Connexion
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.
(credit: Google)
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: Getty Images)
A federal Cyber Safety Review Board has issued its report on what led to last summer's capture of hundreds of thousands of emails by Chinese hackers from cloud customers, including federal agencies. It cites "a cascade of security failures at Microsoft" and finds that "Microsoft's security culture was inadequate" and needs to adjust to a "new normal" of cloud provider targeting.
The report, mandated by President Biden in the wake of the far-reaching intrusion, details the steps that Microsoft took before, during, and after the breach and in each case finds critical failure. The breach was "preventable," even though it cites Microsoft as not knowing precisely how Storm-0558, a "hacking group assessed to be affiliated with the People's Republic of China," got in.
"Throughout this review, the board identified a series of Microsoft operational and strategic decisions that collectively points to a corporate culture that deprioritized both enterprise security investments and rigorous risk management," the report reads.
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 (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: Getty | Peter Cade)
Wild migratory birds likely spread a deadly strain of bird flu to dairy cows in Texas and Kansas, state and federal officials announced this week.
It is believed to be the first time the virus, a highly pathogenic avian influenza (HPAI), has been found in cows in the US. Last week, officials in Minnesota confirmed finding an HPAI case in a young goat, marking the first time the virus has been found in a domestic ruminant in the US.
According to the Associated Press, officials with the Texas Animal Health Commission confirmed the flu virus is the Type A H5N1 strain, which has been ravaging bird populations around the globe for several years. The explosive ongoing spread of the virus has led to many spillover events into mammals, making epidemiologists anxious that the virus could adapt to spread widely in humans.
Enlarge / If you've formatted a disk in Windows in the last 30 years, you may have come across this dialog box. (credit: Andrew Cunningham)
Windows 11 has done a lot to update and modernize long-neglected parts of Windows' user interface, including many Settings menus and venerable apps like Notepad and Paint. But if you dig deep enough, you'll still find parts of the user interface that look and work like they did in the mid-'90s, either for compatibility reasons or because no one ever thought to go back and update them.
Former Microsoft programmer Dave Plummer shared some history about one of those finely aged bits: the Format dialogue box, which is still used in fully updated Windows 11 installs to this day when you format a disk using Windows Explorer.
Plummer says he wrote the Format dialog in late 1994, when the team was busy porting the user interface from the consumer-focused Windows 95 (released in mid-1995) to the more-stable but more resource-intensive Windows NT (NT 4.0, released in mid-1996, was the first to use the 95-style UI).
Enlarge / One day I will own griffons in such spectacular fashion. But I'm currently carrying a too-heavy backpack and clipped through a hut wall.
With all due respect to the Capcom team, which poured itself into Dragon’s Dogma 2 and deserves praise, raises, and time off, let me get right to it: I love this game for how dumb it is.
I mean "dumb" in the way most heavy metal lyrics are dumb, but you find yourself rocking out nonetheless. Dumb like when you laugh uncontrollably at the sight of someone getting conked on the head and falling over backward. Dumb as in the silliest bits of Monty Python and the Holy Grail, just nowhere near as self-aware (unless, due to translation issues, this game actually is self-aware, then I apologize).
Dragon’s Dogma 2 (DD2) reminds me of playing another huge, dumb, enjoyable game: The Elder Scrolls V: Skyrim. Not the first time you play through it, though. I’m talking about the second or third run-through (or that 100-plus-hour save in which you refuse to finish the game), and your admiration of this huge, rich world gives way to utter ridiculousness. You one-shot dragons with your broken stealth-archer build, you put buckets on the heads of NPCs to rob them, and you marvel at how the most effective fast travel is horse tilting. You lunge into possibilities, choose chaos, and appreciate all the ways you can do so.
Enlarge / The Android 15 logo. This is "Android V," if you can't tell from the logo. (credit: Google)
Android 15 continues its march toward release with the Android 15 Developer Preview 2. Android 15 won't be out until around October, but the first preview shipped a month ago. It's time for another one!
Android's satellite messaging support has been in the works for about a year now, and it sounds like Android 15 is going to launch the feature for apps. The new OS is including notifications and better status bar indicators for when you're connected to space. A "NonTerrestrialNetwork" API will let apps know when they're limited to barely there satellite connectivity. Google says Android 15 will let third-party SMS and MMS applications tap into the satellite connectivity APIs, but enhanced messaging with RCS support will be limited to "preloaded" applications only. It seems incredible that Google doesn't have public APIs for third-party RCS apps, but here's your confirmation that Android 15 will continue locking out Play Store apps from RCS.
Android 15's new satellite messaging UI. [credit: Google ]
Android's PDF support can be all over the place depending on what device you have, so Android 15 is including making some big improvements to the built-in PDF render. First it's going to end up as a module so it can be updated via the Play Store. Google says this Android 15 version is getting "advanced features such as rendering password-protected files, annotations, form editing, searching, and selection with copy."
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 / PSVR2 (left) next to the original PSVR. (credit: Kyle Orland / Ars Technica)
It looks like Sony's PlayStation VR2 is not living up to the company's sales expectations just over a year after it first hit the market. Bloomberg reports that the PlayStation-maker has stopped producing new PSVR2 units as it tries to clear out a growing backlog of unsold inventory.
Bloomberg cites "people familiar with [Sony's] plans" in reporting that PSVR2 sales have "slowed progressively" since its February 2023 launch. Sony has produced "well over 2 million" units of the headset, compared to what tracking firm IDC estimates as just 1.69 million unit shipments to retailers through the end of last year. The discrepancy has caused a "surplus of assembled devices... throughout Sony’s supply chain," according to Bloomberg's sources.
IDC estimates a quarterly low of 325,000 PSVR2 units shipped in the usually hot holiday season, compared to a full 1.3 million estimated holiday shipments for Meta's then-new Quest 3 headset, which combined with other Quest products to account for over 3.7 million estimated sales for the full year.
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 / 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 (credit: Aurich Lawson)
Microsoft’s decision to ease off its 23-year competition with Sony and Nintendo over supremacy in games hardware has opened a path for Japan’s return as the world’s undisputed home of the console.
The prospect of a new, less internationalized era of console wars has raised hopes of happier times for the Japanese survivors but has also caused analysts and investors to revisit the question of how much longer the whole genre of dedicated games machines will continue to exist.
Microsoft head of gaming Phil Spencer last month revealed plans to release what would previously have been exclusively Xbox games for use on rival platforms, as part of a new focus on cloud-based gaming.
Enlarge / A view of the Centers for Disease Control and Prevention headquarters in Atlanta. (credit: Getty | Nathan Posner)
COVID-19 is becoming more like the flu and, as such, no longer requires its own virus-specific health rules, the Centers for Disease Control and Prevention said Friday alongside the release of a unified "respiratory virus guide."
In a lengthy background document, the agency laid out its rationale for consolidating COVID-19 guidance into general guidance for respiratory viruses—including influenza, RSV, adenoviruses, rhinoviruses, enteroviruses, and others, though specifically not measles. The agency also noted the guidance does not apply to health care settings and outbreak scenarios.
"COVID-19 remains an important public health threat, but it is no longer the emergency that it once was, and its health impacts increasingly resemble those of other respiratory viral illnesses, including influenza and RSV," the agency wrote.
Enlarge / The Moderna Spikevax COVID-19 vaccine is shown at a CVS in 2023. (credit: Getty | Melissa Phillip/Houston Chronicle )
People ages 65 and up should get another dose of a COVID-19 vaccine this spring, given the age group's higher risk of severe disease and death from the pandemic virus, the Centers for Disease Control and Prevention announced Wednesday.
Earlier today, an advisory committee for the CDC voted overwhelmingly in favor of recommending the spring booster dose. And late this afternoon, CDC Director Mandy Cohen signed off on the recommendation, allowing boosting to begin.
"Today’s recommendation allows older adults to receive an additional dose of this season’s COVID-19 vaccine to provide added protection," Cohen said in a statement. "Most COVID-19 deaths and hospitalizations last year were among people 65 years and older. An additional vaccine dose can provide added protection that may have decreased over time for those at highest risk."
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 / On February 9, Ford suspended shipments of model-year 2024 F-150 Lightning electric pickup trucks to dealerships. (credit: Ford)
Ford has temporarily suspended new shipments of its model-year 2024 F-150 Lightning electric pickup truck to dealerships from its factory in Michigan. It issued the stop-ship order in early February, according to Automotive News, but it's not clear when Ford will lift the measure.
"We started shipping the first newly designed F-150 pickups to dealers last week. MY24 Lightnings started shipping last month. We expect to ramp up shipments in the coming weeks as we complete thorough launch quality checks to ensure these new F-150s meet our high standards and delight customers," Ford told Ars.
Exactly what the problem is with the MY24 F-150 Lightnings is unknown, and Ford says it will "ramp shipments once quality checks are completed."
For some, having to work from home during the COVID-19 pandemic was stressful. Parents balanced job duties while caring for children. Some struggled to set up a home office and adjust to new tools, like video conferencing. Lonely workdays at home added to social isolation. The line between work and life blurred.
For others, working from home was a boon—comfort, convenience, flexibility, no commuting or rush-hour traffic, no office-environment distractions. When the acute aspects of the pandemic receded, some who at first struggled began to settle into a work-from-home (WFH) groove and appreciated the newfound flexibility.
Then, bosses began calling their employees back to the office. Many made the argument that the return-to-office (RTO) policies and mandates were better for their companies; workers are more productive at the office, and face-to-face interactions promote collaboration, many suggested. But there's little data to support that argument. Pandemic-era productivity is tricky to interpret, given that the crisis disrupted every aspect of life. Research from before the pandemic generally suggested remote work improves worker performance—though it often included workers who volunteered to WFH, potentially biasing the finding.
The game includes re-creations of classic battles from the original trilogy... [credit: Lucasfilm ]
There was once a time of outstanding Star Wars computer games before Disney made a deal with publisher EA that led to a plethora of lukewarmly received multiplayer games-as-a-service and mobile titles. (And Jedi: Fallen Order, of course.) Now, a cornerstone of that era is set to be revived in the form of the Star Wars Battlefront Classic Collection.
Announced this morning, the collection includes two beloved LucasArts-developed online multiplayer shooters from the 2000s: Star Wars Battlefront and Star Wars Battlefront 2. Although it debuted during a Nintendo livestream, the collection won't just be coming out on Nintendo Switch—it's coming to PC, PlayStation, and Xbox, too.
The collection will include "restored online play for up to 64 players" of all the previous Galactic Conquest and Campaign mode content from both games on all platforms, including bonus maps and extras that were previously exclusive to specific consoles. Additionally, the Hero Assault mode will now be playable on all ground maps.
Enlarge / The Android 15 logo. This is "Android V," if you can't tell from the logo. (credit: Google)
It's that time of year again. Android is going to start its ~8-month-long beta process with the release of a new major OS version. The Android 15 Developer Preview is out today for the Pixel 6, 7, and 8, Pixel Fold, and Pixel Tablet. This release should mark the end of major OS support for the Pixel 5 and 5a series.
So what's new? It's hard to know too much with only the simple text descriptions we're getting, but we have a few bullet points. "Partial screen sharing" will let users share or record individual app windows instead of the entire screen. Phones don't have much of a difference between an app window and a full screen, but it would be nice if this blocked incoming notifications from showing up on your screen share. It would also be nice for tablets.
Android is surfacing an API that supports the Linux kernel's fs-verity feature. This will let you store a read-only file on a read-write file system and cryptographically sign it to ensure it hasn't been maliciously tampered with. Google apparently wants app developers to use this, saying, "This leads to enhanced security, protecting against potential malware or unauthorized file modifications that could compromise your app's functionality or data."
Enlarge / The Gemini 1.5 logo, released by Google. (credit: Google)
One week after its last major AI announcement, Google appears to have upstaged itself. Last Thursday, Google launched Gemini Ultra 1.0, which supposedly represented the best AI language model Google could muster—available as part of the renamed "Gemini" AI assistant (formerly Bard). Today, Google announced Gemini Pro 1.5, which it says "achieves comparable quality to 1.0 Ultra, while using less compute."
While Ultra 1.0 is possibly still better than Pro 1.5 (what even are we saying here), Ultra was presented as a key selling point of its "Gemini Advanced" tier of its Google One subscription service. And now it's looking a lot less advanced than seven days ago. All this is on top of the confusing name-shuffling Google has been doing recently. (Just to be clear—although it's not really clarifying at all—the free version of Bard/Gemini currently uses the Pro 1.0 model. Got it?)
Google claims that Gemini 1.5 represents a new generation of LLMs that "delivers a breakthrough in long-context understanding," and that it can process up to 1 million tokens, "achieving the longest context window of any large-scale foundation model yet." Tokens are fragments of a word. The first part of the claim about "understanding" is contentious and subjective, but the second part is probably correct. OpenAI's GPT-4 Turbo can reportedly handle 128,000 tokens in some circumstances, and 1 million is quite a bit more—about 700,000 words. A larger context window allows for processing longer documents and having longer conversations. (The Gemini 1.0 model family handles 32,000 tokens max.)
Enlarge (credit: Getty Images)
A core developer of Nginx, currently the world's most popular web server, has quit the project, stating that he no longer sees it as "a free and open source project… for the public good." His fork, freenginx, is "going to be run by developers, and not corporate entities," writes Maxim Dounin, and will be "free from arbitrary corporate actions."
Dounin is one of the earliest and still most active coders on the open source Nginx project and one of the first employees of Nginx, Inc., a company created in 2011 to commercially support the steadily growing web server. Nginx is now used on roughly one-third of the world's web servers, ahead of Apache.
Nginx Inc. was acquired by Seattle-based networking firm F5 in 2019. Later that year, two of Nginx's leaders, Maxim Konovalov and Igor Sysoev, were detained and interrogated in their homes by armed Russian state agents. Sysoev's former employer, Internet firm Rambler, claimed that it owned the rights to Nginx's source code, as it was developed during Sysoev's tenure at Rambler (where Dounin also worked). While the criminal charges and rights do not appear to have materialized, the implications of a Russian company's intrusion into a popular open source piece of the web's infrastructure caused some alarm.
Enlarge / The world's most expensive domino set. (credit: Aurich Lawson)
It's been a busy time in the Xbox rumor mill of late. Last weekend, the Verge reported that Microsoft was considering launching a version of Bethesda's upcoming Indiana Jones and the Great Circle on PlayStation 5, alongside plans to port last year's Hi-Fi Rush to other consoles. That same weekend, Xbox Eras published more lightly sourced rumors suggesting that prominent Xbox exclusive Starfield would be getting a PS5 port.
While Microsoft hasn't directly commented on these reports, Xbox chief Phil Spencer wrote on social media that Microsoft is "planning a business update event for next week, where we look forward to sharing more details with you about our vision for the future of Xbox."
The churning rumor mill has set off something of an existential crisis among some Xbox superfans, content creators, and influencers, who are worried that Microsoft is planning to essentially abandon their favored console. "Genuinely feel terrible for convincing my sister to get an Xbox instead of a PS5," XboxYoda posted in a representative social media take. "Like I actually feel like I let her down... ."
Enlarge / Americans love pickup trucks, but most pickup-truck loving Americans are not ready to go electric yet. Meanwhile, there's almost nothing to buy if you want a smaller, cheaper EV. (credit: Ford)
For the last two years, a small "skunkworks" at the Ford Motor Company has been working on a low-cost electric vehicle platform, according to Ford CEO Jim Farley. Farley revealed the existence of this new platform during the automaker's quarterly financial results call with investors on Tuesday evening. The company is rethinking its electrification strategy, having now faced up to the reality that the current crop of EVs are too expensive for mass-market adoption to take off.
Ford was early to market with its Mustang Mach-E crossover, itself the product of a skunkworks-style development process: an internal group called Team Edison, formed to add some excitement to what was originally going to be a more boring compliance car. The team also took the bold step of making a fully electric version of the country's bestselling vehicle, the F-150 pickup truck.
Demand for the electric F-150 Lightning appeared strong, but a series of price hikes has resulted in really expensive trucks languishing on dealer forecourts and Ford cutting production shifts to reduce output. The Mustang Mach-E is still selling, although with barely any growth year on year.
Enlarge (credit: Gigabyte)
Nvidia launched three new GPUs last month, part of a Super overhaul of the RTX 40-series designed to improve the value of the company's $600-and-up graphics cards.
But today, the company is quietly doing something that it hasn't done in over four years: launching a sub-$200 graphics card. As spotted by TechPowerUp, Nvidia partners like Gigabyte have begun officially announcing a 6GB version of the old RTX 3050 graphics card, albeit with less memory and memory bandwidth, fewer CUDA cores, and lower power requirements.
The announcement follows a few days of leaked retail listings, which generally point to an MSRP of roughly $179 for the new-old card. This would make it Nvidia's first sub-$200 graphics card launch since the GeForce GTX 1650 Super came out in late 2019, a four-year gap caused partially by a cryptocurrency- and pandemic-fueled GPU shortage that lasted from late 2020 into mid-to-late 2022.
Enlarge / The three rulebooks for "fantastic medieval wargames" that started it all, released at some point in late January 1974, as seen in Dungeons & Dragons Art & Arcana: A Visual History. (credit: Wizards of the Coast/Ten Speed Press)
"We have just fromed [sic] Tactical Studies Rules, and we wish to let the wargaming community know that a new line of miniature rules is available."
With this letter, written by Gary Gygax to wargaming zine publisher Jim Lurvey, one of the founders of what would become TSR, announced that a January 1974 release for Dungeons & Dragons was forthcoming. This, plus other evidence compiled by Jon Peterson (as pointed out by the Grognardia blog), points to the last Sunday of January 1974 as the best date for the "anniversary" of D&D. The first sale was in "late January 1974," Gygax later wrote, and on the last Sunday of January 1974, Gygax invited potential customers to drop by his house in the afternoon to try it out.
You could argue whether a final draft, printing, announcement, sale, or first session counts as the true "birth" of D&D, but we have to go with something, and Peterson's reasoning seems fairly sound. Gygax's memory, and a documented session at his own house, are a good point to pin down the celebration of this thing that has shaped a seemingly infinite number of other things.
Enlarge (credit: Getty Images)
On Thursday, OpenAI announced updates to the AI models that power its ChatGPT assistant. Amid less noteworthy updates, OpenAI tucked in a mention of a potential fix to a widely reported "laziness" problem seen in GPT-4 Turbo since its release in November. The company also announced a new GPT-3.5 Turbo model (with lower pricing), a new embedding model, an updated moderation model, and a new way to manage API usage.
"Today, we are releasing an updated GPT-4 Turbo preview model, gpt-4-0125-preview. This model completes tasks like code generation more thoroughly than the previous preview model and is intended to reduce cases of 'laziness' where the model doesn’t complete a task," writes OpenAI in its blog post.
Since the launch of GPT-4 Turbo, a large number of ChatGPT users have reported that the ChatGPT-4 version of its AI assistant has been declining to do tasks (especially coding tasks) with the same exhaustive depth as it did in earlier versions of GPT-4. We've seen this behavior ourselves while experimenting with ChatGPT over time.
Enlarge / Who doesn't want less recoil? Unless, that is, you're someone competing against the person getting this benefit with a $100 "emulation" device. (credit: Cronus)
The Cronus Zen describes itself as a hardware tool for "universal controller compatibility," letting you plug in a third-party controller, an Xbox controller into a PlayStation, or even your keyboard and mouse into a console. But you can also use its scripting engine to "amplify your game" and set up "GamePacks" to do things like reduce recoil animations in games like Call of Duty. And that is where Cronus seems to have gotten into trouble.
As first noted by the Call of Duty news channel CharlieIntel, the latest update to the PlayStation 5's system (24.01-08.60.00) software blocks the Cronus from connecting. The update is "NOT mandatory," Cronus claims in a notice on its website, so Zen players can hold off and keep playing. Still, there is "currently no timetable on a fix … it could be 24 (hours), 24 days, 24 months, we won't know until we've dug into it." There is, for now, a "Remote Play Workaround" for those already too far updated.
Ars attempted to reach Cronus for comment and reached out to Sony as well and will update this post with any new information.
Enlarge (credit: Getty Images)
In 1921, Czech playwright Karel Čapek and his brother Josef invented the word "robot" in a sci-fi play called R.U.R. (short for Rossum's Universal Robots). As Even Ackerman in IEEE Spectrum points out, Čapek wasn't happy about how the term's meaning evolved to denote mechanical entities, straying from his original concept of artificial human-like beings based on chemistry.
In a newly translated column called "The Author of the Robots Defends Himself," published in Lidové Noviny on June 9, 1935, Čapek expresses his frustration about how his original vision for robots was being subverted. His arguments still apply to both modern robotics and AI. In this column, he referred to himself in the third-person:
For his robots were not mechanisms. They were not made of sheet metal and cogwheels. They were not a celebration of mechanical engineering. If the author was thinking of any of the marvels of the human spirit during their creation, it was not of technology, but of science. With outright horror, he refuses any responsibility for the thought that machines could take the place of people, or that anything like life, love, or rebellion could ever awaken in their cogwheels. He would regard this somber vision as an unforgivable overvaluation of mechanics or as a severe insult to life.
This recently resurfaced article comes courtesy of a new English translation of Čapek's play, called R.U.R. and the Vision of Artificial Life, accompanied by 20 essays on robotics, philosophy, politics, and AI. The editor, Jitka Čejková, a professor at the Chemical Robotics Laboratory in Prague, aligns her research with Čapek's original vision. She explores "chemical robots"—microparticles resembling living cells—which she calls "liquid robots."
Enlarge / A photo of David L. Mills taken by Raul654 on April 27, 2005. (credit: Raul654 / Benj Edwards / Getty Images)
On Thursday, Internet pioneer Vint Cerf announced that Dr. David L. Mills, the inventor of Network Time Protocol (NTP), died peacefully at age 85 on January 17, 2024. The announcement came in a post on the Internet Society mailing list after Cerf was informed of David's death by Mills' daughter, Leigh.
"He was such an iconic element of the early Internet," wrote Cerf.
Dr. Mills created the Network Time Protocol (NTP) in 1985 to address a crucial challenge in the online world: the synchronization of time across different computer systems and networks. In a digital environment where computers and servers are located all over the world, each with its own internal clock, there's a significant need for a standardized and accurate timekeeping system.
Enlarge / I normally prefer my performance cars on race circuits, but the Lightning Switchgear impressed this off-road racing novice. (credit: Jonathan Gitlin)
CONCORD, NC—The venerable pickup truck can play a multitude of roles these days. For some, it's nothing more than a work vehicle, something to carry around lumber or tools or tow a trailer full of equipment. For others, it's the new American family car. But some truck owners like to leave the tarmac behind to have a little fun in the wilderness. Mostly, that involves low-speed rock crawling, perhaps up the side of a steep mountain. But it doesn't have to be slow—vehicles like Ford's range of Raptors are designed to do highway speeds across expanses of desert wilderness, largely thanks to very clever dampers and plenty of suspension travel to munch up those bumps and bounces.
Ford is yet to make a Raptor version of its F-150 Lightning electric pickup, but we got an idea of what one could be capable of this week thanks to a ride in the Blue Oval's latest electric demonstrator, the F-150 Lightning Switchgear. It's the result of a collaboration between Ford Performance and RTR Vehicles, a tuning company founded and run by drifting champion and off-road racer Vaughn Gittin Jr., and while it's just a one-off for now, the Lightning Switchgear is a testbed for pushing the boundaries of what we can expect from electric trucks, Ford says. (You may remember RTR previously worked with Ford to create a 1,400-hp Mustang Mach-E in 2020.)
"This is going to focus on chassis and suspension. So to that end, just like you do with any good racing vehicle, you start with the tires," explained Sriram Pakkam, head of F1 and EV demonstrators at Ford.
Enlarge / A vial of the updated 2023-2024 formula of Pfizer's COVID-19 vaccine at a CVS Pharmacy in Eagle Rock, California, on September 14, 2023. (credit: Getty | Irfan Khan)
Staying up to date on COVID-19 vaccines can cut the risk of COVID-related strokes, blood clots, and heart attacks by around 50 percent in people ages 65 years or older and in those with a condition that makes them more vulnerable to those events, according to a new study from the Centers for Disease Control and Prevention.
The finding, published this week in the CDC's Morbidity and Mortality Weekly Report, should help ease concerns that the shots may conversely increase the risk of those events—collectively called thromboembolic events. In January 2023, the CDC and the Food and Drug Administration jointly reported a preliminary safety signal from their vaccine-monitoring systems that indicated mRNA COVID-19 vaccines may increase the risk of strokes in the 21 days after vaccination of people ages 65 and older. Since that initial report, that signal decreased, becoming statistically insignificant. Other vaccine monitoring systems, including international systems, have not picked up such a signal. Further studies (summarized here) have not produced clear or consistent data pointing to a link to strokes.
In May, the FDA concluded that the evidence does not support any safety concern and reported that "scientists believe factors other than vaccination might have contributed to the initial finding."
Enlarge / Jelly Bean is back! (credit: Andrew Cunningham)
It sure looks like Android 15 is going to have lock screen widgets. The Android 14 QPR2 beta landed the other day, and Mishaal Rahman over at Android Authority found a hidden unfinished feature that brings back lock screen widgets. We've expected this to happen since Apple's big lock screen widget release with iOS 16.
Rahman found a new "communal" space feature that resembles lock screen widgets. After enabling the feature and swiping in from the right of the lock screen, a pencil icon will pop up. Tapping the icon opens up a widget list, allowing you to move some widgets to the lock screen. Right now, in this unfinished state, the default lock screen clock and notification panel UI don't know how to get out of the way yet, so you get a pile of widgets with the usual lock screen UI on top. It's a mess.
Any time one smartphone operating system does something, the other tends to copy it, and iOS added lock screen widgets in 2022. Two years later is plenty of time for Google to adjust and copy the feature. The thing is, Android added lock screen widgets in 2012 with Android 4.2. Google removed the feature two years later in Android 5.0, so really, this is Android copying iOS copying Android. Some of this code is apparently making a comeback, as all the widgets available to the lock screen were ones that still had the 10-year-old "keyguard" flag set for Android 4.2.
Enlarge / Four new Ryzen 5000 CPUs, all riffs on existing Ryzen 5000 CPUs. (credit: AMD)
AMD announced the first Ryzen 8000 desktop processors today: a new lineup of socket AM5 CPUs that bring RDNA 3 integrated GPUs and an AI-accelerating NPU to its desktop platform for the first time. But the company also spent some time on new budget chips for its last-generation AM4 platform. The four new Ryzen 5000 processors cover everything from budget office desktops with integrated GPUs to cost-conscious gaming systems.
At the top of the range is the Ryzen 7 5700X3D, an 8-core CPU with an extra 64MB slab of L3 cache stacked on top of the main CPU die. At $249, it will be a little over $100 cheaper than the 5800X3D, but with the same core count, cache size, and a slightly lower maximum clock speed (4.1 GHz, down from 4.5 GHz). AMD compared it favorably to the Core i5-13600K in gaming workloads, a chip that currently retails for a bit over $280.
The Ryzen 7 5700 is a $175 8-core processor without 3D V-Cache that should still perform reasonably well in most workloads, though AMD's spec sheet says that it has less cache than the 5700X and only supports PCI Express 3.0 instead of PCIe 4.0. This indicates that the 5700 is actually a 5700G with the integrated graphics disabled; it will be a bit slower than the Ryzen 5700X, despite their similar names, core counts, and clock speeds. The Ryzen 5 5600GT and 5500GT are 6- and 4-core chips with Vega-based integrated graphics, both intended for lower-end systems. At $140 and $125, they essentially amount to minor clock speed bumps for the existing Ryzen 5 5600G and Ryzen 3 5300G.
Enlarge / The Raspberry Pi 5 inside its official case. (credit: Andrew Cunningham)
I bought an 8GB Raspberry Pi 5 as soon as they went up for preorder, just like I have bought every full-size Pi model since the Pi 3 Model B launched back in 2016, including the Pi 3B+, with its better Wi-Fi and more efficient chip, and the Pi 4, with its substantial performance and RAM boost.
The difference is that I didn't really have anything in mind for the Pi 5 when I bought it. But years of Pi shortages made me worried about its scarcity, and I figured I'd buy first and ask questions later rather than want it later and be totally unable to get one. In the end, it will probably knock each of my other Pis down a level in my tech setup: the Pi 5 becomes the retro emulation box, the Pi 4 becomes the multi-use always-on light-duty server (currently running a combo of HomeBridge, WireGuard, and a dynamic DNS IP address updater), the Pi 3B+ joins the Pi 3B as either "test hardware for small one-off projects" or "the retro emulation box I lent to a friend which may or may not have been ruined when their basement flooded."
Before I did that, though, I wanted to take another crack at trying to use a Pi as an everyday general-purpose desktop computer. The Raspberry Pi's operating system has always included many of the tools you'd need to take a crack at this, including a lightweight desktop environment and a couple of web browser options, and the Pi 4-based Pi 400 variant has always been pitched specifically as a general-purpose computer.
Enlarge (credit: Jonathan Gitlin)
I seem to be thinking a lot about Ford's electric pickup truck, the F-150 Lightning. Earlier this week, we got the news of price cuts and price increases. Before that, there was a pending cut to planned production output. Taken as it is, it's just the all-electric version of America's favorite pickup—and arguably the best version unless you need to pull something on the end of a trailer hitch.
But the Lightning doesn't exist in a vacuum. Depending on who you talk to, it's a clever attempt to get Americans to go electric, an utterly familiar wrapper on a slab of new technology that, yes, still requires the owner to adjust their mindset a bit from the gasoline-powered way of thinking. To others, it's a white elephant, one that costs too much and languishes on dealership forecourts, proof positive that electrification is a thing other countries might bother with, but forget that here at home, cowboy.
I've never found life to be quite that simple, and neither is the Lightning. Here in Washington, DC, the vehicle remains a rare sight—the only time I've seen one in the wild, it belonged to the DC government's fleet of vehicles (its job was inspecting abandoned vehicles). Out west, it's much more common to see electric F-150s on the road, and last year, Ford sold about 40,000 Lightnings, despite halting production for a fire and then again to retool part of the line.
Enlarge / The Ford F-150 Lightning Pro gets a plainer front treatment than the more expensive trims. (credit: Jonathan Gitlin)
Was Ars premature in calling the electric Ford F-150 Lightning pickup truck the most important electric vehicle of the decade? At launch, it seemed like a no-brainer—an all-electric version of the nation's most popular four-wheel vehicle that rode better than any other F-150 and a starting price of just $40,000 sure sounded compelling. In practice, things haven't worked out quite that well. Today, we learned that F-150 Lighting prices are increasing for some trims, and others are being retired altogether.
When Ford first announced Lightning pricing in 2021, the range started at $39,974 for a Lightning Pro—the trim aimed at commercial customers—with the standard range battery, or $49,974 for the Lightning Pro with a larger battery capable of 300 miles of range. The F-150 Lightning XLT was the entry-level model for private customers, which originally cost $52,974, with more expensive Lariat and Platinum trims that topped out at $90,874.
But Ford raised those prices before too long, then raised them again. Coupled with a post-pandemic trend of outrageous additional dealer markups, this spelled bad news for F-150 Lightning sales—the entry-level F-150 Lightning Pro cost $59,974 by the middle of 2023, $20,000 more than the 2021 pricing. In July 2023, the automaker cut F-150 Lightning prices heavily, but prices remained significantly higher than at launch.
Enlarge (credit: Aurich Lawson/Getty Images)
The mince pies have been eaten, the crackers have been cracked, and the days are starting to get longer. That means it's time to look back on the best vehicles we tested in 2023. It has been a good year for electric vehicles, which accounted for almost one in ten new vehicles sold in the US this year. We've also driven some rather good hybrids, as well as a pair of sports cars that reminded us that there's still room for enthusiast cars. Read on to find out which cars made the cut.
You'd be hard-pressed to spot the difference between the 2023 Polestar 2 and the 2024 Polestar 2, but the improvements are obvious when you drive one. (credit: Jonathan Gitlin)
In addition to claiming the top spot in 2023, Polestar might also win a prize for the most significant reengineering job for a midlife refresh. Normally, an automaker might restyle the bumpers or change the headlights and tweak the interior when it gives a model its spruce-up after a few years on sale. Not Polestar—it mostly left the cosmetics alone but moved the electric motor in the single-motor Polestar 2 from under the hood, where it drove the front wheels, to the rear, where it now drives the rear wheels.
Combined with a bit of a bump in power (ok, 29 percent more power and 48 percent more torque), the result is a real driver's car, with better steering and handling than the front-wheel drive Polestar 2 it replaces. There's more standard equipment than before, and it's more efficient, too. Only about 30 percent of US Polestar customers have picked the single-motor model in the past, but they're missing out. The twin-motor car might be faster, but it's less engaging to drive, has less range, and costs a whole bunch more.
The trailer for Resident Evil 4 on iOS
Apple's AAA gaming ambitions for the iPhone 15 Pro saw both a release and a delay this week.
When Apple unveiled the iPhone 15 Pro and touted its AAA gaming capabilities in September, the company named three upcoming games as showcases: the Resident Evil 4 remake, Death Stranding, and Assassin's Creed Mirage. All would arrive to iOS and all would require an iPhone 15 Pro to play.
Resident Evil 4 launched on iOS, iPadOS, and macOS today. And a few days ago, publisher 505 Games announced in a post to X that Death Stranding—which was expected to launch this month—has been delayed to "a new release date in early 2024" because it "needs a little more time."
Enlarge / A Long March 5 rocket, the largest launcher in China's inventory, deployed a classified Chinese military satellite into orbit Friday. (credit: CASC)
China's largest rocket apparently wasn't big enough to launch the country's newest spy satellite, so engineers gave the rocket an upgrade.
The Long March 5 launcher flew with a payload fairing some 20 feet (6.2 meters) taller than its usual nose cone when it took off on Friday with a Chinese military spy satellite. This made the Long March 5, with a height of some 200 feet, the tallest rocket China has ever flown.
Adding to the intrigue, the Chinese government claimed the spacecraft aboard the Long March 5 rocket, named Yaogan-41, is a high-altitude optical remote-sensing satellite. These types of surveillance satellites usually fly much closer to Earth to obtain the sharpest images possible of an adversary's military forces and strategically important sites.
Enlarge / The iPhone 15 lineup.
Over the past couple of years of reviewing the iPhone, we've often jokingly called them "smartcameras" rather than smartphones, as the camera features are really what sell people on upgrading to new models.
So, for our final Apple gift guide, we'll revisit some of what we explored in our iPhone 15 and iPhone 15 Pro review with a special focus on the cameras. If you're looking to grab a new iPhone for yourself or someone in your family, which camera is best?
The idea here is to provide a top-level, quick summary of the features of each iPhone camera as they pertain to specific uses to make for an easy buying guide for last-minute holiday shoppers who want a quick answer. We'll go over each phone and survey its features, detailing their relevant uses and noting some recommendations and considerations along the way.
Enlarge / An illustration of a robot holding a French flag, figuratively reflecting the rise of AI in France due to Mistral. It's hard to draw a picture of an LLM, so a robot will have to do. (credit: Getty Images)
On Monday, Mistral AI announced a new AI language model called Mixtral 8x7B, a "mixture of experts" (MoE) model with open weights that reportedly truly matches OpenAI's GPT-3.5 in performance—an achievement that has been claimed by others in the past but is being taken seriously by AI heavyweights such as OpenAI's Andrej Karpathy and Jim Fan. That means we're closer to having a ChatGPT-3.5-level AI assistant that can run freely and locally on our devices, given the right implementation.
Mistral, based in Paris and founded by Arthur Mensch, Guillaume Lample, and Timothée Lacroix, has seen a rapid rise in the AI space recently. It has been quickly raising venture capital to become a sort of French anti-OpenAI, championing smaller models with eye-catching performance. Most notably, Mistral's models run locally with open weights that can be downloaded and used with fewer restrictions than closed AI models from OpenAI, Anthropic, or Google. (In this context "weights" are the computer files that represent a trained neural network.)
Mixtral 8x7B can process a 32K token context window and works in French, German, Spanish, Italian, and English. It works much like ChatGPT in that it can assist with compositional tasks, analyze data, troubleshoot software, and write programs. Mistral claims that it outperforms Meta's much larger LLaMA 2 70B (70 billion parameter) large language model and that it matches or exceeds OpenAI's GPT-3.5 on certain benchmarks, as seen in the chart below.
Enlarge / Electric Ford F-150 Lightnings being built at the Rouge Electric Vehicle Center in Dearborn, Michigan. (credit: Ford)
On Monday, Automotive News reported that Ford's suppliers have been told by the automaker that from January it is halving the production rate of its F-150 Lightning from 3,200 trucks a week down to 1,600 trucks a week.
Ford debuted a fully electric version of its best-selling F-150 pickup truck in 2022. You'd be hard-pressed to tell the electric F-150 Lightning from a gas- or diesel-burning F-150—bar some aerodynamic detailing here and there, they all use the same body, and the EV hides its batteries neatly between the chassis rails.
That conservatism in design appeared to be a winning strategy with the pickup crowd. Ford's order books were flooded with over 200,000 reservations well before the truck hit the streets, spurring the automaker to announce last January that it would double its original production plan and aim for an annual production rate of 150,000 trucks a year.
Enlarge (credit: Getty Images | Benj Edwards)
In late November, some ChatGPT users began to notice that ChatGPT-4 was becoming more "lazy," reportedly refusing to do some tasks or returning simplified results. Since then, OpenAI has admitted that it's an issue, but the company isn't sure why. The answer may be what some are calling "winter break hypothesis." While unproven, the fact that AI researchers are taking it seriously shows how weird the world of AI language models has become.
"We've heard all your feedback about GPT4 getting lazier!" tweeted the official ChatGPT account on Thursday. "We haven't updated the model since Nov 11th, and this certainly isn't intentional. model behavior can be unpredictable, and we're looking into fixing it."
On Friday, an X account named Martian openly wondered if LLMs might simulate seasonal depression. Later, Mike Swoopskee tweeted, "What if it learned from its training data that people usually slow down in December and put bigger projects off until the new year, and that’s why it’s been more lazy lately?"
Enlarge (credit: Getty Images)
Grok, the AI language model created by Elon Musk's xAI, went into wide release last week, and people have begun spotting glitches. On Friday, security tester Jax Winterbourne tweeted a screenshot of Grok denying a query with the statement, "I'm afraid I cannot fulfill that request, as it goes against OpenAI's use case policy." That made ears perk up online since Grok isn't made by OpenAI—the company responsible for ChatGPT, which Grok is positioned to compete with.
Interestingly, xAI representatives did not deny that this behavior occurs with its AI model. In reply, xAI employee Igor Babuschkin wrote, "The issue here is that the web is full of ChatGPT outputs, so we accidentally picked up some of them when we trained Grok on a large amount of web data. This was a huge surprise to us when we first noticed it. For what it’s worth, the issue is very rare and now that we’re aware of it we’ll make sure that future versions of Grok don’t have this problem. Don’t worry, no OpenAI code was used to make Grok."
In reply to Babuschkin, Winterbourne wrote, "Thanks for the response. I will say it's not very rare, and occurs quite frequently when involving code creation. Nonetheless, I'll let people who specialize in LLM and AI weigh in on this further. I'm merely an observer."
Enlarge (credit: Aurich Lawson)
Back in April, we ran a series of useful and/or somewhat goofy prompts through Google's (then-new) PaLM-powered Bard chatbot and OpenAI's (slightly older) ChatGPT-4 to see which AI chatbot reigned supreme. At the time, we gave the edge to ChatGPT on five of seven trials, while noting that "it's still early days in the generative AI business."
Now, the AI days are a bit less “early," and this week's launch of a new version of Bard powered by Google's new Gemini language model seemed like a good excuse to revisit that chatbot battle with the same set of carefully designed prompts. That's especially true since Google's promotional materials emphasize that Gemini Ultra beats GPT-4 in "30 of the 32 widely used academic benchmarks" (though the more limited “Gemini Pro" currently powering Bard fares significantly worse in those not-completely-foolproof benchmark tests).
This time around, we decided to compare the new Gemini-powered Bard to both ChatGPT-3.5—for an apples-to-apples comparison of both companies’ current “free" AI assistant products—and ChatGPT-4 Turbo—for a look at OpenAI’s current “top of the line" waitlisted paid subscription product (Google’s top-level “Gemini Ultra" model won’t be publicly available until next year). We also looked at the April results generated by the pre-Gemini Bard model to gauge how much progress Google’s efforts have made in recent months.
Enlarge / It's a hefty plug, but it has to be so that an F-150 Lightning can send power back to the home through an 80-amp Ford Charge Station Pro. (credit: Ford)
Modern EVs have some pretty huge batteries, but like their gas-powered counterparts, the main thing they do is sit in one place, unused. The Ford F-150 Lightning was built with two-way power in mind, and soon it might have a use outside emergency scenarios.
Ford and Resideo, a Honeywell Home thermostat brand, recently announced the EV-Home Power Partnership. It's still in the testing phases, but it could help make EVs a more optimal purchase. Put simply, you could charge your EV when it's cheap, and when temperatures or demand make grid power time-of-use expensive (or pulled from less renewable sources), you could use your truck's battery to power the AC. That would also help with grid reliability, should enough people implement such a backup.
The F-150 Lightning already offers a whole-home backup power option, one that requires the professional installation of an 80-amp Ford Charge Station Pro and a home transfer switch to prevent problems when the grid switches back on. Having a smart thermostat allows for grid demand response, so the F-150 would be able to more actively use its vehicle-to-home (V2H) abilities.
Enlarge / This is what Ample's battery modules look like. (credit: Ample)
A small fleet of rideshare Fiat 500e electric vehicles will become testbeds for battery-swap technology in 2024. The experiment is being conducted by Ample, a startup working on battery swaps, and Stellantis, Fiat's parent company, the Verge reported today.
This isn't Ample's first test of its battery-swapping technology; in 2021 it started a small trial in the Bay Area to demo its modular battery, which replaces the existing traction battery in an EV and allows Ample's automated swap stations to switch out depleted packs for charged ones. But the fact that this deal was made with an OEM like Stellantis is still significant.
As we detailed last time we looked at Ample's technology, the EVs require some engineering work for this to all be possible. Ample has to design a structural frame to replace the existing battery pack that will instead contain the swappable modules, while still conforming to the engineering requirements of the original pack—down to the same fasteners, bolts, and connectors.
Lateo.net - Flux RSS en pagaille (pour en ajouter : @ moi) 
