Enlarge (credit: Getty Images | Benj Edwards)

On Thursday, Meta unveiled early versions of its Llama 3 open-weights AI model that can be used to power text composition, code generation, or chatbots. It also announced that its Meta AI Assistant is now available on a website and is going to be integrated into its major social media apps, intensifying the company's efforts to position its products against other AI assistants like OpenAI's ChatGPT, Microsoft's Copilot, and Google's Gemini.

Like its predecessor, Llama 2, Llama 3 is notable for being a freely available, open-weights large language model (LLM) provided by a major AI company. Llama 3 technically does not quality as "open source" because that term has a specific meaning in software (as we have mentioned in other coverage), and the industry has not yet settled on terminology for AI model releases that ship either code or weights with restrictions (you can read Llama 3's license here) or that ship without providing training data. We typically call these releases "open weights" instead.

At the moment, Llama 3 is available in two parameter sizes: 8 billion (8B) and 70 billion (70B), both of which are available as free downloads through Meta's website with a sign-up. Llama 3 comes in two versions: pre-trained (basically the raw, next-token-prediction model) and instruction-tuned (fine-tuned to follow user instructions). Each has a 8,192 token context limit.

Enlarge / Seeing power armor in the Fallout series inspired many players to jump back into their own suits (and start stockpiling fusion cores). (credit: Amazon/MGM Studios)

Long-time Fallout fans are used to long waits between titles, and, depending on their preferences, inconsistent results. But when Amazon's Fallout series showed up on Prime and absolutely nailed it, it spurred a lot of players to crack open their libraries and commence some post-apocalyptic replaying. And maybe first-time playing, too.

Fallout 76, the online multiplayer title that is the most recent full release, saw perhaps the biggest delta. The game hit its all-time peak of 43,887 simultaneous players on Wednesday, April 17, according to SteamCharts, roughly one week after the Amazon series' debut. For the year leading up to that peak, Fallout 76 had hovered around 7,000-10,000 players through most of 2024, and then jumped after the series' debut. Of course, that number only counts PC players, and only those on Steam; the game, which launched simultaneously on consoles, and is available on Microsoft's Game Pass, likely has many more players.

SteamDB, another Steam stats tracker, suggested on X (formerly Twitter) that the Fallout game series as a whole had more than doubled its concurrent player count by April 14.

Enlarge / The Mercedes-Benz E-Class is the brand's workhorse, covering millions of km a year ferrying German taxi passengers around. (credit: Peter Nelson)

Mercedes-Benz's E Class badge possesses a lot of canon. When asked to picture a '90s-or-newer Mercedes full-size sedan, it's hard not to conjure up mental snapshots of W124-and-up generations schlepping around well-heeled suburban communities. I bet a lot of folks also picture the taxi lineup outside any German train station, too.

Well, at least I do. But I'm not talking about envisioning surly Munich cabbies but rather the faithful four-doors they putter around in. A lot of them are E Classes, which further proves that the cars have always been quintessential Mercedes luxury workhorses.

The brand-new 2024 E 350 4Matic, with its thrifty turbo-four powerplant, might have the most direct lineage in the US market to those cream-colored rickety diesels. It's easy on fuel for its size and ready to soak up all the miles across both traffic-ridden cityscapes and high-speed stretches of highway. And with its expansive suite of tech and mild hybrid propulsion, it's an overall great next chapter for the badge.

Enlarge / Is that sooty rocket lifting off with the CRS-3 mission in 2014 a reused booster? No, it is not. (credit: SpaceX)

Ten years ago today, when a Falcon 9 rocket took off from Florida, something strange happened. Dramatically, as the rocket lifted off, a fountain of dirty water splashed upward alongside the vehicle, coating the rocket in grime.

Following the ultimately successful liftoff of this third cargo Dragon mission to the International Space Station, SpaceX founder Elon Musk was asked about the incident during a news conference. He offered a fairly generic answer without going into the details.

"We sprayed a bunch of water all around the pad," Musk said. "Essentially what happened is we splashed dirty water on ourselves. So it’s a little embarrassing, but no harm done."

Enlarge / An AI-generated image from DALL-E 2 created with the prompt "A painting by Grant Wood of an astronaut couple, american gothic style." (credit: AI Pictures That Go Hard / X)

When OpenAI's DALL-E 2 debuted on April 6, 2022, the idea that a computer could create relatively photorealistic images on demand based on just text descriptions caught a lot of people off guard. The launch began an innovative and tumultuous period in AI history, marked by a sense of wonder and a polarizing ethical debate that reverberates in the AI space to this day.

Last week, OpenAI turned off the ability for new customers to purchase generation credits for the web version of DALL-E 2, effectively killing it. From a technological point of view, it's not too surprising that OpenAI recently began winding down support for the service. The 2-year-old image generation model was groundbreaking for its time, but it has since been surpassed by DALL-E 3's higher level of detail, and OpenAI has recently begun rolling out DALL-E 3 editing capabilities.

But for a tight-knit group of artists and tech enthusiasts who were there at the start of DALL-E 2, the service's sunset marks the bittersweet end of a period where AI technology briefly felt like a magical portal to boundless creativity. "The arrival of DALL-E 2 was truly mind-blowing," illustrator Douglas Bonneville told Ars in an interview. "There was an exhilarating sense of unlimited freedom in those first days that we all suspected AI was going to unleash. It felt like a liberation from something into something else, but it was never clear exactly what."

Enlarge / The Framework Laptop 13. (credit: Andrew Cunningham)

Since Framework showed off its first prototypes in February 2021, we've generally been fans of the company's modular, repairable, upgradeable laptops.

Not that the company's hardware releases to date have been perfect—each Framework Laptop 13 model has had quirks and flaws that range from minor to quite significant, and the Laptop 16's upsides struggle to balance its downsides. But the hardware mostly does a good job of functioning as a regular laptop while being much more tinkerer-friendly than your typical MacBook, XPS, or ThinkPad.

But even as it builds new upgrades for its systems, expands sales of refurbished and B-stock hardware as budget options, and promotes the re-use of its products via external enclosures, Framework has struggled with the other side of computing longevity and sustainability: providing up-to-date software.

Enlarge / An image of a boy amazed by flying letters. (credit: Getty Images)

Some weeks in AI news are eerily quiet, but during others, getting a grip on the week's events feels like trying to hold back the tide. This week has seen three notable large language model (LLM) releases: Google Gemini Pro 1.5 hit general availability with a free tier, OpenAI shipped a new version of GPT-4 Turbo, and Mistral released a new openly licensed LLM, Mixtral 8x22B. All three of those launches happened within 24 hours starting on Tuesday.

With the help of software engineer and independent AI researcher Simon Willison (who also wrote about this week's hectic LLM launches on his own blog), we'll briefly cover each of the three major events in roughly chronological order, then dig into some additional AI happenings this week.

Gemini Pro 1.5 general release

On Tuesday morning Pacific time, Google announced that its Gemini 1.5 Pro model (which we first covered in February) is now available in 180-plus countries, excluding Europe, via the Gemini API in a public preview. This is Google's most powerful public LLM so far, and it's available in a free tier that permits up to 50 requests a day.

Enlarge (credit: Benj Edwards)

On Monday, OpenAI announced that visitors to the ChatGPT website in some regions can now use the AI assistant without signing in. Previously, the company required that users create an account to use it, even with the free version of ChatGPT that is currently powered by the GPT-3.5 AI language model. But as we have noted in the past, GPT-3.5 is widely known to provide more inaccurate information compared to GPT-4 Turbo, available in paid versions of ChatGPT.

Since its launch in November 2022, ChatGPT has transformed over time from a tech demo to a comprehensive AI assistant, and it has always had a free version available. The cost is free because "you're the product," as the old saying goes. Using ChatGPT helps OpenAI gather data that will help the company train future AI models, although free users and ChatGPT Plus subscription members can both opt out of allowing the data they input into ChatGPT to be used for AI training. (OpenAI says it never trains on inputs from ChatGPT Team and Enterprise members at all.)

Opening ChatGPT to everyone could provide a frictionless on-ramp for people who might use it as a substitute for Google Search or potentially gain new customers by providing an easy way for people to use ChatGPT quickly, then offering an upsell to paid versions of the service.

Enlarge / Teams is being decoupled from the other Office apps worldwide, six months after Microsoft did the same thing for the EU. (credit: Microsoft/Andrew Cunningham)

Months after unbundling the apps in the European Union, Microsoft is taking the Office and Teams breakup worldwide. Reuters reports that Microsoft will begin selling Teams and the other Microsoft 365 apps to new commercial customers as separate products with separate price tags beginning today.

"To ensure clarity for our customers, we are extending the steps we took last year to unbundle Teams from M365 and O365 in the European Economic Area and Switzerland to customers globally," a Microsoft spokesperson told Ars. "Doing so also addresses feedback from the European Commission by providing multinational companies more flexibility when they want to standardize their purchasing across geographies."

The unbundling is a win for other team communication apps like Slack and videoconferencing apps like Zoom, both of which predate Teams but haven't had the benefits of the Office apps' huge established user base.

Enlarge / Hertz currently has more than a thousand Bolt EUVs for sale as they leave its rental car fleet. (credit: Chevrolet)

Electric vehicles have many advantages over cars that still use internal combustion engines. They're far more efficient, they're quieter, and they usually have much more torque than their gasoline-powered equivalents. But we're still far from achieving price parity between powertrains. In other words, EVs are expensive.

One place you can find some bargains, though, is the rental company Hertz, which currently has more than 2,100 EVs for sale, more than half of which are affordable enough to qualify for the IRS used clean vehicle tax credit.

Hertz has been adding a lot of EVs to its fleet as part of the company's decarbonization plan. In 2021, it revealed plans to purchase 100,000 Teslas. However, the controversial car maker had delivered fewer than half of those two years later, and long repair times for customer-inflicted damage have seen the rental agency divest itself of many of those Teslas and diversify its fleet, adding plenty of Polestars, Kias, and Chevrolets.

Enlarge (credit: Getty Images / Benj Edwards)

On Tuesday, Anthropic's Claude 3 Opus large language model (LLM) surpassed OpenAI's GPT-4 (which powers ChatGPT) for the first time on Chatbot Arena, a popular crowdsourced leaderboard used by AI researchers to gauge the relative capabilities of AI language models. "The king is dead," tweeted software developer Nick Dobos in a post comparing GPT-4 Turbo and Claude 3 Opus that has been making the rounds on social media. "RIP GPT-4."

Since GPT-4 was included in Chatbot Arena around May 10, 2023 (the leaderboard launched May 3 of that year), variations of GPT-4 have consistently been on the top of the chart until now, so its defeat in the Arena is a notable moment in the relatively short history of AI language models. One of Anthropic's smaller models, Haiku, has also been turning heads with its performance on the leaderboard.

"For the first time, the best available models—Opus for advanced tasks, Haiku for cost and efficiency—are from a vendor that isn't OpenAI," independent AI researcher Simon Willison told Ars Technica. "That's reassuring—we all benefit from a diversity of top vendors in this space. But GPT-4 is over a year old at this point, and it took that year for anyone else to catch up."

Enlarge (credit: peterschreiber.media | Getty Images)

The US Justice Department on Monday unsealed an indictment charging seven men with hacking or attempting to hack dozens of US companies in a 14-year campaign furthering an economic espionage and foreign intelligence gathering by the Chinese government.

All seven defendants, federal prosecutors alleged, were associated with Wuhan Xiaoruizhi Science & Technology Co., Ltd. a front company created by the Hubei State Security Department, an outpost of the Ministry of State Security located in Wuhan province. The MSS, in turn, has funded an advanced persistent threat group tracked under names including APT31, Zirconium Violet Typhoon, Judgment Panda, and Altaire.

Relentless 14-year campaign

“Since at least 2010, the defendants … engaged in computer network intrusion activity on behalf of the HSSD targeting numerous US government officials, various US economic and defense industries and a variety of private industry officials, foreign democracy activists, academics and parliamentarians in response to geopolitical events affecting the PRC,” federal prosecutors alleged. “These computer network intrusion activities resulted in the confirmed and potential compromise of work and personal email accounts, cloud storage accounts and telephone call records belonging to millions of Americans, including at least some information that could be released in support of malign influence targeting democratic processes and institutions, and economic plans, intellectual property, and trade secrets belonging to American businesses, and contributed to the estimated billions of dollars lost every year as a result of the PRC’s state-sponsored apparatus to transfer US technology to the PRC.”

Enlarge / Sometimes your infernal-engine-powered heart just isn't in it. (credit: Larian Studios/Hasbro)

Swen Vincke, director of the colossal entity that is Baldur's Gate 3, is not leaving the door open to future expansions of that already fully packed game.

At this week's Game Developer's Conference (GDC), Vincke made it clear during a talk and in interviews that Larian Studios is not going to make any major new content for Baldur's Gate 3 (BG3)—nor start work on Baldur's Gate 4, nor make anything, really, inside the framework of Dungeons & Dragons' Fifth Edition (5e).

Not that Vincke or his team are bitter. Their hearts just aren't in it. They had actually started work on BG3 downloadable content and gave some thought to Baldur's Gate 4, Vincke told IGN. "But we hadn’t really had closure on BG3 yet and just to jump forward on something new felt wrong." On top of that, the team had new ideas that didn't fit D&D 5e, which "is not an easy system to put into a video game," Vincke said.

Enlarge (credit: AMD)

Last summer, AMD debuted the latest version of its FidelityFX Super Resolution (FSR) upscaling technology. While version 2.x focused mostly on making lower-resolution images look better at higher resolutions, version 3.0 focused on AMD's "Fluid Motion Frames," which attempt to boost FPS by generating interpolated frames to insert between the ones that your GPU is actually rendering.

Today, the company is announcing FSR 3.1, which among other improvements decouples the upscaling improvements in FSR 3.x from the Fluid Motion Frames feature. FSR 3.1 will be available "later this year" in games whose developers choose to implement it.

Fluid Motion Frames and Nvidia's equivalent DLSS Frame Generation usually work best when a game is already running at a high frame rate, and even then can be more prone to mistakes and odd visual artifacts than regular FSR or DLSS upscaling. FSR 3.0 was an all-or-nothing proposition, but version 3.1 should let you pick and choose what features you want to enable.

Enlarge / A Falcon 9 rocket lifts off Thursday from Cape Canaveral, Florida. (credit: NASA/SpaceX)

Upgrades at SpaceX's most-used launch pad in Florida got a trial run Thursday with the liftoff of a Falcon 9 rocket with a Dragon cargo ship heading for the International Space Station.

SpaceX's Cargo Dragon spacecraft launched at 4:55 pm EDT (20:55 UTC) Thursday from Space Launch Complex 40 (SLC-40) at Cape Canaveral Space Force Station in Florida. This mission, known as CRS-30, is SpaceX's 30th resupply mission to the space station since 2012.

The automated Dragon supply ship took off on top of a Falcon 9 rocket, heading for a monthlong stay at the International Space Station, where it will deliver more than 6,000 pounds of hardware, fresh food, and experiments for the lab's seven-person crew.

Enlarge

When OpenAI launched its GPT-4 AI model a year ago, it created a wave of immense hype and existential panic from its ability to imitate human communication and composition. Since then, the biggest question in AI has remained the same: When is GPT-5 coming out? During interviews and media appearances around the world, OpenAI CEO Sam Altman frequently gets asked this question, and he usually gives a coy or evasive answer, sometimes coupled with promises of amazing things to come.

According to a new report from Business Insider, OpenAI is expected to release GPT-5, an improved version of the AI language model that powers ChatGPT, sometime in mid-2024—and likely during the summer. Two anonymous sources familiar with the company have revealed that some enterprise customers have recently received demos of GPT-5 and related enhancements to ChatGPT.

One CEO who recently saw a version of GPT-5 described it as "really good" and "materially better," with OpenAI demonstrating the new model using use cases and data unique to his company. The CEO also hinted at other unreleased capabilities of the model, such as the ability to launch AI agents being developed by OpenAI to perform tasks automatically.

Enlarge (credit: Kevin Noki)

Have you ever worked on a hobby project where modifying and compiling the source code for a Linux-based emulator was possibly the easiest and most straightforward part of the whole thing?

Kevin Noki really, really wanted a functioning Macintosh Plus, complete with a functioning, auto-ejecting disk drive that it could boot from. The German maker already had a Mac Plus (1Mb) from eBay, but it had both a busted power supply and floppy drive. Rather than carve out the busted Plus' one-of-a-kind internals and slap a Raspberry Pi in there like some DIY slacker, Noki went… a different path.

Noki 3D-printed his own Macintosh, the "Brewintosh." I would like you to consider what you think that last sentence means and then wipe your expectations clean. I have watched the entire 48-minute journey of Noki's Brewintosh, which is both very soothing on some ASMR-adjacent gut level and also low-key maddening for the way it plays down all the individual accomplishments along the way. Any one of the Brewintosh's pieces would be my entire weekend, and my spouse would not enjoy my mood while I was sunk into it.

Enlarge

A couple of weeks ago, Apple released macOS Sonoma 14.4 with the usual list of bug fixes, security patches, and a couple of minor new features. Since then, users and companies have been complaining of a long list of incompatibilities, mostly concerning broken external accessories like USB hubs and printers but also extending to software like Java.

MacRumors has a good rundown of the list of issues, which has been steadily getting longer as people have run into more problems. It started with reports of malfunctioning USB hubs, sourced from users on Reddit, the Apple Support Communities forums, and elsewhere—USB hubs built into various displays stopped functioning for Mac users after the 14.4 update.

Other issues surfaced in the days after people started reporting problems with their USB hubs, including some instances of broken printer drivers, unexpected app crashes for some Java users, and problems launching apps that rely on the PACE anti-piracy software (and iLok hardware dongles) to authenticate.

Enlarge (credit: Getty Images)

It seems like AI large language models (LLMs) are everywhere these days due to the rise of ChatGPT. Now, a software developer named Ishan Anand has managed to cram a precursor to ChatGPT called GPT-2—originally released in 2019 after some trepidation from OpenAI—into a working Microsoft Excel spreadsheet. It's freely available and is designed to educate people about how LLMs work.

"By using a spreadsheet anyone (even non-developers) can explore and play directly with how a 'real' transformer works under the hood with minimal abstractions to get in the way," writes Anand on the official website for the sheet, which he calls "Spreadsheets-are-all-you-need." It's a nod to the 2017 research paper "Attention is All You Need" that first described the Transformer architecture that has been foundational to how LLMs work.

Anand packed GPT-2 into an XLSB Microsoft Excel binary file format, and it requires the latest version of Excel to run (but won't work on the web version). It's completely local and doesn't do any API calls to cloud AI services.

Enlarge / National Transportation Safety Board Chair Jennifer Homendy testifies about the Boeing door-plug investigation before the Senate Commerce, Science, and Transportation Committee on March 6, 2024, in Washington, DC. (credit: Getty Images | Kevin Dietsch )

A government investigation into a Boeing 737 Max 9 plane's door-plug blowout has been hampered by a lack of repair records and security camera footage, the National Transportation Safety Board's chair told US senators. Boeing was "unable to find the records" and told the NTSB that the security camera footage was overwritten.

"To date, we still do not know who performed the work to open, reinstall, and close the door plug on the accident aircraft," NTSB Chair Jennifer Homendy wrote Wednesday in a letter to leaders of the Senate Commerce, Science, and Transportation Committee. "Boeing has informed us that they are unable to find the records documenting this work. A verbal request was made by our investigators for security camera footage to help obtain this information; however, they were informed the footage was overwritten. The absence of those records will complicate the NTSB's investigation moving forward."

A Boeing spokesperson told Ars today that under the company's standard practice, "video recordings are maintained on a rolling 30-day basis" before being overwritten. The NTSB's preliminary report on the investigation said the airplane was delivered to Alaska Airlines on October 31, 2023, after a repair in a Boeing factory. On January 5, the plane was forced to return to Portland International Airport in Oregon when a passenger door plug blew off the aircraft during flight.

Enlarge / SpaceX's Starship soars through the sky over South Texas, powered by 33 methane-burning Raptor engines. (credit: Stephen Clark/Ars Technica)

SpaceX's new-generation Starship rocket, the most powerful and largest launcher ever built, flew halfway around the world following liftoff from South Texas on Thursday, accomplishing a key demonstration of its ability to carry heavyweight payloads into low-Earth orbit.

SpaceX's third towering Starship rocket, standing some 397 feet (121 meters) tall and wider than the fuselage of a 747 jumbo jet, lifted off at 8:25 am CDT (13:25 UTC) Thursday from SpaceX's Starbase launch facility on the Texas Gulf Coast east of Brownsville. SpaceX delayed the liftoff time by nearly an hour and a half to wait for boats to clear out of restricted waters near the launch base.

Hitting its marks

The successful launch builds on two Starship test flights last year that achieved some, but not all, of their objectives and appears to put the privately funded rocket program on course to begin launching satellites, allowing SpaceX to ramp up the already-blistering pace of Starlink deployments.

Enlarge / SpaceX's third flight-ready Starship rocket underwent a countdown dress rehearsal earlier this month. (credit: SpaceX)

The Federal Aviation Administration said Wednesday it has approved a commercial launch license for the third full-scale test flight of SpaceX's giant Starship rocket.

This is the final regulatory hurdle before SpaceX can launch Starship from South Texas. The third flight of the world's most powerful rocket, following a pair of test launches last year, is scheduled for Thursday morning.

SpaceX's Starship rocket and Super Heavy booster will take off from the Starbase test site on the Texas Gulf Coast, a few miles north of the US-Mexico border. The launch window for the nearly 400-foot-tall (121-meter) rocket opens at 7 am CDT (12:00 UTC) Thursday and extends for 110 minutes. You can watch the launch live on SpaceX's X page or on one of several third-party YouTube livestreams. We have embedded a livestream from LabPadre and Spaceflight Now here.

Enlarge / Polestar is running a few months late with its next EV, but US deliveries should begin within the next three months. (credit: Polestar)

Polestar is almost ready to start selling its next electric vehicle here in the US. The Polestar 3 SUV has finally gone into production, and the automaker says deliveries to US customers should start in the second quarter of this year. That's a little behind schedule—in October 2022, we were told that the Polestar 3 would go on sale here toward the end of 2023.

There's just one powertrain at launch, the Polestar 3 long-range dual motor, which starts at $73,400. This offers 489 hp (365 kW) and 620 lb-ft (840 Nm), which should give the EV a range of about 315 miles (507 km) from its 111 kWh battery pack. And Polestar is making the Pilot Pack standard equipment, which includes a head-up display and some advanced driver assistance systems.

There will also be the Polestar 3 with the Plus Pack, for $78,900. This adds different 21-inch wheels, an upgraded interior with either "MicroTech" fabric or wool covering the seats, heated rear seats to go with the heated front seats, and an upgraded Bowers and Wilkins sound system with 25 speakers and a subwoofer.

Longtime readers and first-person shooter fans might remember how excited we were for the announcement of Timesplitters 4 in late 2008. Unfortunately, an intriguing trailer was the last we saw of the game before the project was killed amidst Free Radical's bankruptcy later that year.

Today, though, we actually have a chance to see and play what we missed over 15 years ago. That's thanks to an eagle-eyed eBay trawler who grabbed an aging PlayStation 3 development kit that happened to have a prototype of Timesplitters 4 sitting on its hard drive.

Reddit user Flimsy-Zebra3775 posted about the find earlier this month, asking the PS3 subreddit community for guidance on what looked like a working prototype of the long-lost game. After reportedly asking the seller "a lot of probing questions" and confirming that the prototype could be booted, Flimnsy-Zebra reportedly paid 525 pounds (about $670) for the dev kit, saying, "for history and my collection, it's worth it."

Enlarge / The 13- and 15-inch M3 MacBook Air. (credit: Andrew Cunningham)

When Apple upgraded its Macs with the M2 chip, some users noticed that storage speeds were actually quite a bit lower than they were in the M1 versions. Both the 256GB M2 MacBook Air and the 512GB M2 MacBook Pro had their storage speeds roughly halved compared to M1 Macs with the same storage capacities.

Teardowns revealed that this was because Apple was using fewer physical flash memory chips to provide the same amount of storage. Modern SSDs achieve their high speeds partly by reading from and writing to multiple NAND flash chips simultaneously, a process called "interleaving." When there's only one flash chip to access, speeds go down.

Early teardowns of the M3 MacBook Air suggest that Apple may have reversed course here, at least for some Airs. The Max Tech YouTube channel took a 256GB M3 Air apart, showing a pair of 128GB NAND flash chips rather than the single 256GB chip that the M2 Air used. BlackMagic Disk Speed Test performance increases accordingly; read and write speeds for the 256GB M2 Air come in at around 1,600 MB/s, while the M3 Air has read speeds of roughly 2,900 MB/s and write speeds of about 2,100 MB/s. That's roughly in line with the M1 Air's performance.

Enlarge / From left to right, the Rivian R2, Rivian R3, and Rivian R3X. (credit: Rivian)

Today in California, Rivian CEO and founder RJ Scaringe unveiled new additions to the electric vehicle startup's model lineup. Details of the new R2 had leaked earlier this week, although not entirely accurately—the new smaller EV will start at $45,000 (not $47,000) when it goes on sale in 2026.

At first glance of the R2, you might be excused for thinking you're looking at a Rivian R1S, for both SUVs share a lot of styling details—not just Rivian's distinctive daylight running lights but also the overall shape of the vehicle. But the new car is shorter in both length (by 15.7 inches/400 mm) and height, and only offers two rows of seating, not three.

Convenience improvements include a pair of gloveboxes under the dashboard and a flashlight that's stored in the door. At the back, the glass rear window can drop down to allow access to the cargo area. Both rows of seats fold flat, in case you want to camp in your car.

The countdown continues for the hotly anticipated debut of 3 Body Problem, Netflix's eight-episode sci-fi series adapted from the award-winning novel The Three-Body Problem by Liu Cixin, the first book in his Remembrance of Earth's Past trilogy. Those attending the SXSW Film & Television Festival in Austin will get to see the series' world premiere tonight. The rest of us have to wait until later this month, but in the meantime, the streaming platform has released a final trailer.

(Some spoilers for the novel below.)

The 3-Body Problem's narrative is told in a nonlinear fashion, jumping between a young astrophysicist, Ye Wenjie, who witnesses her father being beaten to death by Red Guards during the Cultural Revolution, and Ye's return to Tsinghua University as an established professor many years later. During the earlier timeline, Ye figures out a means of sending an interstellar message to possible extraterrestrial civilizations and receives a response from a planet called Trisolaris. (As its name implies, the planet has three suns, which wreak havoc on Trisolaris via unpredictable "chaotic periods"—hence the novel's title, which refers to a classic problem in celestial mechanics.) Despite being warned that the aliens intend to invade and conquer Earth, Ye responds to the message and invites them to do so, disillusioned by the state of the world.

Enlarge (credit: Getty Images)

On Monday, Anthropic prompt engineer Alex Albert caused a small stir in the AI community when he tweeted about a scenario related to Claude 3 Opus, the largest version of a new large language model launched on Monday. Albert shared a story from internal testing of Opus where the model seemingly demonstrated a type of "metacognition" or self-awareness during a "needle-in-the-haystack" evaluation, leading to both curiosity and skepticism online.

Metacognition in AI refers to the ability of an AI model to monitor or regulate its own internal processes. It's similar to a form of self-awareness, but calling it that is usually seen as too anthropomorphizing, since there is no "self" in this case. Machine-learning experts do not think that current AI models possess a form of self-awareness like humans. Instead, the models produce humanlike output, and that sometimes triggers a perception of self-awareness that seems to imply a deeper form of intelligence behind the curtain.

In the now-viral tweet, Albert described a test to measure Claude's recall ability. It's a relatively standard test in large language model (LLM) testing that involves inserting a target sentence (the "needle") into a large block of text or documents (the "haystack") and asking if the AI model can find the needle. Researchers do this test to see if the large language model can accurately pull information from a very large processing memory (called a context window), which in this case is about 200,000 tokens (fragments of words).

Enlarge / The Anthropic Claude 3 logo. (credit: Anthropic)

On Monday, Anthropic released Claude 3, a family of three AI language models similar to those that power ChatGPT. Anthropic claims the models set new industry benchmarks across a range of cognitive tasks, even approaching "near-human" capability in some cases. It's available now through Anthropic's website, with the most powerful model being subscription-only. It's also available via API for developers.

Claude 3's three models represent increasing complexity and parameter count: Claude 3 Haiku, Claude 3 Sonnet, and Claude 3 Opus. Sonnet powers the Claude.ai chatbot now for free with an email sign-in. But as mentioned above, Opus is only available through Anthropic's web chat interface if you pay $20 a month for "Claude Pro," a subscription service offered through the Anthropic website. All three feature a 200,000-token context window. (The context window is the number of tokens—fragments of a word—that an AI language model can process at once.)

We covered the launch of Claude in March 2023 and Claude 2 in July that same year. Each time, Anthropic fell slightly behind OpenAI's best models in capability while surpassing them in terms of context window length. With Claude 3, Anthropic has perhaps finally caught up with OpenAI's released models in terms of performance, although there is no consensus among experts yet—and the presentation of AI benchmarks is notoriously prone to cherry-picking.

Enlarge / Apple's M3 MacBook Pro should be able to drive a pair of external displays soon, as long as the lid is closed. (credit: Apple)

One long-standing limitation of Apple's most basic Mac chips—the plain-old, no-adjective M1, M2, and M3—has been their inability to work with more than a single external monitor at a time.

This was one of the only ways in which the Apple Silicon era has been a step back from the Intel era, where most Macs supported at least two external displays, plus the screen built into the MacBook Air or Pro you were using. (Would an integrated Intel GPU actually work well with that many screens connected? Usually not. But at least you could try.) And it was one factor that might push a user in the direction of a more-expensive MacBook Pro with an M1 Pro, M1 Max, M2 Pro, or M2 Max, all of which have always supported multiple external displays.

When Apple launched the $1,599 M3 version of the MacBook Pro last fall, the one-external-display limitation was still in place. But today's announcement of the M3 MacBook Air came with a small but pleasant surprise for anyone who wants their Mac to do double-duty as a laptop and a desktop—if the laptop's lid is closed, the M3 can now run a pair of external displays over its Thunderbolt ports. And Apple confirmed to 9to5Mac that the 14-inch M3 MacBook Pro would have the same functionality enabled via a future macOS update (we've asked Apple to verify whether it's coming in the imminent macOS 14.4 update or at some future date).

Enlarge / Apple is refreshing the MacBook Air with M3 chips but leaving everything else about the 2022 redesign intact. (credit: Apple)

Apple has quietly refreshed its MacBook Air lineup, bringing new chips (and in some cases, new prices) to its most popular laptops. New 13- and 15-inch MacBook Airs include Apple's latest-generation M3 chip, while the old M2 MacBook Air now replaces 2020's M1 MacBook Air as Apple's $999 entry-level laptop. The new 13- and 15-inch M3 systems start at $1,099 and $1,299; they can be ordered today and will be released on March 8.

The new Airs use the same design as the M2 versions. Compared to older M1 and late-Intel-era Airs, they have slightly larger displays with a prominent notch, a non-tapered but still thin-and-light chassis, larger trackpads, modestly refined keyboards, and a MagSafe port for charging.

All of the new Airs use the M3, with no options to upgrade to faster or more capable processors (this means the Air is still restricted to just a single external display when the built-in display is on, though the M3 can now drive two external displays when the lid is closed, something older chips didn't offer). The $1,099 13-inch Air does use a slightly cut-down version of the chip with 8 GPU cores instead of 10, with the 10-core GPU available as a $100 upgrade; all 15-inch models use the fully enabled M3 with the 10-core GPU.

Enlarge (credit: Bloomberg / Contributor | Bloomberg)

Amid safety scandals involving "many loose bolts" and widespread problems with Boeing's 737 Max 9s, Boeing is apparently considering buying back Spirit AeroSystems, the key supplier behind some of Boeing's current manufacturing problems, sources told The Wall Street Journal.

Spirit was initially spun out from Boeing Commercial Airplanes in 2005, and Boeing had planned to keep it that way. Last year, Boeing CEO Dave Calhoun sought to dispel rumors that Boeing might reacquire Spirit as federal regulators launched investigations into both companies. But now Calhoun appears to be "softening that stance," the WSJ reported.

According to the WSJ's sources, no deal has formed yet, but Spirit has initiated talks with Boeing and "hired bankers to explore strategic options." Sources also confirmed that Spirit is weighing whether to sell its operations in Ireland, which manufactures parts for Boeing rival Airbus.

Enlarge (credit: Microsoft)

For the third year running, Microsoft is releasing a batch of small- to medium-sized Windows 11 updates in February. This year's collection of tweaks includes the now-obligatory smattering of generative AI features but also some window-snapping changes, better Android phone integration, and accessibility upgrades, among other things.

Starting with the non-AI features: If you've paired an Android phone with your PC in the Your Phone app, Microsoft says that "soon" you will be able to use the phone's camera as a webcam during video calls. This mirrors a similar Mac-to-iPhone feature that Apple added to macOS a couple of years ago; though most PCs these days are sold with webcams, the camera on any reasonably recent Android phone will be a visual upgrade.

Window snapping is getting some "intelligent suggestions" that Microsoft says will "help you quickly organize open apps based on how you use them." And the Widgets view gets a new "focused" view and subcategories that will let you easily switch between widget boards if you want different widgets for different contexts (one for home and one for work, for example).

Enlarge / Ford was the first OEM to announce it was switching to J3400, and it's the first automaker to gain access to the Tesla Supercharger network. (credit: Ford)

Today, Ford electric vehicles gained access to the Tesla Supercharger network. Last May, the Blue Oval was the first automaker to throw its lot in with what was then called the North American Charging Standard and is now known as J3400. Ford proved to be the first domino falling, and with Stellantis' announcement earlier this month that it too would move to J3400, the more compact DC fast-charging plug will be the de facto standard in the next couple of years.

Until Ford made the switch, every non-Tesla EV in North America had settled on the Combined Charging Standard 1 plug (with the exception of the Nissan Leaf, which still uses CHAdeMO). CCS1 and J3400 use the same electronic communication protocols—only the actual plug and socket are different.

But it will take some time for car makers to start building J3400 ports into their EVs. That should begin next year, probably with the introduction of model year 2026. This means that EVs older than MY26 will need to use a passive adapter to mate a J3400 charger cable with a CCS1-equipped EV.

Enlarge / BMW calls the X2 a Sport Activity Coupe. (credit: Jonathan Gitlin)

Last week, we told you about our first drive in the new Mini Countryman John Cooper Works, Mini's new little crossover. This week, it's the turn of a related model, built on the same vehicle architecture: BMW's new X2 crossover, or "Sports Activity Coupe" in BMW-speak. As we'll find out, the BMW shares more than one trait with the Countryman JCW.

BMW had an array of X2 M35is, all painted the same "Frozen Tampa Bay" shade of green, which starts in the US at $51,400. There's a less-powerful $42,000 X2 xDrive28i coming here as well, but North American customers will not be offered the battery-electric iX2—BMW's product planners evidently didn't think importing the diminutive EV would be profitable. Outside the US, BMW expects 1 in 5 X2s to be electric.

The first-generation X2 (and the more upright-looking X1) were divisive cars even by BMW standards. The new one is slightly bigger than before, at 179.3 inches (4,554 mm) long, 72.6 inches (1,844 mm) wide, and 62.6 inches (1,590 mm) tall. That translates to more rear legroom and more cargo volume at the back, but it's not a massive machine—a touch bigger than the Audi Q3 but a bit smaller than a Mercedes-Benz GLB.

Enlarge (credit: Busà Photography | Moment Unreleased)

OpenAI is now boldly claiming that The New York Times "paid someone to hack OpenAI’s products" like ChatGPT to "set up" a lawsuit against the leading AI maker.

In a court filing Monday, OpenAI alleged that "100 examples in which some version of OpenAI’s GPT-4 model supposedly generated several paragraphs of Times content as outputs in response to user prompts" do not reflect how normal people use ChatGPT.

Instead, it allegedly took The Times "tens of thousands of attempts to generate" these supposedly "highly anomalous results" by "targeting and exploiting a bug" that OpenAI claims it is now "committed to addressing."

Enlarge (credit: Floogle/X)

Nintendo has made some bold, weird choices with its hardware designs. But none were so bold and weird as 1995's Virtual Boy, a "woefully premature commercial curio," as one Ars writer put it, that "quickly passed unlamented into history," as remarked another. The awkward red-on-black tabletop headset system wasn't so much ahead of its time as beamed in from an alternate reality. In this reality, it didn't sell much and was largely forgotten.

Nintendo has seemed eager to let the Virtual Boy fade from the collective memory, but clever coders have labored to keep the system accessible outside vintage hardware collections. The latest, and perhaps most accessible, is Red Viper, which plays Virtual Boy games on a (lightly hacked) Nintendo 3DS, the other Nintendo system on which 3D features were underappreciated. It is full-speed, it supports homebrew games, you can change the drawing color to something other than red, and it is free. It's built on top of the work of earlier 3DS emulator r3dragon, which itself drew heavily from the Reality Boy project for Windows.

Red Viper makes use of the 3DS's top screen for game display and turns the lower screen into a system options panel. It maps the Virtual Boy's own face buttons onto the touchscreen. In the Twitter thread announcing Red Viper's general release, coder Floogle notes that the emulator is only roughly translating the Virtual Boy's 50 Hz refresh to the 3DS' 60 Hz by pushing a frame every 20 ms. There is, Floogle supposes, some hardware headroom for improvement.

Enlarge / A photo of "Willy's Chocolate Experience" (inset), which did not match AI-generated promises, shown in the background. (credit: Stuart Sinclair)

On Saturday, event organizers shut down a Glasgow-based "Willy's Chocolate Experience" after customers complained that the unofficial Wonka-inspired event, which took place in a sparsely decorated venue, did not match the lush AI-generated images listed on its official website (archive here). According to Sky News, police were called to the event, and "advice was given."

"What an absolute shambles of an event," wrote Stuart Sinclar on Facebook after paying 35 pounds per ticket for himself and his kids. "Took 2 minutes to get through to then see a queue of people surrounding the guy running it complaining ... The kids received 2 jelly babies and a quarter of a can of Barrs limeade."

The Willy's Chocolate Experience website, which promises "a journey filled with wondrous creations and enchanting surprises at every turn," features five AI-generated images (likely created with OpenAI's DALL-E 3) that evoke a candy-filled fantasy wonderland inspired by the Willy Wonka universe and the recent Wonka film. But in reality, Sinclair was met with a nearly empty location with a few underwhelming decorations and a tiny bouncy castle. In one photo shared by Sinclair, a rainbow arch leads to a single yellow gummy bear and gum drop sitting on a bare concrete floor.

Enlarge / A shopper looks at a meat display on June 20, 2022, at the Market 32 Supermarket in South Burlington, Vermont. Niacin can be found in foods such as red meat, poultry, fish, fortified cereals and breads, brown rice, nuts, legumes, and bananas. (credit: Getty | Robert Nickelsberg)

In the early 20th century, the deadliest nutrient-related disease in US history ravaged the American South. Pellagra, a disease caused by a deficiency in niacin and/or tryptophan, is marked by the four "D's": diarrhea, dermatitis that leads to gruesome skin plaques, dementia, and death. At its peak during the Great Depression, pellagra killed nearly 7,000 Southerners a year. Between 1906 and 1940, researchers estimate that the epidemic struck roughly 3 million Americans, killing around 100,000.

The deadly epidemic led to voluntary—and eventually mandatory—fortification of wheat and other cereals with niacin (aka Vitamin B3). By the middle of the century, pellagra nearly vanished from the US. But, decades later, the public health triumph may be backfiring. With Americans' diets more reliant than ever on processed, niacin-fortified foods, the average niacin intake in the US is now nearing what's considered the tolerable upper limit of the nutrient, according to a federal health survey. And an extensive study recently published in Nature Medicine suggests that those excess amounts of niacin may be exacerbating cardiovascular disease, increasing risks of heart attacks, strokes, and death.

The study, led by Stanley Hazen, chair of Cardiovascular and Metabolic Sciences at Cleveland Clinic's Lerner Research Institute, connected high blood levels of a breakdown product of niacin—and to a lesser extent, tryptophan—to an elevated risk of major adverse cardiovascular events (MACE). And this elevated risk appears to be independent of known risk factors for those events, such as high cholesterol.

Enlarge / Stable Diffusion 3 generation with the prompt: studio photograph closeup of a chameleon over a black background. (credit: Stability AI)

On Thursday, Stability AI announced Stable Diffusion 3, an open-weights next-generation image-synthesis model. It follows its predecessors by reportedly generating detailed, multi-subject images with improved quality and accuracy in text generation. The brief announcement was not accompanied by a public demo, but Stability is opening up a waitlist today for those who would like to try it.

Stability says that its Stable Diffusion 3 family of models (which takes text descriptions called "prompts" and turns them into matching images) range in size from 800 million to 8 billion parameters. The size range accommodates allowing different versions of the model to run locally on a variety of devices—from smartphones to servers. Parameter size roughly corresponds to model capability in terms of how much detail it can generate. Larger models also require more VRAM on GPU accelerators to run.

Since 2022, we've seen Stability launch a progression of AI image-generation models: Stable Diffusion 1.4, 1.5, 2.0, 2.1, XL, XL Turbo, and now 3. Stability has made a name for itself as providing a more open alternative to proprietary image-synthesis models like OpenAI's DALL-E 3, though not without controversy due to the use of copyrighted training data, bias, and the potential for abuse. (This has led to lawsuits that are unresolved.) Stable Diffusion models have been open-weights and source-available, which means the models can be run locally and fine-tuned to change their outputs.

Enlarge (credit: Bloomberg / Contributor | Bloomberg)

A judge has dismissed a complaint from a parent and guardian of a girl, now 15, who was sexually assaulted when she was 12 years old after Snapchat recommended that she connect with convicted sex offenders.

According to the court filing, the abuse that the girl, C.O., experienced on Snapchat happened soon after she signed up for the app in 2019. Through its "Quick Add" feature, Snapchat "directed her" to connect with "a registered sex offender using the profile name JASONMORGAN5660." After a little more than a week on the app, C.O. was bombarded with inappropriate images and subjected to sextortion and threats before the adult user pressured her to meet up, then raped her. Cops arrested the adult user the next day, resulting in his incarceration, but his Snapchat account remained active for three years despite reports of harassment, the complaint alleged.

Two years later, at 14, C.O. connected with another convicted sex offender on Snapchat, a former police officer who offered to give C.O. a ride to school and then sexually assaulted her. The second offender is also currently incarcerated, the judge's opinion noted.

Enlarge / Generations from Gemini AI from the prompt, "Paint me a historically accurate depiction of a medieval British king." (credit: @stratejake / X)

On Thursday morning, Google announced it was pausing its Gemini AI image-synthesis feature in response to criticism that the tool was inserting diversity into its images in a historically inaccurate way, such as depicting multi-racial Nazis and medieval British kings with unlikely nationalities.

"We're already working to address recent issues with Gemini's image generation feature. While we do this, we're going to pause the image generation of people and will re-release an improved version soon," wrote Google in a statement Thursday morning.

As more people on X began to pile on Google for being "woke," the Gemini generations inspired conspiracy theories that Google was purposely discriminating against white people and offering revisionist history to serve political goals. Beyond that angle, as The Verge points out, some of these inaccurate depictions "were essentially erasing the history of race and gender discrimination."

Enlarge / A new lens-free and compact system for facial recognition scans a bust of Michelangelo’s David and reconstructs the image using less power than existing 3D-surface imaging systems. (credit: W-C Hsu et al., Nano Letters, 2024)

Facial recognition is a common feature for unlocking smartphones and gaming systems, among other uses. But the technology currently relies upon bulky projectors and lenses, hindering its broader application. Scientists have now developed a new facial recognition system that employs flatter, simpler optics that also require less energy, according to a recent paper published in the journal Nano Letters. The team tested their prototype system with a 3D replica of Michelangelo's famous David sculpture and found it recognized the face as well as existing smartphone facial recognition can.

The current commercial 3D imaging systems in smartphones (like Apple's iPhone) extract depth information via structured light. A dot projector uses a laser to project a pseudorandom beam pattern onto the face of the person looking at a locked screen. It does so thanks to several other built-in components: a collimator, light guide, and special lenses (known as diffractive optical elements, or DOEs) that break the laser beam apart into an array of some 32,000 infrared dots. The camera can then interpret that projected beam pattern to confirm the person's identity.

Packing in all those optical components like lasers makes commercial dot projectors rather bulky, so it can be harder to integrate for some applications such as robotics and augmented reality, as well as the next generation of facial recognition technology. They also consume significant power. So Wen-Chen Hsu, of National Yang Ming Chiao Tung University and the Hon Hai Research Institute in Taiwan, and colleagues turned to ultrathin optical components known as metasurfaces for a potential solution. These metasurfaces can replace bulkier components for modulating light and have proven popular for depth sensors, endoscopes, tomography. and augmented reality systems, among other emerging applications.

Enlarge / Japan's second H3 rocket lifted off from the Tanegashima Space Center at 9:22 am local time Saturday. (credit: Yang Guang/Xinhua via Getty Images)

Japan's new H3 rocket took off Friday on its second test flight; its success is an important milestone for the launch vehicle poised to power nearly all of the Japanese space program's missions into orbit over the next decade.

The 187-foot-tall (57-meter) H3 rocket lifted off from the Tanegashima Space Center in southwestern Japan at 7:22 pm EST Friday (00:22 UTC Saturday) with a dummy payload and two smaller satellites.

Two hydrogen-fueled main engines and a pair of strap-on solid rocket boosters ignited with 1.6 million pounds of thrust, vaulting the H3 rocket off its launch pad on a trajectory east from Tanegashima, then south over the Pacific Ocean. The strap-on boosters burned out and jettisoned about two minutes into the flight, and the core stage's LE-9 engines fired for nearly five minutes.

Enlarge / SpaceX's fully stacked Starship rocket and Super Heavy booster on a launch pad in South Texas. (credit: SpaceX)

One of the largest launch pads at Cape Canaveral Space Force Station will become vacant later this year after the final flight of United Launch Alliance's Delta IV Heavy rocket. SpaceX is looking to make the sprawling facility a new home for the Starship launch vehicle.

The environmental review for SpaceX's proposal to take over Space Launch Complex 37 (SLC-37) at Cape Canaveral is getting underway now, with three in-person public meetings and one virtual meeting scheduled for March to collect comments from local residents, according to a new website describing the plan.

Then, federal agencies, led by the Department of the Air Force, will develop an environmental impact statement to evaluate how Starship launch and landing operations will affect the land, air, and water around SLC-37, which sits on Space Force property on the Atlantic coastline.

Enlarge / Snapshots from three videos generated using OpenAI's Sora.

On Thursday, OpenAI announced Sora, a text-to-video AI model that can generate 60-second-long photorealistic HD video from written descriptions. While it's only a research preview that we have not tested, it reportedly creates synthetic video (but not audio yet) at a fidelity and consistency greater than any text-to-video model available at the moment. It's also freaking people out.

"It was nice knowing you all. Please tell your grandchildren about my videos and the lengths we went to to actually record them," wrote Wall Street Journal tech reporter Joanna Stern on X.

"This could be the 'holy shit' moment of AI," wrote Tom Warren of The Verge.

Enlarge / SpaceX launched a Falcon 9 rocket Wednesday with six missile-tracking satellites for the US military. (credit: SpaceX)

Two prototype satellites for the Missile Defense Agency and four missile-tracking satellites for the US Space Force rode a SpaceX Falcon 9 rocket into orbit Wednesday from Florida's Space Coast.

These satellites are part of a new generation of spacecraft designed to track hypersonic missiles launched by China or Russia and perhaps emerging missile threats from Iran or North Korea, which are developing their own hypersonic weapons.

Hypersonic missiles are smaller and more maneuverable than conventional ballistic missiles, which the US military's legacy missile defense satellites can detect when they launch. Infrared sensors on the military's older-generation missile tracking satellites are tuned to pick out bright thermal signatures from missile exhaust.

Enlarge / No way would Zuckerberg be photographed wearing a Vision Pro, but let's just imagine he's looking at a picture of one in his headset here... (credit: @zuck Instagram | Aurich Lawson)

Since the launch of the Apple Vision Pro, it's not been hard to find countless thoughts and impressions on the headset from professional reviewers and random purchasers. But among all those hot takes, the opinions of Meta CEO Mark Zuckerberg stand out for a few reasons—not least of which is that he and his company have spent years of development time and lost tens of billions of dollars creating the competing Quest headset line.

For that reason alone, Zuckerberg's Instagram-posted thoughts on the Vision Pro can't be considered an impartial take on the device's pros and cons. Still, Zuckerberg's short review included its fair share of fair points, alongside some careful turns of phrase that obscure the Quest's relative deficiencies.

 

View this post on Instagram

 

A post shared by Mark Zuckerberg (@zuck)

Enlarge / Stellantis' EV offerings in the US are few and far between. (credit: Stellantis)

The North American electric vehicle fast charging plug war is over, bar some shouting. This week, Stellantis—which owns Dodge, Ram, Alfa Romeo, Fiat, and others—announced that it will adopt the new SAE J3400 charging standard for its battery electric vehicles beginning in 2025.

As a result, all the major automakers that sell BEVs in North America have made the switch to J3400, a process that began last May when Ford revealed it was ditching the Combined Charging Standard 1 (CCS1) socket for what was then called the North American Charging Standard (NACS).

In fact, the technical changes between the two are mild—NACS, now known as J3400, uses the same electronic communication protocols as CCS1. The real draw for Ford, and for each automaker who announced the switch since, has been negotiated access for their customers to Tesla's Supercharger network, which far outstrips the CCS1 infrastructure in the US and Canada.

Enlarge / File photo of an X-37B spaceplane. (credit: Boeing)

It turns out some of the informed speculation about the US military's latest X-37B spaceplane mission was pretty much spot-on.

When the semi-classified winged spacecraft launched on December 28, it flew into orbit on top of a SpaceX Falcon Heavy rocket, which is much larger than the Atlas V and Falcon 9 rockets used to launch the X-37B on its previous missions.

This immediately sparked speculation that the X-37B would reach higher altitudes than its past flights, which remained in low-Earth orbit at altitudes of a few hundred miles. A discovery from Tomi Simola, a satellite tracking hobbyist living near Helsinki, Finland, appears to confirm this suspicion.