails, broadband network.

Chinese officials have long signaled their interest in deploying a satellite network, or maybe several, to beam broadband Internet signals across China and other nations within its sphere of influence.

Two serious efforts are underway in China to develop a rival to SpaceX’s Starlink network, which the Chinese government has banned in its territory. The first batch of 18 satellites for one of those Chinese networks launched into low-Earth orbit Tuesday.

A Long March 6A rocket delivered the 18 spacecraft into a polar orbit following liftoff at 2: 42 am EDT (06: 42 UTC) from the Taiyuan launch base in northern China’s Shanxi province. The Long March 6A is one of China’s newest rockets—and the country’s first to employ strap-on solid rocket boosters—with the ability to deploy a payload of up to 4.5 metric tons (9,900 pounds) into a 700-kilometer (435-mile) Sun-synchronous orbit.

The rocket placed its payload of 18 Qianfan satellites into the proper orbit, and the launch mission was a complete success, according to the China Aerospace Science and Technology Corporation, the largest state-owned contractor for the Chinese space program.

Qianfan translates to “Thousand Sails,” and the 18 satellites launched Tuesday are the first of potentially thousands of spacecraft planned by Shanghai Spacecom Satellite Technology (SSST), a company backed by Shanghai’s municipal government. The network developed by SSST is also called the “Spacesail Constellation.”

Shanghai officials only began releasing details of this constellation last year. A filing with the International Telecommunication Union suggests the developers of Shanghai-based megaconstellation initially plan to deploy 1,296 satellites at an altitude of about 1,160 kilometers (721 miles).

Xinhua, China’s state-run news agency, said the constellation “will provide global users with low-latency, high-speed and ultra-reliable satellite broadband Internet services.”

Opening the floodgates?

SSST’s network was previously known as G60 Starlink, referencing a major cross-country highway in China and the project’s intent to imitate SpaceX’s broadband service.

Thousand Sails may eventually consist of more than 14,000 satellites, but like other Internet megaconstellations, the size of the fleet will likely grow at a rate commensurate with demand. It will take many years for SSST to deploy a 14,000-satellite constellation, if it ever does. SpaceX has rolled out several generations of Starlink satellites to offer new services and more capacity to meet customer uptake.

Chinese officials have released few details about the Qianfan satellites. But the project’s backers have said the spacecraft has a “standardized and modular” flat-panel design. “It meets the needs of stacking multiple satellites with one rocket,” said Shanghai Gesi Aerospace Technology, a joint venture set up by SSST and the Chinese Academy of Sciences to oversee manufacturing of Qianfan satellites.

This sounds a lot like the design of SpaceX’s Starlink satellites, which are flat-packed for launch on Falcon 9 rockets. SpaceX pioneered this way of launching and deploying large numbers of satellites. The approach used for Starlink, and apparently for Qianfan, streamlines the integration of multiple satellites with their launcher on the ground. It also simplifies their separation from the rocket once in orbit.

The new Qianfan satellite factory in Shanghai can produce up to 300 spacecraft per year, project officials said in December. Officials previously said the first 108 satellites for the Thousand Sails constellation would launch this year.

SSST announced in February it had raised more than $900 million from Chinese state-backed investment funds, Shanghai’s municipal government, and sources of venture capital. SSST’s origin is linked to a Chinese joint venture with a Germany-based company called KLEO Connect, which intended to develop a smaller constellation of low-Earth orbit satellites for data relay services.

China launched four technology demonstration satellites, purportedly related to the KLEO Connect venture, to test telecom hardware and electric propulsion systems in orbit. The joint venture fell apart with a flurry of lawsuits, and the German government last year blocked a complete takeover of KLEO Connect by its Chinese investors.

Now, SSST is going it alone with the Thousand Sails network. It has rapidly scaled up satellite manufacturing capacity in Shanghai. But outside of Starlink, companies with ideas for megaconstellations have run into serious headwinds.

OneWeb filed for bankruptcy in 2020 before eventually launching its entire first-generation network of 633 Internet satellites. Amazon has pushed back the full-scale deployment of its Project Kuiper megaconstellation, and the launch of the first operational Kuiper Internet satellites may be delayed again to 2025. The future of the European Union’s IRIS² satellite Internet network is in doubt after disagreements among European governments on funding the project.

The Thousand Sails constellation is less well-known than another planned Chinese satellite Internet network known as Guowang, or “national network,” which is supported by China’s central government. Guowang is owned by a state-backed company called SatNet, and its architecture will consist of 13,000 satellites. However, China has not yet launched any spacecraft for the Guowang project.

It’s unclear if the Thousand Sails network and the Guowang constellation will be direct competitors. They could be geared to different segments of the broadband market. In either case, China’s restrictive Internet policies with terrestrial networks will likely spill over into the satellite segment.

Chinese officials recognize the military utility of satellite Internet services like Starlink, which has supported Ukrainian military forces fighting Russian troops since 2022. A homegrown Starlink-like service would, no doubt, prove useful for China’s military.

Alongside potential domestic civilian users, China could use its satellite Internet networks as a diplomatic tool to build on existing partnerships between the Chinese government and developing countries. This could “lead to a leapfrogging moment, where African countries opt for the Chinese Internet constellation over Western providers due to the fact that much of their infrastructure is already Chinese-built,” the Royal United Services Institute, a UK think tank, wrote in a report last year.

While there are open questions about how China will use its satellite megaconstellations, their deployment will require a significant increase in the country’s launch capacity, driving the development of new commercial rockets, including reusable boosters, to lower costs and increase their flight rate.

China begins launching a megaconstellation, and it sounds a lot like Starlink Read More »

“so-tired”:-disney+,-hulu,-espn+-prices-increase-by-up-to-25-percent-in-october

“So tired”: Disney+, Hulu, ESPN+ prices increase by up to 25 percent in October

The cycle continues —

Not even ad tiers are safe as Disney looks to coax people into bundle packages.

The Doctor and Ruby in 1960s

Enlarge / A scene from the new season of Doctor Who, which is streaming on Disney+.

Disney+

Disney+, Hulu, and ESPN+ will get more expensive as of October 17, whether users have a subscription with or without ads. After most recently jacking up streaming prices in October 2023, The Walt Disney Company is raising subscription fees by as much as 25 percent, depending on the streaming service and plan.

Here’s how pricing will look in October compared to now:

Now As of October 17
Disney+ with ads $8/month $10/month
Disney+ without ads $14/month

$140/year
$16/month

$160/year
Hulu with ads $8/year

$80/year
$10/month

$100/year
Hulu without ads $18/month $19/month
Hulu and Live TV with ads $77/month  $83/month
Hulu and Live TV without ads $90/month $96/month
Disney+ and Hulu with ads $10/month $11/month
Disney+ and Hulu without ads $20/month No change
ESPN+ $11/month

$110/year
$12/month

$120/year
Disney+, Hulu, and ESPN+ with ads $15 No change
Disney+, Hulu, and ESPN+ without ads $25 No change

Disney didn’t announce any pricing changes for the bundle that contains Disney+, sister streaming service Hulu, and Warner Bros. Discovery’s rival streaming platform, Max.

Based on the updated pricing, Disney is seemingly trying to coax people to sign up for one of its streaming bundles, which combine multiple services for a lower price than if the services were each subscribed to individually. Streaming companies have been trying to use bundles to deter people from frequently canceling their streaming subscriptions. But as we’ve written before, streaming bundles don’t address subscribers’ complaints around incessant price hikes, content quality, confusing packages, or features.

Another price hike

One of the biggest problems that streaming subscribers, especially long-term ones, are facing is ever-rising prices. Disney already increased prices in October 2023, meaning Disney+, Hulu, and ESPN+ are facing their second price hikes in about a year.

Early reactions online, including on forums like Reddit, show people dissatisfied with streaming price hikes that don’t seem to align with the quality of content available. For example, Reddit user Montysucker wrote: “easy[,] cancel now,” adding:

“The enshittification of media in the last few years is insane and it’s wild how seemingly no one cares anymore about making something that is actually enjoyable to watch and not their egotistic[al] pipe dream.”

Of course, many expressed being overwhelmed with continuing to see streaming prices rise, as Slow_Investment_2211 wrote:

On October 12, 2023, as Variety summarized, Disney+ without ads went up by 27 percent, from $11 to $14 per month. Hulu without ads went up 20 percent ($15 to $18/month). Hulu with Live packages each also increased by $7 at the time, while ESPN+ pricing increased by $1.

Disney paired the price hike announcement with the revealing of new upcoming features for Disney+. However, the new linear channels are little comfort for people who don’t use Disney+.

The new channels will be ABC News Live, which Disney+ users can access on September 4, and channels “focused on preschool content, featuring TV series and shorts available on Disney+.” Disney+ will also get four more channels (or as Disney’s calling them, playlists) that show: 1) “Seasonal Content” from Disney+; 2) “Epic Stories” from big franchises like Marvel and Star Wars; 3) “Throwbacks” with “nostalgic pop culture”; and 4) “Real Life” documentaries.

It’s possible that Disney will introduce more channels to appeal to more users. But with all the price hikes that streaming subscribers have endured over the past few years, many would prefer avoiding price bumps that are partially for extra channels that they may never watch. Charging for unwanted content in media packages that are already priced questionably is reminiscent of cable, something that streaming was initially supposed to replace, not replicate.

Subscribers to Disney’s trio of streaming services will be unlikely to be alone in facing price hikes for long; analysts suspect Netflix pricing will also increase this year.

“So tired”: Disney+, Hulu, ESPN+ prices increase by up to 25 percent in October Read More »

this-ai-humanoid-robot-helped-assemble-bmws-at-us-factory

This AI humanoid robot helped assemble BMWs at US factory

why humanoid, though? —

BMW has no timeline to integrate humanoid robots into its production lines.

A humanoid robot

Enlarge / This is the Figure 02, a new humanoid robot that recently tried its hand(s) at assembling BMW chassis in South Carolina.

BMW

Robots have been working in car factories for decades now, starting with machines performing some welds on a General Motors production line back in 1961. Now, robots work alongside people on production lines, excelling at tasks like manipulating parts too heavy for humans to easily lift or welding or bonding with more precision than we can manage.

Those robots mostly look like big multi-axis arms, but a new breed of two-armed, two-legged robots is being tested in car factories. BMW is the latest automaker to try them at its factory in Spartanburg, South Carolina.

Unlike Tesla, which hopes to develop its own bipedal ‘bot to work on its production line sometime next year, BMW has brought in a robot from Figure AI. The Figure 02 robot has hands with sixteen degrees of freedom and human-equivalent strength.

“We are excited to unveil Figure 02, our second-generation humanoid robot, which recently completed successful testing at the BMW Group Plant Spartanburg. Figure 02 has significant technical advancements, which enable the robot to perform a wide range of complex tasks fully autonomously,” said Brett Adcock, founder and CEO of Figure AI.

Figure 02 assembled chassis parts.

Enlarge / Figure 02 assembled chassis parts.

BMW

BMW wanted to test how to integrate a humanoid robot into its production process—how to have the robot communicate with the production line software and human workers and determine what requirements would be necessary to add robots to the mix.

The Figure robot was given the job of inserting sheet metal parts into fixtures as part of the process of making a chassis. BMW says this required particular dexterity and that it’s an ergonomically awkward and tiring task for humans.

Now that the trial is over, Figure’s robot is no longer working at Spartanburg, and BMW says it has “no definite timetable established” to add humanoid robots to its production lines. “The developments in the field of robotics are very promising. With an early-test operation, we are now determining possible applications for humanoid robots in production. We want to accompany this technology from development to industrialization,” said Milan Nedeljković, BMW’s board member responsible for production.

This AI humanoid robot helped assemble BMWs at US factory Read More »

elon-musk-sues-openai,-sam-altman-for-making-a-“fool”-out-of-him

Elon Musk sues OpenAI, Sam Altman for making a “fool” out of him

“Altman’s long con” —

Elon Musk asks court to void Microsoft’s exclusive deal with OpenAI.

Elon Musk and Sam Altman share the stage in 2015, the same year that Musk alleged that Altman's

Enlarge / Elon Musk and Sam Altman share the stage in 2015, the same year that Musk alleged that Altman’s “deception” began.

After withdrawing his lawsuit in June for unknown reasons, Elon Musk has revived a complaint accusing OpenAI and its CEO Sam Altman of fraudulently inducing Musk to contribute $44 million in seed funding by promising that OpenAI would always open-source its technology and prioritize serving the public good over profits as a permanent nonprofit.

Instead, Musk alleged that Altman and his co-conspirators—”preying on Musk’s humanitarian concern about the existential dangers posed by artificial intelligence”—always intended to “betray” these promises in pursuit of personal gains.

As OpenAI’s technology advanced toward artificial general intelligence (AGI) and strove to surpass human capabilities, “Altman set the bait and hooked Musk with sham altruism then flipped the script as the non-profit’s technology approached AGI and profits neared, mobilizing Defendants to turn OpenAI, Inc. into their personal piggy bank and OpenAI into a moneymaking bonanza, worth billions,” Musk’s complaint said.

Where Musk saw OpenAI as his chance to fund a meaningful rival to stop Google from controlling the most powerful AI, Altman and others “wished to launch a competitor to Google” and allegedly deceived Musk to do it. According to Musk:

The idea Altman sold Musk was that a non-profit, funded and backed by Musk, would attract world-class scientists, conduct leading AI research and development, and, as a meaningful counterweight to Google’s DeepMind in the race for Artificial General Intelligence (“AGI”), decentralize its technology by making it open source. Altman assured Musk that the non-profit structure guaranteed neutrality and a focus on safety and openness for the benefit of humanity, not shareholder value. But as it turns out, this was all hot-air philanthropy—the hook for Altman’s long con.

Without Musk’s involvement and funding during OpenAI’s “first five critical years,” Musk’s complaint said, “it is fair to say” that “there would have been no OpenAI.” And when Altman and others repeatedly approached Musk with plans to shift OpenAI to a for-profit model, Musk held strong to his morals, conditioning his ongoing contributions on OpenAI remaining a nonprofit and its tech largely remaining open source.

“Either go do something on your own or continue with OpenAI as a nonprofit,” Musk told Altman in 2018 when Altman tried to “recast the nonprofit as a moneymaking endeavor to bring in shareholders, sell equity, and raise capital.”

“I will no longer fund OpenAI until you have made a firm commitment to stay, or I’m just being a fool who is essentially providing free funding to a startup,” Musk said at the time. “Discussions are over.”

But discussions weren’t over. And now Musk seemingly does feel like a fool after OpenAI exclusively licensed GPT-4 and all “pre-AGI” technology to Microsoft in 2023, while putting up paywalls and “failing to publicly disclose the non-profit’s research and development, including details on GPT-4, GPT-4T, and GPT-4o’s architecture, hardware, training method, and training computation.” This excluded the public “from open usage of GPT-4 and related technology to advance Defendants and Microsoft’s own commercial interests,” Musk alleged.

Now Musk has revived his suit against OpenAI, asking the court to award maximum damages for OpenAI’s alleged fraud, contract breaches, false advertising, acts viewed as unfair to competition, and other violations.

He has also asked the court to determine a very technical question: whether OpenAI’s most recent models should be considered AGI and therefore Microsoft’s license voided. That’s the only way to ensure that a private corporation isn’t controlling OpenAI’s AGI models, which Musk repeatedly conditioned his financial contributions upon preventing.

“Musk contributed considerable money and resources to launch and sustain OpenAI, Inc., which was done on the condition that the endeavor would be and remain a non-profit devoted to openly sharing its technology with the public and avoid concentrating its power in the hands of the few,” Musk’s complaint said. “Defendants knowingly and repeatedly accepted Musk’s contributions in order to develop AGI, with no intention of honoring those conditions once AGI was in reach. Case in point: GPT-4, GPT-4T, and GPT-4o are all closed source and shrouded in secrecy, while Defendants actively work to transform the non-profit into a thoroughly commercial business.”

Musk wants Microsoft’s GPT-4 license voided

Musk also asked the court to null and void OpenAI’s exclusive license to Microsoft, or else determine “whether GPT-4, GPT-4T, GPT-4o, and other OpenAI next generation large language models constitute AGI and are thus excluded from Microsoft’s license.”

It’s clear that Musk considers these models to be AGI, and he’s alleged that Altman’s current control of OpenAI’s Board—after firing dissidents in 2023 whom Musk claimed tried to get Altman ousted for prioritizing profits over AI safety—gives Altman the power to obscure when OpenAI’s models constitute AGI.

Elon Musk sues OpenAI, Sam Altman for making a “fool” out of him Read More »

gamification-gets-drivers-to-put-down-their-phones,-study-finds

Gamification gets drivers to put down their phones, study finds

use bluetooth —

Gamification plus cash prizes worked even better.

Close-up Of A Man's Hand Typing Text Message On Mobile Phone While Driving Car

Enlarge / You should not do this while you are driving.

Getty Images

Distracted driving isn’t only a result of drivers using their phones when they should be paying attention. But it is a significant cause of the problem, accounting for at least 13 percent of distracted driving deaths and rising to 1 in 5 for young drivers. Now, a study conducted with customers of the Progressive insurance company has tested different strategies to get those drivers to put their phones down in the car, and it found two that significantly reduced handheld use, with the effect persisting after the end of the study.

The study recruited 1,653 customers already enrolled in its Progressive Snapshot program, which involves the use of a smartphone app that detects phone use while driving. Before the start of the trial, the participants all averaged more than 6.4 minutes per hour of handheld use while driving—Progressive says its safest customers have handheld usage of less than 1 minute per hour while driving.

Five test groups

The drivers were split into five different arms, each with increasing amounts of intervention. The first group just received education about the problem, such as statistics about state laws that ban phone use while driving, increased crash risks, and recommendations to use hands-free options like a phone mount or casting interface instead.

The second arm received the educational materials and a free phone mount with the message “Driving? Park your phone here.” The authors were particularly interested to see whether this arm worked, given the relatively low-cost and one-time nature of this intervention.

The third arm received both educational materials and the mount, but the participants were also asked to commit in writing to reducing their phone use while driving. The researchers informed these participants about their baseline phone use while driving and then gave them weekly goals to reduce their phone use to below 1 minute per hour over the course of the 10-week study. This arm also got regular tips on helpful habits, like mindfulness reminders, encouraging prompts, or using a phone’s “do not disturb” mode while driving.

Arm four included all the same interventions as arm three but added competitive gamification as well. Each Monday evening during the 10-week trial, the participants in arm four received a message with their handheld usage goal for the coming week and whether they met the previous week’s goal. Points were awarded for meeting the goal and were taken away for backsliding. Additionally, participants with similar baseline usage were grouped together in tens to form leaderboards so individuals could compete against (anonymized) peers.

Finally, arm five included the same interventions as arm four and added a financial incentive. Participants who scored enough points over the course of the 10-week study shared a $2,000 prize, each taking home $15.63 in the end. This arm was also told they would earn $5 for finishing at the head of the weekly leaderboard, which displayed a “total winnings” column in this arm.

What worked?

Perhaps unsurprisingly, none of the interventions in groups one, two, or three resulted in those participants significantly reducing their handheld use while driving. But the drivers in arm four had a 20.5 percent reduction in the amount of handheld usage while driving during the study.

What’s more, this effect persisted throughout the rest of the Progressive Snapshot period (a variable-length post-intervention period lasting 25–65 days) following the end of the 10-week study, with this arm still showing 16.2 percent less handheld usage compared to the control.

Paying drivers on top of competitive gamification was the most effective way to get them to put down their phones. This group reduced its handheld usage by 27.6 percent, or 89 seconds/hour, compared to the control. That reduction was maintained at the same level throughout the post-intervention period for this arm.

Gamification has already been adopted to varying degrees of success by automakers and their UI designers to try to encourage hybrid and EV drivers to be more efficient. So it’s not entirely surprising to see the same approach can work to modify other types of driving behavior. It’s even possible that insurance companies could start financially incentivizing drivers to behave better, assuming prizes cost less than the amount saved by paying for fewer crashes and claims.

PNAS, 2024. DOI: 10.1073/pnas.2320603121  (About DOIs).

Gamification gets drivers to put down their phones, study finds Read More »