Author name: Tim Belzer

what-if-ai-doesn’t-just-keep-getting-better-forever?

What if AI doesn’t just keep getting better forever?

AI / /

For years now, many AI industry watchers have looked at the quickly growing capabilities of new AI models and mused about exponential performance increases continuing well into the future. Recently, though, some of that AI “scaling law” optimism has been replaced by fears that we may already be hitting a plateau in the capabilities of large language models trained with standard methods.

A weekend report from The Information effectively summarized how these fears are manifesting amid a number of insiders at OpenAI. Unnamed OpenAI researchers told The Information that Orion, the company’s codename for its next full-fledged model release, is showing a smaller performance jump than the one seen between GPT-3 and GPT-4 in recent years. On certain tasks, in fact, the upcoming model “isn’t reliably better than its predecessor,” according to unnamed OpenAI researchers cited in the piece.

On Monday, OpenAI co-founder Ilya Sutskever, who left the company earlier this year, added to the concerns that LLMs were hitting a plateau in what can be gained from traditional pre-training. Sutskever told Reuters that “the 2010s were the age of scaling,” where throwing additional computing resources and training data at the same basic training methods could lead to impressive improvements in subsequent models.

“Now we’re back in the age of wonder and discovery once again,” Sutskever told Reuters. “Everyone is looking for the next thing. Scaling the right thing matters more now than ever.”

What’s next?

A large part of the training problem, according to experts and insiders cited in these and other pieces, is a lack of new, quality textual data for new LLMs to train on. At this point, model makers may have already picked the lowest hanging fruit from the vast troves of text available on the public Internet and published books.

What if AI doesn’t just keep getting better forever? Read More »

amazon-ready-to-use-its-own-ai-chips,-reduce-its-dependence-on-nvidia

Amazon ready to use its own AI chips, reduce its dependence on Nvidia

AI, Amazon, CPUs, datacenter CPUs, syndication, Trainium 2 / /

Amazon now expects around $75 billion in capital spending in 2024, with the majority on technology infrastructure. On the company’s latest earnings call, chief executive Andy Jassy said he expects the company will spend even more in 2025.

This represents a surge on 2023, when it spent $48.4 billion for the whole year. The biggest cloud providers, including Microsoft and Google, are all engaged in an AI spending spree that shows little sign of abating.

Amazon, Microsoft, and Meta are all big customers of Nvidia, but are also designing their own data center chips to lay the foundations for what they hope will be a wave of AI growth.

“Every one of the big cloud providers is feverishly moving towards a more verticalized and, if possible, homogenized and integrated [chip technology] stack,” said Daniel Newman at The Futurum Group.

“Everybody from OpenAI to Apple is looking to build their own chips,” noted Newman, as they seek “lower production cost, higher margins, greater availability, and more control.”

“It’s not [just] about the chip, it’s about the full system,” said Rami Sinno, Annapurna’s director of engineering and a veteran of SoftBank’s Arm and Intel.

For Amazon’s AI infrastructure, that means building everything from the ground up, from the silicon wafer to the server racks they fit into, all of it underpinned by Amazon’s proprietary software and architecture. “It’s really hard to do what we do at scale. Not too many companies can,” said Sinno.

After starting out building a security chip for AWS called Nitro, Annapurna has since developed several generations of Graviton, its Arm-based central processing units that provide a low-power alternative to the traditional server workhorses provided by Intel or AMD.

Amazon ready to use its own AI chips, reduce its dependence on Nvidia Read More »

there-are-some-things-the-crew-8-astronauts-aren’t-ready-to-talk-about

There are some things the Crew-8 astronauts aren’t ready to talk about

astronauts, Crew-8, human spaceflight, international space station, NASA, Science, Space, spacex / /

“I did not say I was uncomfortable talking about it. I said we’re not going to talk about it.”

NASA astronaut Michael Barratt works with a spacesuit inside the Quest airlock of the International Space Station on May 31. Credit: NASA

The astronauts who came home from the International Space Station last month experienced some drama on the high frontier, and some of it accompanied them back to Earth.

In orbit, the astronauts aborted two spacewalks, both under unusual circumstances. Then, on October 25, one of the astronauts was hospitalized due to what NASA called an unspecified “medical issue” after splashdown aboard a SpaceX Crew Dragon capsule that concluded the 235-day mission. After an overnight stay in a hospital in Florida, NASA said the astronaut was released “in good health” and returned to their home base in Houston to resume normal post-flight activities.

The space agency did not identify the astronaut or any details about their condition, citing medical privacy concerns. The three NASA astronauts on the Dragon spacecraft included commander Matthew Dominick, pilot Michael Barratt, and mission specialist Jeanette Epps. Russian cosmonaut Alexander Grebenkin accompanied the three NASA crew members. Russia’s space agency confirmed he was not hospitalized after returning to Earth.

Dominick, Barratt, and Epps answered media questions in a post-flight press conference Friday, but they did not offer more information on the medical issue or say who experienced it. NASA initially sent all four crew members to the hospital in Pensacola, Florida, for evaluation, but Grebenkin and two of the NASA astronauts were quickly released and cleared to return to Houston. One astronaut remained behind until the next day.

“Spaceflight is still something we don’t fully understand,” said Barratt, a medical doctor and flight surgeon. “We’re finding things that we don’t expect sometimes. This was one of those times, and we’re still piecing things together on this, and so to maintain medical privacy and to let our processes go forward in an orderly manner, this is all we’re going to say about that event at this time.”

NASA typically makes astronaut health data available to outside researchers, who regularly publish papers while withholding identifying information about crew members. NASA officials often tout gaining knowledge about the human body’s response to spaceflight as one of the main purposes of the International Space Station. The agency is subject to federal laws, including the Health Insurance Portability and Accountability Act (HIPAA) of 1996, restricting the release of private medical information.

“I did not say I was uncomfortable talking about it,” Barratt said. “I said we’re not going to talk about it. I’m a medical doctor. Space medicine is my passion … and how we adapt, how we experience human spaceflight is something that we all take very seriously.”

Maybe some day

Barratt said NASA will release more information about the astronaut’s post-flight medical issue “in the fullness of time.” This was Barratt’s third trip to space and the first spaceflight for Dominick and Epps.

One of the most famous incidents involving hospitalized astronauts was in 1975, before the passage of the HIPAA medical privacy law, when NASA astronauts Thomas Stafford, Deke Slayton, and Vance Brand stayed at a military hospital in Hawaii nearly two weeks after inhaling toxic propellant fumes that accidentally entered their spacecraft’s internal cabin as it descended under parachutes. They were returning to Earth at the end of the Apollo-Soyuz mission, in which they docked their Apollo command module to a Soviet Soyuz spacecraft in orbit.

NASA’s view—and perhaps the public’s, too—of medical privacy has changed in the nearly 50 years since. On that occasion, NASA disclosed that the astronauts suffered from lung irritation, and officials said Brand briefly passed out from the fumes after splashdown, remaining unconscious until his crewmates fitted an oxygen mask tightly over his face. NASA and the military also made doctors available to answer media questions about their condition.

The medical concern after splashdown last month was not the only part of the Crew-8 mission that remains shrouded in mystery. Dominick and NASA astronaut Tracy Dyson were supposed to go outside the International Space Station for a spacewalk June 13, but NASA called off the excursion, citing a “spacesuit discomfort issue.” NASA replaced Dominick with Barratt and rescheduled the spacewalk for June 24 to retrieve a faulty electronics box and collect microbial samples from the exterior of the space station. But that excursion ended after just 31 minutes, when Dyson reported a water leak in the service and cooling umbilical unit of her spacesuit.

While Barratt discussed the water leak in some detail Friday, Dominick declined to answer a question from Ars regarding the suit discomfort issue. “We’re still reviewing and trying to figure all the details,” he said.

Aging suits

Regarding the water leak, Barratt said he and Dyson noticed her suit had a “spewing umbilical, which was quite dramatic, actually.” The decision to abandon the spacewalk was a “no-brainer,” he said.

“It was not a trivial leak, and we’ve got footage,” Barratt said. “Anybody who was watching NASA TV at the time could see there was basically a snowstorm, a blizzard, spewing from the airlock because we already had the hatch open. So we were seeing flakes of ice in the airlock, and Tracy was seeing a lot of them on her helmet, on her gloves, and whatnot. Dramatic is the right word, to be real honest.”

Dyson, who came back to Earth in September on a Russian Soyuz spacecraft, reconnected the leaking umbilical with her gloves and helmet covered with ice, with restricted vision. “Tracy’s actions were nowhere short of heroic,” Barratt said.

Once the leak stabilized, the astronauts closed the hatch and began repressurizing the airlock.

“Getting the airlock closed was kind of me grabbing her legs and using her as an end effector to lever that thing closed, and she just made it happen,” Barratt said. “So, yeah,  there was this drama. Everything worked out fine. Again, normal processes and procedures saved our bacon.”

Barratt said the leak wasn’t caused by any procedural error as the astronauts prepared their suits for the spacewalk.

“It was definitely a hardware issue,” he said. “There was a little poppet valve on the interface that didn’t quite seat, so really, the question became why didn’t that seat? We solved that problem by changing out the whole umbilical.”

By then, NASA’s attention on the space station had turned to other tasks, such as experiments, the arrival of a new cargo ship, and testing of Boeing’s Starliner crew capsule docked at the complex, before it ultimately departed and left its crew behind. The spacewalk wasn’t urgent, so it had to wait. NASA now plans to attempt the spacewalk again as soon as January with a different set of astronauts.

Barratt thinks the spacesuits on the space station are good to go for the next spacewalk. However, the suits are decades old, and their original designs date back more than 40 years, when NASA developed the units for use on the space shuttle. Efforts to develop a replacement suit for use in low-Earth orbit have stalled. In June, Collins Aerospace dropped out of a NASA contract to build new spacesuits for servicing the International Space Station and future orbiting research outposts.

“None of our spacesuits are spring chickens, so we will expect to see some hardware issues with repeated use and not really upgrading,” Barratt said.

Photo of Stephen Clark

Stephen Clark is a space reporter at Ars Technica, covering private space companies and the world’s space agencies. Stephen writes about the nexus of technology, science, policy, and business on and off the planet.

There are some things the Crew-8 astronauts aren’t ready to talk about Read More »

ftx-sues-binance-for-$1.76b-in-battle-of-crypto-exchanges-founded-by-convicts

FTX sues Binance for $1.76B in battle of crypto exchanges founded by convicts

Binance, FTX, Policy / /

Lawsuit seeks “at least $1.76 billion that was fraudulently transferred” by SBF.

Former Binance CEO Changpeng Zhao arrives at federal court in Seattle for sentencing on Tuesday, April 30, 2024. Credit: Getty Images | Changpeng Zhao

The bankruptcy estate of collapsed cryptocurrency exchange FTX has sued the company’s former rival Binance in an attempt to recover $1.76 billion or more. The lawsuit seeks “at least $1.76 billion that was fraudulently transferred to Binance and its executives at the FTX creditors’ expense, as well as compensatory and punitive damages to be determined at trial.”

The complaint filed yesterday in US Bankruptcy Court in Delaware names Binance and co-founder and former CEO Changpeng Zhao among the defendants. FTX founder Sam Bankman-Fried sold 20 percent of his crypto exchange to Binance in November 2019, but Binance exited that investment in 2021, the lawsuit said.

“As Zhao would later remark, he decided to exit his position in FTX because of personal grievances he had against Bankman-Fried,” the lawsuit said. “In July 2021, the parties negotiated a deal whereby FTX bought back Binance’s and its executives’ entire stakes in both FTX Trading and [parent company] WRS. Pursuant to that deal, FTX’s Alameda Research division directly funded the share repurchase with a combination of FTT (FTX’s exchange token), BNB (Binance’s exchange token), and BUSD (Binance’s dollar-pegged stablecoin). In the aggregate, those tokens had a fair market value of at least $1.76 billion.”

Because FTX and Alameda were balance-sheet insolvent by early 2021, the $1.76 billion transfer “was a constructive fraudulent transfer based on a straightforward application” of bankruptcy law, and an intentional fraudulent transfer “because the transfer was made in furtherance of Bankman-Fried’s scheme,” the lawsuit said.

Alameda could not fund the transaction because of its insolvency, the lawsuit said. “Indeed, as Bankman-Fried’s second-in-command, Caroline Ellison, would later testify, she contemporaneously told Bankman-Fried ‘we don’t really have the money for this, we’ll have to borrow from FTX to do it,'” the lawsuit said.

The complaint alleges that after the 2021 divestment, Zhao “set out to destroy” FTX, and accuses Binance and Zhao of fraud, injurious falsehood, intentional misrepresentation, and unjust enrichment.

Binance is far from the only entity being sued by FTX. The firm filed 23 lawsuits in the bankruptcy court on Friday “as part of a broader effort to claw back money for creditors of the bankrupt company,” Bloomberg reported. Defendants in other suits include Anthony Scaramucci and his hedge fund SkyBridge Capital, Crypto.com, and the Mark Zuckerberg-founded FWD.US.

Lawsuit cites SBF’s false statements

Ellison, who was sentenced to two years in prison, testified that Alameda funded the repurchase with about $1 billion of FTX Trading capital received from depositors, the lawsuit said. It continued:

Ellison further testified that Bankman-Fried dismissed her concerns about financial resources, telling her that, notwithstanding the need to use customer deposits, the repurchase was “really important, we have to get it done.” Indeed, as discussed below, one of the reasons Bankman-Fried viewed the transaction as “really important” was precisely because of his desire to conceal his companies’ insolvency and send a false signal of strength to the market. In connection with the share repurchase, Bankman-Fried was asked directly by a reporter whether Alameda funded the entire repurchase using its own assets, expressing surprise that Alameda could have done so given the purchase price and what was publicly known regarding Alameda’s financial resources. In response, Bankman-Fried falsely stated: “The purchase was entirely from Alameda. Yeah, it had a good last year :P” (i.e., an emoji for a tongue sticking out).

The transaction contributed to FTX’s downfall, according to the lawsuit. It “left the platform in an even greater imbalance, which Bankman-Fried attempted to cover up in a pervasive fraud that infected virtually all aspects of FTX’s business,” FTX’s complaint said. Bankman-Fried is serving a 25-year prison sentence.

Because FTX trading was insolvent in July 2021 when the Binance share repurchase was completed, “the FTX Trading shares acquired through the share repurchase were actually worthless based on a proper accounting of FTX Trading’s assets and liabilities,” the lawsuit said.

Zhao allegedly “set out to destroy”

FTX claims that once Zhao divested himself of the equity stake in FTX, “Zhao then set out to destroy his now-unaffiliated competitor” because FTX was “a clear threat to Binance’s market dominance.” Zhao resigned from Binance last year after agreeing to plead guilty to money laundering violations and was sentenced to four months in prison. He was released in September.

FTX’s lawsuit alleges that “Zhao’s succeed-at-all-costs business ethos was not limited to facilitating money laundering. Beginning on November 6, 2022, Zhao sent a series of false, misleading, and fraudulent tweets that were maliciously calculated to destroy his rival FTX, with reckless disregard to the harm that FTX’s customers and creditors would suffer. As set forth herein in more detail, Zhao’s false tweets triggered a predictable avalanche of withdrawals at FTX—the proverbial run on the bank that Zhao knew would cause FTX to collapse.”

Zhao’s tweet thread said Binance liquidated its remaining FTT “due to recent revelations.” The lawsuit alleges that “contrary to Zhao’s denial, Binance’s highly publicized apparent liquidation of its FTT was indeed a ‘move against a competitor’ and was not, as Zhao indicated, ‘due to recent revelations.'”

“As Ellison testified, ‘if [Zhao] really wanted to sell his FTT, he wouldn’t preannounce to the market that he was going to sell it. He would just sell it […] his real aim in that tweet, as I saw it, was not to sell his FTT, but to hurt FTX and Alameda,'” the lawsuit said.

The lawsuit further claims that while FTX was “in freefall, Zhao sent additional false tweets calculated, in part, to prevent FTX from seeking and obtaining alternative financing to cauterize the run on the institution by customers deceived by the tweets. Collectively and individually, these false public statements destroyed value that would have otherwise been recoverable by FTX’s stakeholders.”

Binance calls lawsuit “meritless”

On November 8, 2022, Bankman-Fried and Zhao agreed to a deal in which “Binance would acquire FTX Trading and inject capital sufficient to address FTX’s liquidity issues,” the lawsuit said. But the next day, Binance published tweets saying it was backing out of the deal “as a result of corporate due diligence.”

When Zhao agreed to the deal on November 8, he had “already been made aware of the ‘mishandled’ customer funds during his conversation with Bankman-Fried,” the lawsuit said. “This is contrary to Binance’s representation in the November 9 Tweets that he learned that fact after entering into the Letter of Intent. In addition, Zhao was also aware that the Debtors were insolvent when he entered into the Letter of Intent.”

In the 24 hours between the November 8 agreement and the November 9 tweets, “no new material information was provided to Zhao and Binance in the diligence process that would have revealed new issues” causing Binance to exit the deal, according to the lawsuit.

Binance said it will fight FTX’s lawsuit. “The claims are meritless, and we will vigorously defend ourselves,” a Binance spokesperson said in a statement provided to Ars.

The defendants also included “Does 1-1,000,” people who allegedly received fraudulent transfers in 2021 and “whose true names, identities and capacities are presently unknown to the Plaintiffs.” FTX is seeking recovery of fraudulent transfers from all defendants. FTX also asked the court to award punitive damages and find that Binance and Zhao committed fraud, injurious falsehood, intentional misrepresentation, and unjust enrichment.

Photo of Jon Brodkin

Jon is a Senior IT Reporter for Ars Technica. He covers the telecom industry, Federal Communications Commission rulemakings, broadband consumer affairs, court cases, and government regulation of the tech industry.

FTX sues Binance for $1.76B in battle of crypto exchanges founded by convicts Read More »

russia:-fine,-i-guess-we-should-have-a-grasshopper-rocket-project,-too

Russia: Fine, I guess we should have a Grasshopper rocket project, too

Science, Space, Uncategorized / /

Like a lot of competitors in the global launch industry, Russia for a long time dismissed the prospects of a reusable first stage for a rocket.

As late as 2016, an official with the Russian agency that develops strategy for the country’s main space corporation, Roscosmos, concluded, “The economic feasibility of reusable launch systems is not obvious.” In the dismissal of the landing prospects of SpaceX’s Falcon 9 rocket, Russian officials were not alone. Throughout the 2010s, competitors including space agencies in Europe and Japan, and US-based United Launch Alliance, all decided to develop expendable rockets.

However, by 2017, when SpaceX re-flew a Falcon 9 rocket for the first time, the writing was on the wall. “This is a very important step, we sincerely congratulate our colleague on this achievement,” then-Roscosmos CEO Igor Komarov said at the time. He even spoke of developing reusable components, such as rocket engines capable of multiple firings.

A Russian Grasshopper

That was more than seven years ago, however, and not much has happened in Russia since then to foster the development of a reusable rocket vehicle. Yes, Roscosmos unveiled plans for the “Amur” rocket in 2020, which was intended to have a reusable first stage and methane-fueled engines and land like the Falcon 9. But its debut has slipped year for year—originally intended to fly in 2026, its first launch is now expected no earlier than 2030.

Now, however, there is some interesting news from Moscow about plans to develop a prototype vehicle to test the ability to land the Amur rocket’s first stage vertically.

According to the state-run news agency, TASS, construction of this test vehicle will enable the space corporation to solve key challenges. “Next year preparation of an experimental stage of the (Amur) rocket, which everyone is calling ‘Grasshopper,’ will begin,” said Igor Pshenichnikov, the Roscosmos deputy director of the department of future programs. The Russian news article was translated for Ars by Rob Mitchell.

Russia: Fine, I guess we should have a Grasshopper rocket project, too Read More »

review:-catching-up-with-the-witchy-brew-of-agatha-all-along

Review: Catching up with the witchy brew of Agatha All Along

Agatha All Along, culture, Disney Plus, Entertainment, marvel studios, MCU Phase Five, streaming television, TV reviews / /

Down, down, down the road

Spoilers ahead! This imaginative sequel to WandaVision is a reminder of just how good the MCU can be.

Kathryn Hahn stars as Agatha Harkness, reprising her WandaVision role. Credit: Disney+

The MCU’s foray into streaming television has produced mixed results, but one of my favorites was the weirdly inventive, oh-so-meta WandaVision. I’m happy to report that the spinoff sequel,  Agatha All Along, taps into that same offbeat creativity, giving us a welcome reminder of just how good the MCU can be when it’s firing on all storytelling cylinders.

(Spoilers below, including for WandaVision and Multiverse of Madness. We’ll give you another heads up when major spoilers for Agatha All Along are imminent.)

The true identity of nosy next-door neighbor Agnes—played to perfection by Kathryn Hahn—was the big reveal of 2021’s WandaVision, even inspiring a jingle that went viral. Agnes turned out to be a powerful witch named Agatha Harkness, who had studied magic for centuries and was just dying to learn the source of Wanda’s incredible power. Wanda’s natural abilities were magnified by the Mind Stone, but Agatha realized that Wanda was a wielder of “chaos magic.” She was, in fact, the Scarlet Witch. In the finale, Wanda trapped Agatha in her nosy neighbor persona while releasing the rest of the town of Westview from her grief-driven Hex.

Then Wanda presumably died in Doctor Strange and the Multiverse of Madness (and count me among those who thought her arc in that film was a massive fail on Marvel’s part). What happened to Agatha? It seems the hex is still in place but went a bit wonky. Agatha All Along opens like a true crime serial (cf. Mare of Easttown) with Agatha/Agnes as the rebellious, socially challenged tough detective called to investigate a body found in the woods outside Westview. Then a young Teen (Joe Locke) breaks the hex and asks her to show him the way to the legendary Witches’ Road, a journey involving a series of trials. The reward: at the end of the road, the surviving witches get what they most desire. Agatha wants her powers back and Teen—well, his motives are murkier, as is his identity, which is guarded by a sigil.

Agatha and Teen first have to assemble a coven: Lilia (Patti LuPone), a divination witch; Jennifer (Sasheer Zamata), a potions witch; Alice (Ali Ahn), a protection witch; and Sharon Davis (Debra Jo Rupp, reprising her WandaVision role), standing in for a green witch on account of her gardening skills. They sing the spell in the form of a ballad—”Down the Witches’ Road,” a killer earworm that recurs throughout the series and is already spawning lots of cover versions. The entrance appears and the journey begins. As if the Witches’ Road weren’t dangerous enough, Agatha is also being pursued by her ex, Rio Vidal (Aubrey Plaza), a powerful green witch, as well as the Salem Seven, vengeful wraiths of Agatha’s first coven, who (we learned in a WandaVision flashback) she killed by draining their powers when they attacked her.

Trapped in a reality-warping spell, Agatha is apparently a detective now. YouTube/Marvel Studios

A large part of WandaVision‘s delight came from the various sitcom styles featured in each episode. Agatha All Along has its own take on that approach: each trial takes on the setting and style of witches from popular culture (even the ending credits play on this). One evokes the New England WASP-y style of the 1998 film Practical Magic; another plays on Stevie Nicks’ Bohemian “white witch” phase with elements of the 1972 film Season of the Witch; yet another trial dresses the coven in high school summer camp 1980s garb.

There are nods to the Wicked Witch of the West and Glinda from the Wizard of Oz, Malificent, and the hag version of Snow White’s Evil Queen in the seventh episode, “Death’s Hand in Mine.” It might just be the best single episode of all the Marvel series. This is Lilia’s trial, requiring her to use her divination skills to navigate a deadly tarot reading. Every wrong card releases one of the many swords suspended above the table.

Throughout the journey, Lilia has uttered seemingly random nonsensical things. Here we learn this is because she experiences life out of temporal sequence, moving between past and present while peering into the future. Suddenly all those earlier sprinkled breadcrumbs make sense, a testament to the skillful writing and directing—not to mention LuPone’s powerful performance. (Apparently she requested a script with the events in linear order to better evoke the necessary emotions when shooting scenes out of sequence.)

To glory at the end

(WARNING: Major spoilers below. Stop reading now if you haven’t finished the series.) 

By this time the coven has already lost two members: Sharon Davis (who didn’t even last the first trial), replaced by Rio; and Alice, who tried to help Agatha when the latter was briefly possessed during a ouija board trial—only to have Agatha do what she always does and drain Alice of all her power. Lilia’s tarot reading reveals that Death has been traveling with them all along in the form of Rio. Yes, Agatha’s ex is Death, aka “the original Green Witch.” They end up losing Lilia, too; she sacrifices herself to take out the Salem Seven after letting the surviving coven members escape. We see her falling to her death and then show up as a child in her homeland for her very first divination lesson—the cycle of life and death come full circle.

Agatha likes her new look for this trial. Marvel/Disney+

We soon discover that Rio/Death is mostly there because of Teen. There was much fan speculation about his identity in the run-up to the series release and fans guessed correctly: it’s Wanda and Vision’s son, Billy Maximoff, whose soul found its way into the body of a dying teenager named William Kaplan just as Wanda’s hex was unraveling him and his twin, Tommy, out of existence. That’s why he went on the Witches’ Road: to find Tommy. But this also makes him an aberration in Death’s eyes that must be removed to restore the balance. The catch: Billy has to sacrifice himself; in this unusual case, Death cannot simply take him.

Agatha initially agrees to manipulate Billy into doing just that, then has a last-minute change of heart. She kisses Rio/Death and thereby embraces her fate, sacrificing herself so Billy can live. From the start she had a soft spot for the teen, accompanied by references to her long-dead son. The backstory is quite moving and key to Agatha’s unexpected change of heart. Her son’s fate was revealed in the finale. Death came for him when Agatha was in labor but agreed to grant her “time.” How much time? Six or seven years, during which mother and son bonded and wandered from village to village, with Agatha occasionally killing more covens to absorb their power. But Death did not forget, and with Nicky (Abel Lysenko) gone, Agatha indulged all her worst impulses.

Which brings us to the Big Twist: Agatha and her son made up the ballad of the Witches’ Road, singing it in local taverns and slowly building up the legend. The Witches’ Road never existed. Agatha used the legend over centuries to lure witches into a trap to steal their powers. That was her intention at the start of the series, too, except this time—a portal opened. Billy, it seems, inherited Wanda’s ability to warp and shape reality, even subconsciously. He wanted the road to be real and so it was.

The reveal is skillfully done and ties everything up in a nice satisfying bow, with one exception. The writers just couldn’t let Agatha go completely; she returns as a ghost and joins Billy on his search for his brother Tommy. That’s a creative choice that leaves the door open for a second season, and I strongly suspect we’ll get one. But Ghost Agatha will be a tough plot point to crack. And it rather undercuts the pivotal moment of Agatha’s sacrifice—actually doing something that doesn’t directly benefit herself. On the whole, though, Agatha All Along is marvelously entertaining, binge-able fun with just enough emotional resonance and heartbreak to add some depth.

All episodes of Agatha All Along are now streaming on Disney+.

Photo of Jennifer Ouellette

Jennifer is a senior reporter at Ars Technica with a particular focus on where science meets culture, covering everything from physics and related interdisciplinary topics to her favorite films and TV series. Jennifer lives in Baltimore with her spouse, physicist Sean M. Carroll, and their two cats, Ariel and Caliban.

Review: Catching up with the witchy brew of Agatha All Along Read More »

claude-ai-to-process-secret-government-data-through-new-palantir-deal

Claude AI to process secret government data through new Palantir deal

AI, AI ethics, Amazon, Anthropic, Biz & IT, chatgpt, chatgtp, Claude, Claude 3, Claude 3.5, confabulation, ethical ai, futurism, government contracts, large language models, LLaMA, machine learning, Meta, military AI, openai, palantir, US Defense Department, US intelligence agencies, Victor Tangermann / /

An ethical minefield

Since its founders started Anthropic in 2021, the company has marketed itself as one that takes an ethics- and safety-focused approach to AI development. The company differentiates itself from competitors like OpenAI by adopting what it calls responsible development practices and self-imposed ethical constraints on its models, such as its “Constitutional AI” system.

As Futurism points out, this new defense partnership appears to conflict with Anthropic’s public “good guy” persona, and pro-AI pundits on social media are noticing. Frequent AI commentator Nabeel S. Qureshi wrote on X, “Imagine telling the safety-concerned, effective altruist founders of Anthropic in 2021 that a mere three years after founding the company, they’d be signing partnerships to deploy their ~AGI model straight to the military frontlines.“

Anthropic's

Anthropic’s “Constitutional AI” logo.

Credit: Anthropic / Benj Edwards

Anthropic’s “Constitutional AI” logo. Credit: Anthropic / Benj Edwards

Aside from the implications of working with defense and intelligence agencies, the deal connects Anthropic with Palantir, a controversial company which recently won a $480 million contract to develop an AI-powered target identification system called Maven Smart System for the US Army. Project Maven has sparked criticism within the tech sector over military applications of AI technology.

It’s worth noting that Anthropic’s terms of service do outline specific rules and limitations for government use. These terms permit activities like foreign intelligence analysis and identifying covert influence campaigns, while prohibiting uses such as disinformation, weapons development, censorship, and domestic surveillance. Government agencies that maintain regular communication with Anthropic about their use of Claude may receive broader permissions to use the AI models.

Even if Claude is never used to target a human or as part of a weapons system, other issues remain. While its Claude models are highly regarded in the AI community, they (like all LLMs) have the tendency to confabulate, potentially generating incorrect information in a way that is difficult to detect.

That’s a huge potential problem that could impact Claude’s effectiveness with secret government data, and that fact, along with the other associations, has Futurism’s Victor Tangermann worried. As he puts it, “It’s a disconcerting partnership that sets up the AI industry’s growing ties with the US military-industrial complex, a worrying trend that should raise all kinds of alarm bells given the tech’s many inherent flaws—and even more so when lives could be at stake.”

Claude AI to process secret government data through new Palantir deal Read More »

space-policy-is-about-to-get-pretty-wild,-y’all

Space policy is about to get pretty wild, y’all

Features, Science, Space / /

Saddle up, space cowboys. It may get bumpy for a while.

President Donald Trump steps on the stage at Kennedy Space Center after the successful launch of the Demo-2 crew mission in May 2020. Credit: NASA/Bill Ingalls

The global space community awoke to a new reality on Wednesday morning.

The founder of this century’s most innovative space company, Elon Musk, successfully used his fortune, time, and energy to help elect Donald Trump to president of the United States. Already, Musk was the dominant Western player in space. SpaceX launches national security satellites and NASA astronauts and operates a megaconstellation. He controls the machines that provide essential space services to NASA and the US military. And now, thanks to his gamble on backing Trump, Musk has strong-armed himself into Trump’s inner circle.

Although he may not have a cabinet-appointed position, Musk will have a broad portfolio in the new administration for as long as his relations with Trump remain positive. This gives Musk extraordinary power over a number of areas, including spaceflight. Already this week, he has been soliciting ideas and input from colleagues. The New York Times reported that Musk has advised Trump to hire key employees from SpaceX into his administration, including at the Department of Defense. This reflects the huge conflict of interest that Musk will face when it comes to space policy. His actions could significantly benefit SpaceX, of which he is the majority owner and has the final say in major decisions.

It will be a hugely weird dynamic. Musk is unquestionably in a position for self-dealing. Normally, such conflicts of interest would be frowned on within a government, but Trump has already shown a brazen disregard for norms, and there’s no reason to believe that will change during his second go at the presidency. One way around this could be to give Musk a “special adviser” tag, which means he would not have to comply with federal conflict-of-interest laws.

So it’s entirely possible that the sitting chief executive of SpaceX could be the nation’s most important adviser on space policy, conflicts be damned. Musk possesses flaws as a leader, but it is difficult to argue against results. His intuitions for the industry, such as pushing hard for reusable launch and broadband Internet from space, have largely been correct. In a vacuum, it is not necessarily bad to have someone like Musk providing a vision for US spaceflight in the 21st century. But while space may be a vacuum, there is plenty of oxygen in Washington, DC.

Being a space journalist got a lot more interesting this week—and a lot more difficult. As I waded through this reality on Wednesday, I began to reach out to sources about what is likely to happen. It’s way too early to have much certainty, but we can begin to draw some broad outlines for what may happen to space policy during a second Trump presidency. Buckle up—it could be a wild ride.

Bringing efficiency to NASA?

Let’s start with NASA and firmly establish what we mean. The US space agency does some pretty great things, but it’s also a bloated bureaucracy. That’s by design. Members of Congress write budgets and inevitably seek to steer more federal dollars to NASA activities in the areas they represent. Two decades ago, an engineer named Mike Griffin—someone Musk sought to hire as SpaceX’s first chief engineer in 2002—became NASA administrator under President George W. Bush.

Griffin recognized NASA’s bloat. For starters, it had too many field centers. NASA simply doesn’t need 10 major outposts across the country, as they end up fighting one another for projects and funding. However, Griffin knew he would face a titanic political struggle to close field centers, on par with federal efforts to close duplicative military bases during the “Base Realignment and Closure” process after the Cold War. So Griffin instead sought to make the best of the situation with his “Ten Healthy Centers” initiative. Work together, he told his teams across the country.

Essentially, then, for the last two decades, NASA programs have sought to leverage expertise across the agency. Consider the development of the Orion spacecraft, which began nearly 20 years ago. The following comment comes from Julie Kramer-White from an oral history interview conducted in 2016. Kramer is a long-time NASA engineer who was chief engineer of Orion at the time.

“I’ll tell you the truth, ten healthy centers is a pain in the butt,” she said. “The engineering team is a big engineering team, and they are spread across 9 of the 10 Centers… Our guys don’t think anything about a phone call that’s got people from six different centers. You’re trying to balance the time zone differences, and of course that’s got its own challenge with Europe as well but even within the United States with the different centers managing the time zone issue. I would say as a net technically, it’s a good thing. From a management perspective, boy, it’s a hassle.”

Space does not get done fast or efficiently by committee. But that’s how NASA operates—committees within committees, reviewed by committees.

Musk has repeatedly said he wants to bring efficiency to the US government and vowed to identify $2 trillion in savings. Well, NASA would certainly be more efficient with fewer centers—each of which has its own management layers, human resources setups, and other extensive overhead. But will the Trump administration really have the stomach to close centers? Certainly the congressional leadership from a state like Ohio would fight tooth and nail for Glenn Research Center. This offers an example of how bringing sweeping change to the US government in general, and NASA in particular, will run into the power of the purse held by Congress.

One tool NASA has used in recent years to increase efficiency is buying commercial services rather than leading the development of systems, such as the Orion spacecraft. This most prominent example is cargo and crew transportation to the International Space Station, but NASA has extended this approach to all manner of areas, from space communications to lunar landers to privately operated space stations. Congress has not always been happy with this transition because it has lessened its influence over steering funding directly to centers. NASA has nonetheless continued to push for this change because it has lowered agency costs, allowing it to do more.

Yet here again we run into conflicts of interest with Musk. The primary reason that NASA’s transition toward buying services has been a success is due to SpaceX. Private companies not named SpaceX have struggled to compete as NASA awards more fixed-price contracts for space services. Given Congress’ love for directing space funds to local centers, it’s unlikely to let Musk overhaul the agency in ways that send huge amounts of new business to SpaceX.

Where art thou, Artemis?

The biggest question is what to do with the Artemis program to return humans to the Moon. Ars wrote extensively about some of the challenges with this program a little more than a month ago, and Michael Bloomberg, founder of Bloomberg News, wrote a scathing assessment of Artemis recently under the headline “NASA’s $100 billion Moon mission is going nowhere.”

It is unlikely that outright cancellation of Artemis is on the table—after all, the first Trump administration created Artemis six years ago. However, Musk is clearly focused on sending humans to Mars, and the Moon-first approach of Artemis was championed by former Vice President Mike Pence, who is long gone. Trump loves grand gestures, and Musk has told Trump it will be possible to send humans to Mars before the end of his term. (That would be 2028, and it’s almost impossible to see this happening for a lot of reasons.) The Artemis architecture was developed around a “Moon-then-Mars” philosophy—as in, NASA will send humans to the Moon now, with Mars missions pushed into a nebulous future. Whatever Artemis becomes, it is likely to at least put Mars on equal footing to the Moon.

Notably, Musk despises NASA’s Space Launch System rocket, a central element of Artemis. He sees the rocket as the epitome of government bloat. And it’s not hard to understand why. The Space Launch System is completely expendable and costs about 10 to 100 times as much to launch as his own massive Starship rocket.

The key function the SLS rocket and the Orion spacecraft currently provide in Artemis is transporting astronauts from Earth to lunar orbit and back. There are ways to address this. Trump could refocus Artemis on using Starship to get humans to Mars. Alternatively, he could direct NASA to kludge together some combination of Orion, Dragon, and Falcon rockets to get astronauts to the Moon. He might also direct NASA to use the SLS for now but cancel further upgrades to it and a lunar space station called Gateway.

“The real question is how far is a NASA landing team and beachhead team are willing to go in destabilizing the program of record,” one policy source told Ars. “I can’t see Trump and Vance being less willing to shake up NASA than they are other public policy zones.”

What does seem clear is that, for the first time in 15 years, canceling the Space Launch System rocket or dramatically reducing its influence is on the table. This will be an acid test for Musk and Trump’s rhetoric on government efficiency, since the base of support for Artemis is in the deep-red South: states like Alabama, Mississippi, Louisiana, and Florida.

Will they really cut jobs there in the name of efficiency?

Regulatory reform

Reducing government regulations is one area in which the pathway for Musk and Trump is clear. The first Trump administration pushed to reduce regulations on US businesses almost from day one. In spaceflight, this produced Space Policy Directive-2 in 2018. Some progress was made, but it was far from total.

For spaceflight, Musk’s goal is to get faster approval for Starship test flights and licensing for the (literally) hundreds of launches SpaceX is already conducting annually. This will be broadly supported by the second Trump administration. During Trump’s first term, some of the initiatives in Space Policy Directive-2 were slowed or blocked by the Federal Aviation Administration and NASA, but the White House push will be even harder this time.

A looser regulatory environment should theoretically lead to more and more rapid progress in commercial space capabilities.

It’s worth noting here that if you spend any time talking to space startup executives, they all have horror stories about interacting with the FAA or other agencies. Pretty much everyone agrees that regulators could be more efficient but also that they need more resources to process rules in a timely manner. The FAA and Federal Communications Commission have important jobs when it comes to keeping people on the ground safe and keeping orbits sustainable in terms of traffic and space junk.

The second Trump administration will have some important allies on this issue in Congress. Ted Cruz, the US Senator from Texas, will likely chair the Senate Committee on Commerce, Science, and Transportation, which oversees legislation for space activities. He is one of the senators who has shown the most interest in commercial space, and he will support pro-business legislation—that is, laws that allow companies freer rein and regulatory agencies fewer teeth. How far this gets will depend on whether Republicans keep the House or Democrats take control.

Other areas of change

Over the course of the last seven decades, space has largely been a non-partisan topic.

But Musk’s deepening involvement in US space policy could pose a serious problem to this, as he’s now viewed extremely negatively by many Democrats. It seems probable that many people in Congress will oppose any significant shift of NASA’s focus from the Moon to Mars, particularly because it aligns with Musk’s long-stated goal of making humans a multiplanetary species.

There are likely to be battles in space science, as well. Traditionally, Republican presidents have cut funding for Earth science missions, and Democrats have increased funding to better study and understand climate change. Generally, given the administration’s likely focus on human spaceflight, space science will probably take a back seat and may lose funding.

Another looming issue is Mars Sample Return, which NASA is reconsidering due to budget and schedule issues. Presently, the agency intends to announce a new plan for retrieving rock and soil samples from Mars and returning them to Earth in December.

But if Musk and Trump are bent on sending humans to Mars as soon as possible, there is little sense in the space agency spending billions of dollars on a robotic sample return mission. Astronauts can just bring them back inside Starship.

Finally, at present, NASA has rich partnerships with space agencies around the world. In fact, it was the first Trump administration that created the Artemis Accords a little more than four years ago to develop an international coalition to return to the Moon. Since then, the United States and China have both been signing up partners in their competition to establish a presence at the South Pole of the Moon.

One huge uncertainty is how some of NASA’s long-established partners, especially in Europe, where there is bound to be tension around Ukraine and other issues with the Trump administration, will react at the US space agency’s exploration plans. Europeans are already wary of SpaceX’s prowess in global spaceflight and likely will not want to be on board with any space activities that further Musk’s ambitions.

These are just some of the high-level questions facing NASA and US spaceflight. There are many others. For example, how will Trump’s proposed tariffs on key components impact the national security and civil space supply chain? And there’s the Department of Defense, where the military already has multibillion dollar contracts with SpaceX, and there are bound to be similar conflicts and ethical concerns.

No one can hear you scream in space, but there will be plenty of screaming about space in the coming months.

Photo of Eric Berger

Eric Berger is the senior space editor at Ars Technica, covering everything from astronomy to private space to NASA policy, and author of two books: Liftoff, about the rise of SpaceX; and Reentry, on the development of the Falcon 9 rocket and Dragon. A certified meteorologist, Eric lives in Houston.

Space policy is about to get pretty wild, y’all Read More »

dna-shows-pompeii’s-dead-aren’t-who-we-thought-they-were

DNA shows Pompeii’s dead aren’t who we thought they were

ancient DNA, Archaeology, forensic archaeology, Pompeii, Science / /

People have long been fascinated by the haunting plaster casts of the bodies of people who died in Pompeii when Mount Vesuvius erupted in 79 CE. Archaeologists have presented certain popular narratives about who these people might have been and how they might have been related. But ancient DNA analysis has revealed that those preferred narratives were not entirely accurate and may reflect certain cultural biases, according to a new paper published in the journal Current Biology. The results also corroborate prior research suggesting that the people of ancient Pompeii were the descendants of immigrants from the Eastern Mediterranean.

As previously reported, the eruption of Mount Vesuvius released thermal energy roughly equivalent to 100,000 times the atomic bombs dropped on Hiroshima and Nagasaki at the end of World War II, spewing molten rock, pumice, and hot ash over the cities of Pompeii and Herculaneum in particular. The vast majority of people in Pompeii and Herculaneum—the cities hardest hit—perished from asphyxiation, choking on the thick clouds of noxious gas and ash. But at least some of the Vesuvian victims probably died instantaneously from the intense heat of fast-moving lava flows, with temperatures high enough to boil brains and explode skulls.

In the first phase, immediately after the eruption, a long column of ash and pumice blanketed the surrounding towns, most notably Pompeii and Herculaneum. By late night or early morning, pyroclastic flows (fast-moving hot ash, lava fragments, and gases) swept through and obliterated what remained, leaving the bodies of the victims frozen in seeming suspended action.

In the 19th century, an archaeologist named Giuseppe Fiorelli figured out how to make casts of those frozen bodies by pouring liquid plaster into the voids where the soft tissue had been. Some 1,000 bodies have been discovered in the ruins, and 104 plaster casts have been preserved. Restoration efforts of 86 of those casts began about 10 years ago, during which researchers took CT scans and X-rays to see if there were complete skeletons inside. Those images revealed that there had been a great deal of manipulation of the casts, depending on the aesthetics of the era in which they were made, including altering some features of the bodies’ shapes or adding metal rods to stabilize the cast, as well as frequently removing bones before casting.

DNA shows Pompeii’s dead aren’t who we thought they were Read More »

law-enforcement-operation-takes-down-22,000-malicious-ip-addresses-worldwide

Law enforcement operation takes down 22,000 malicious IP addresses worldwide

Biz & IT, cybercrime, Interpol, police, Security / /

An international coalition of police agencies has taken a major whack at criminals accused of running a host of online scams, including phishing, the stealing of account credentials and other sensitive data, and the spreading of ransomware, Interpol said recently.

The operation, which ran from the beginning of April through the end of August, resulted in the arrest of 41 people and the takedown of 1,037 servers and other infrastructure running on 22,000 IP addresses. Synergia II, as the operation was named, was the work of multiple law enforcement agencies across the world, as well as three cybersecurity organizations.

A global response

“The global nature of cybercrime requires a global response which is evident by the support member countries provided to Operation Synergia II,” Neal Jetton, director of the Cybercrime Directorate at INTERPOL, said. “Together, we’ve not only dismantled malicious infrastructure but also prevented hundreds of thousands of potential victims from falling prey to cybercrime. INTERPOL is proud to bring together a diverse team of member countries to fight this ever-evolving threat and make our world a safer place.”

Among the highlights of Operation Synergia II were:

Hong Kong (China): Police supported the operation by taking offline more than 1,037 servers linked to malicious services.

Mongolia: Investigations included 21 house searches, the seizure of a server and the identification of 93 individuals with links to illegal cyber activities.

Macau (China): Police took 291 servers offline.

Madagascar: Authorities identified 11 individuals with links to malicious servers and seized 11 electronic devices for further investigation.

Estonia: Police seized more than 80GB of server data, and authorities are now working with INTERPOL to conduct further analysis of data linked to phishing and banking malware.

The three private cybersecurity organizations that were part of Operation Synergia II were Group-IB, Kaspersky, and Team Cymru. All three used the telemetry intelligence in their possession to identify malicious servers and made it available to participating law enforcement agencies. The law enforcement agencies conducted investigations that resulted in house searches, the disruption of malicious cyber activities, the lawful seizures of servers and other electronic devices, and arrests.

Law enforcement operation takes down 22,000 malicious IP addresses worldwide Read More »

amazon’s-mass-effect-tv-series-is-actually-going-to-be-made

Amazon’s Mass Effect TV series is actually going to be made

Amazon, amazon prime video, culture, gaming, Mass Effect, TV / /

Confirming previous rumors, Variety reports that Amazon will be moving ahead with producing a TV series based on the popular Mass Effect video game franchise. The writing and production staff involved might not inspire confidence from fans, though.

The series’ writer and executive producer is slated to be Daniel Casey, who until now was best known as the primary screenwriter on F9: The Fast Saga, one of the late sequels in the Fast and the Furious franchise. He was also part of a team of writers behind the relatively little-known 2018 science fiction film Kin.

Karim Zreik will also produce, and his background is a little more encouraging; his main claim to fame is in the short-lived Marvel Television unit, which produced relatively well-received series like Daredevil and Jessica Jones for Netflix before Disney+ launched with its Marvel Cinematic Universe shows.

Another listed producer is Ari Arad, who has some background in video game adaptations, including the Borderlands and Uncharted movies, as well as the much-maligned live-action adaptation of Ghost in the Shell.

So yeah, it’s a bit of a mixed bag here. No plot details have been released, but it seems likely that the show will tell a new story rather than focus on the saga of Commander Shepherd from the games, since the games were all about the player inhabiting that character with their own choices. That’s only a guess, though.

Amazon is currently riding high after the smash success of another video game TV series, Fallout, which impressed both longtime and new fans when it debuted to critical acclaim and record viewing numbers earlier this year.

Amazon’s Mass Effect TV series is actually going to be made Read More »