- Human versus autonomous car race ends before it begins
A2RL admits that this is a hard problem, and that's refreshing.
TOKYO—Racing is hard. It's hard on the teams, it's hard on the owner's bank account, it's hard on the cars, and it's especially hard on the drivers. Driving at the edge for a few hours in a vehicle cockpit that's only slightly wider than your frame can take a toll.
The A2RL (Abu Dhabi Autonomous Racing League) removes one of those elements from its vehicles but, in doing so, creates a whole new list of complexities. Say goodbye to the human driver and hello to 95 kilograms of computers and a whole suite of sensors. That setup was poised to be part of a demonstration "race" against former F1 driver Daniil Kvyat at Suzuka Circuit in Japan during the Super Formula season finale.
But again, racing is hard, and replacing humans doesn't change that. The people who run and participate in A2RL are aware of this, and while many organizations have made it a sport of overselling AI, A2RL is up-front about the limitations of the current state of the technology. One example of the technology's current shortcomings: The vehicles can't swerve back and forth to warm up the tires.
- Ars Technica’s top 20 video games of 2024
A relatively light year still had its fair share of interactive standouts.
When we introduced last year's annual list of the best games in this space, we focused on how COVID delays led to a 2023 packed with excellent big-name blockbusters and intriguing indies that seemed to come out of nowhere. The resulting flood of great titles made it difficult to winnow the year's best down to our traditional list of 20 titles.
In 2024 we had something close to the opposite problem. While there were definitely a few standout titles that were easy to include on this year's list (Balatro, UFO 50, and Astro Bot likely chief among them), rounding out the list to a full 20 proved more challenging than it has in perhaps any other year during my tenure at Ars Technica (way back in 2012!). The games that ended up on this year's list are all strong recommendations, for sure. But many of them might have had more trouble making a Top 20 list in a packed year like 2023.
We'll have to wait to see if the release calendar seesaws back to a quality-packed one in 2025, but the forecast for big games like Civilization 7, Avowed, Doom: The Dark Ages, Grand Theft Auto 6, and many, many more has us thinking that it might. In the meantime, here are our picks for the 20 best games that came out in 2024, in alphabetical order.
- Green sea turtle gets relief from “bubble butt” syndrome thanks to 3D printing
Boat collision left Charlotte stranded at the surface and in danger of predation.
Charlotte, a green sea turtle, was hit by a boat back in 2008. This left it with an affliction colloquially referred to as the “bubble butt,” a kind of floating syndrome that makes it impossible for a turtle to dive. Most sea turtles suffering from issues like this simply die at sea, since the condition leaves them stranded at the surface where they can’t forage, sleep, and avoid predators like sharks. But fate had other plans for Charlotte.
Charlotte didn’t end up as a shark’s lunch and didn’t starve to death floating helplessly in the ocean. Instead, it got rescued shortly after the boat accident and eventually found a home at Mystic Aquarium in Stonington, Connecticut, where it received professional care. That was the first time Charlotte got lucky. The second time came when a collaboration formed: Adia, a company specializing in 3D-printing solutions; Formlabs, one of the world’s leading manufacturers of 3D printers; and New Balance Athletic, a sportswear giant based in Boston. This team chose Charlotte as a technology showcase, which basically turned the turtle into an Oscar Pistorius of the sea—just without the criminal conviction.
Weights and diet
Sea turtles are marine reptiles, which means they don’t have gills like fish—they need air to breathe. The lungs also play a key role in their buoyancy regulation system, which allows them to rest for extended periods of time at the sea floor or float at a precisely chosen depth. A sea turtle can precisely choose the depth at which it achieves neutral buoyancy by inhaling the exactly right volume of air.
- Exploring an undersea terrain sculpted by glaciers and volcanoes
Researchers found a fossilized seascape while studying the impact of a volcanic eruption.
On May 2, 2008, the Chaitén volcano in Chile awoke with unexpected fury after more than 9,000 years of dormancy. The eruption blasted rocks and ash a dozen miles into the air, and then heavy rainfall swept the fallen debris up in immense mudflows. A river of rubble carved a destructive path through the nearby town of Chaitén before surging into the sea. The town, practically split in two by the torrent that cut through its middle, was evacuated as ash blanketed over 200,000 square kilometers of surrounding land.
While the terrestrial aftermath was plain to see, captured by both the local media and satellites, the impact on the sea was unknown.
Mists shroud the hillsides at the entrance to Chaitén Bay, as seen from aboard the Schmidt Ocean Institute’s <em>Falkor (too)</em> research vessel. Credit: Alex Ingle / Schmidt Ocean InstituteThe eruption released over 750 billion liters of lava—enough to fill more than 300,000 Olympic-sized swimming pools—mainly in the form of rock fragments. The debris flowed through rivers into the Northern Patagonian Sea, just six miles away.
- 12 days of OpenAI: The Ars Technica recap
Did OpenAI's big holiday event live up to the billing?
Over the past 12 business days, OpenAI has announced a new product or demoed an AI feature every weekday, calling the PR event "12 days of OpenAI." We've covered some of the major announcements, but we thought a look at each announcement might be useful for people seeking a comprehensive look at each day's developments.
The timing and rapid pace of these announcements—particularly in light of Google's competing releases—illustrates the intensifying competition in AI development. What might normally have been spread across months was compressed into just 12 business days, giving users and developers a lot to process as they head into 2025.
Humorously, we asked ChatGPT what it thought about the whole series of announcements, and it was skeptical that the event even took place. "The rapid-fire announcements over 12 days seem plausible," wrote ChatGPT-4o, "But might strain credibility without a clearer explanation of how OpenAI managed such an intense release schedule, especially given the complexity of the features."
- Horizon: Zero Dawn gets the graphical remaster a modern classic deserves
This is how to do a remaster.
At their best, "remastered" video games keep terrific older titles viable on new generations of hardware and for new generations of fans. At their worst, they can feel like a cash-in.
So it was with some trepidation that I recently fired up the "remastered" Horizon: Zero Dawn, a game which won me over years ago with its PS4 version due to the simple fact that it was ONE OF THE BEST VIDEO GAMES OF ALL TIME and featured ONE OF THE BEST PROTAGONISTS OF ALL TIME in one of the BEST STORIES OF ALL TIME. (Yes, I like superlatives, which are some of the BEST WORDS OF ALL TIME. But the game world really was terrific.) Even my kids were won over, playing through the game and its sequel multiple times.
The game tells the story of a future Earth long after catastrophe—in the form of an autonomous robotic swarm—has ruined the planet. But it's not mere dystopia, though one does come across many wrecked and overgrown spaces from that earlier age. Horizon instead focuses on how humans, having lost most of their past knowledge, rebuilt a world in tribal fashion, a world populated by animal-inspired machines. The game's story operates ambitiously in two timelines and features massive killer robots, cults, and mad Sun Kings, all set against the gorgeous background of the American West.
- Rocket Report: ULA has a wild idea; Starliner crew will stay in orbit even longer
ULA's Vulcan rocket is at least several months away from flying again, and Stoke names its engine.
Welcome to Edition 7.24 of the Rocket Report! This is the last Rocket Report of the year, and what a year it's been. So far, there have been 244 rocket launches to successfully reach orbit this year, a record for annual launch activity. And there are still a couple of weeks to go before the calendar turns to 2025. Time is running out for Blue Origin to launch its first heavy-lift New Glenn rocket this year, but if it flies before January 1, it will certainly be one of the top space stories of 2024.
As always, we welcome reader submissions. If you don't want to miss an issue, please subscribe using the box below (the form will not appear on AMP-enabled versions of the site). Each report will include information on small-, medium-, and heavy-lift rockets as well as a quick look ahead at the next three launches on the calendar.
Corkscrew in the sky. A Japanese space startup said its second attempt to launch a rocket carrying small satellites into orbit had been terminated minutes after liftoff Wednesday and destroyed itself again, nine months after the company’s first launch attempt in an explosion, the Associated Press reports. The startup that developed the rocket, named Space One, launched the Kairos rocket from a privately owned coastal spaceport in Japan's Kansai region. Company executive and space engineer Mamoru Endo said an abnormality in the first stage engine nozzle or its control system is likely to have caused an unstable flight of the rocket, which started spiraling in mid-flight and eventually destroyed itself about three minutes after liftoff, using its autonomous safety mechanism.
- OpenAI announces o3 and o3-mini, its next simulated reasoning models
o3 matches human levels on ARC-AGI benchmark, and o3-mini exceeds o1 at some tasks.
On Friday, during Day 12 of its "12 days of OpenAI," OpenAI CEO Sam Altman announced its latest AI "reasoning" models, o3 and o3-mini, which build upon the o1 models launched earlier this year. The company is not releasing them yet but will make these models available for public safety testing and research access today.
The models use what OpenAI calls "private chain of thought," where the model pauses to examine its internal dialog and plan ahead before responding, which you might call "simulated reasoning" (SR)—a form of AI that goes beyond basic large language models (LLMs).
The company named the model family "o3" instead of "o2" to avoid potential trademark conflicts with British telecom provider O2, according to The Information. During Friday's livestream, Altman acknowledged his company's naming foibles, saying, "In the grand tradition of OpenAI being really, truly bad at names, it'll be called o3."
- Startup set to brick $800 kids robot is trying to open source it first
Most owners still won't be refunded for the emotional support toy.
Earlier this month, startup Embodied announced that it is going out of business and taking its Moxie robot with it. The $800 robots, aimed at providing emotional support for kids ages 5 to 10, would soon be bricked, the company said, because they can’t perform their core features without the cloud. Following customer backlash, Embodied is trying to create a way for the robots to live an open sourced second life.
Embodied CEO Paolo Pirjanian shared a document via a LinkedIn blog post today saying that people who used to be part of Embodied’s technical team are developing a “potential” and open source way to keep Moxies running. The document reads:
This initiative involves developing a local server application (‘OpenMoxie’) that you can run on your own computer. Once available, this community-driven option will enable you (or technically inclined individuals) to maintain Moxie’s basic functionality, develop new features, and modify her capabilities to better suit your needs—without reliance on Embodied’s cloud servers.
The notice says that after releasing OpenMoxie, Embodied plans to release “all necessary code and documentation” for developers and users.
- Man who claims he invented bitcoin faces prison after filing $1.1 trillion suit
UK judge issues 1-year suspended prison sentence as Wright hides in Asia.
Craig Wright, the man who claims he invented bitcoin and has been filing lawsuits asserting intellectual property rights, was sentenced to a year in prison yesterday for committing contempt of court.
The sentence is suspended and can be enforced if Wright continues violating court rulings—but he may be able to avoid imprisonment by staying away from countries that have extradition agreements with the UK. Wright defied an order to attend a court hearing in person this week and said he is in Asia.
Wright "was sentenced for contempt of court on Thursday" for bringing a 911 billion pound ($1.1 trillion) lawsuit "against Twitter founder Jack Dorsey's payments company Block in Britain," Reuters wrote.
- The next two FIFA Women’s World Cups will only air on Netflix
Let's hope Netflix fixes its live buffering problems beforehand.
If you want to watch the next two FIFA Women’s World Cups in the US, you’ll need a Netflix subscription.
FIFA confirmed the news today, marking an unexpected change for the sports event, which has historically played on free-to-air broadcast channels. The shift to a streaming platform inevitably makes it more costly and hurts viewer accessibility, while likely injecting FIFA with a lot of cash.
Netflix and FIFA haven’t said how much Netflix is paying for exclusive airing rights. But Netflix and other streaming services have been paying out hefty, sometimes record-setting sums to air live sporting events as the company seeks to earn more revenue from commercials and draw more viewers. Netflix, for example, paid $5 billion to swipe the World Wrestling Entertainment’s weekly RAW program from the USA cable network for 10 years, starting next month.
- New AA-powered AirTag case promises 10-year lifespan
Aluminum AirTag case replaces coin cell with 2 AA Lithium batteries for extended lifespan.
On Wednesday, Elevation Lab announced TimeCapsule, a new $20 battery case purported to extend Apple AirTag battery life from one year to 10 years. The product replaces the standard CR2032 coin cell battery in the Bluetooth-based location tracker with two AA batteries to provide extended power capacity.
The TimeCapsule case requires users to remove their AirTag's original back plate and battery, then place the Apple device onto contact points inside the waterproof enclosure. The company recommends using Energizer Ultimate Lithium AA batteries, which it claims provide 14 times more power capacity than the stock coin cell battery configuration.
The CNC-machined aluminum case is aimed at users who place AirTags in vehicles, boats, or other applications where regular battery changes prove impractical. The company sells the TimeCapsule through its website and Amazon.
- Google will apparently offer “AI Mode” right on its main search page
The most direct push for Google's Gemini chat so far.
Google will soon take more steps to make AI a part of search, exposing more users to its Gemini agent, according to recent reports and app teardowns.
"AI Mode," shown at the top left of the web results page and inside the Google app, will provide an interface similar to a Gemini AI chat, according to The Information.
This tracks with a finding from Android Authority earlier this month, which noted a dedicated "AI mode" button inside an early beta of the Google app. This shortcut also appeared on Google's Android search widget, and a conversation history button was added to the Google app. Going even deeper into the app, 9to5Google found references to "aim" (AI mode) and "ai_mode" which suggest a dedicated tab in the Google app, with buttons for speaking to an AI or sending it pictures.
- Automakers excoriated by Senators for fighting right-to-repair
OEMs also called out for selling data they collect on drivers.
Yesterday, US Senators Jeff Merkley (D-OR), Elizabeth Warren (D-MA), and Joshua Hawley (R-MO) sent letters to the heads of Ford, General Motors, and Tesla, as well as the US heads of Honda, Hyundai, Nissan, Stellantis, Subaru, Toyota, and Volkswagen, excoriating them over their opposition to the right-to-repair movement.
"We need to hit the brakes on automakers stealing your data and undermining your right-to-repair," said Senator Merkley in a statement to Ars. "Time and again, these billionaire corporations have a double standard when it comes to your privacy and security: claiming that sharing vehicle data with repair shops poses cybersecurity risks while selling consumer data themselves. Oregon has one of the strongest right-to-repair laws in the nation, and that’s why I’m working across the aisle to advance efforts nationwide that protect consumer rights."
Most repairs aren’t at dealerships
The Senators point out that 70 percent of car parts and services currently come from independent outlets, which are seen as trustworthy and providing good value for money, "while nearly all dealerships receive the worst possible rating for price."
- Louisiana bars health dept. from promoting flu, COVID, mpox vaccines: Report
Staff worried they were no longer helping people and the ban would lead to deaths.
Louisiana's health department has been barred from advertising or promoting vaccines for flu, COVID-19, and mpox, according to reporting by NPR, KFF Health News, and New Orleans Public Radio WWNO.
Their investigative report—based on interviews with multiple health department employees who spoke on the condition of anonymity for fear of retaliation—revealed that employees were told of the startling policy change in meetings in October and November and that the policy would be implemented quietly and not put into writing.
Ars Technica has contacted the health department for comment and will update this post with any new information.
- The AI war between Google and OpenAI has never been more heated
Potentially groundbreaking AI releases have been coming in fast, sending experts' heads spinning.
Over the past month, we've seen a rapid cadence of notable AI-related announcements and releases from both Google and OpenAI, and it's been making the AI community's head spin. It has also poured fuel on the fire of the OpenAI-Google rivalry, an accelerating game of one-upmanship taking place unusually close to the Christmas holiday.
"How are people surviving with the firehose of AI updates that are coming out," wrote one user on X last Friday, which is still a hotbed of AI-related conversation. "in the last <24 hours we got gemini flash 2.0 and chatGPT with screenshare, deep research, pika 2, sora, chatGPT projects, anthropic clio, wtf it never ends."
Rumors travel quickly in the AI world, and people in the AI industry had been expecting OpenAI to ship some major products in December. Once OpenAI announced "12 days of OpenAI" earlier this month, Google jumped into gear and seemingly decided to try to one-up its rival on several counts. So far, the strategy appears to be working, but it's coming at the cost of the rest of the world being able to absorb the implications of the new releases.
- VPN used for VR game cheat sells access to your home network
Big Mama VPN tied to network which offers access to residential IP addresses.
In the hit virtual reality game Gorilla Tag, you swing your arms to pull your primate character around—clambering through virtual worlds, climbing up trees and, above all, trying to avoid an infectious mob of other gamers. If you’re caught, you join the horde. However, some kids playing the game claim to have found a way to cheat and easily “tag” opponents.
Over the past year, teenagers have produced video tutorials showing how to side-load a virtual private network (VPN) onto Meta’s virtual reality headsets and use the location-changing technology to get ahead in the game. Using a VPN, according to the tutorials, introduces a delay that makes it easier to sneak up and tag other players.
While the workaround is likely to be an annoying but relatively harmless bit of in-game cheating, there’s a catch. The free VPN app that the video tutorials point to, Big Mama VPN, is also selling access to its users’ home internet connections—with buyers essentially piggybacking on the VR headset’s IP address to hide their own online activity.
- Why AI language models choke on too much text
Compute costs scale with the square of the input size. That's not great.
Large language models represent text using tokens, each of which is a few characters. Short words are represented by a single token (like "the" or "it"), whereas larger words may be represented by several tokens (GPT-4o represents "indivisible" with "ind," "iv," and "isible").
When OpenAI released ChatGPT two years ago, it had a memory—known as a context window—of just 8,192 tokens. That works out to roughly 6,000 words of text. This meant that if you fed it more than about 15 pages of text, it would “forget” information from the beginning of its context. This limited the size and complexity of tasks ChatGPT could handle.
Today’s LLMs are far more capable:
- We’re about to fly a spacecraft into the Sun for the first time
"Quite simply, we want to find the birthplace of the solar wind."
Almost no one ever writes about the Parker Solar Probe anymore.
Sure, the spacecraft got some attention when it launched. It is, after all, the fastest moving object that humans have ever built. At its maximum speed, goosed by the gravitational pull of the Sun, the probe reaches a velocity of 430,000 miles per hour, or more than one-sixth of 1 percent the speed of light. That kind of speed would get you from New York City to Tokyo in less than a minute.
And the Parker Solar Probe also has the distinction of being the first NASA spacecraft named after a living person. At the time of its launch, in August 2018, physicist Eugene Parker was 91 years old.
- Not to be outdone by OpenAI, Google releases its own “reasoning” AI model
Gemini 2.0 Flash Thinking is Google's take on so-called AI reasoning models.
It's been a really busy month for Google as it apparently endeavors to outshine OpenAI with a blitz of AI releases. On Thursday, Google dropped its latest party trick: Gemini 2.0 Flash Thinking Experimental, which is a new AI model that uses runtime "reasoning" techniques similar to OpenAI's o1 to achieve "deeper thinking" on problems fed into it.
The experimental model builds on Google's newly released Gemini 2.0 Flash and runs on its AI Studio platform, but early tests conducted by TechCrunch reporter Kyle Wiggers reveal accuracy issues with some basic tasks, such as incorrectly counting that the word "strawberry" contains two R's.
These so-called reasoning models differ from standard AI models by incorporating feedback loops of self-checking mechanisms, similar to techniques we first saw in early 2023 with hobbyist projects like "Baby AGI." The process requires more computing time, often adding extra seconds or minutes to response times. Companies have turned to reasoning models as traditional scaling methods at training time have been showing diminishing returns.