News: Technology News

A song of hype and fire: The 10 biggest AI stories of 2023

Mon, 18 Dec 2023 16:44:14 +0000

When AI hype hit fever pitch—and a market leader nearly tore itself apart.

"Here, There, and Everywhere" isn't just a Beatles song. It's also a phrase that recalls the spread of generative AI into the tech industry during 2023. Whether you think AI is just a fad or the dawn of a new tech revolution, it's been impossible to deny that AI news has dominated the tech space for the past year.

We've seen a large cast of AI-related characters emerge that includes tech CEOs, machine learning researchers, and AI ethicists—as well as charlatans and doomsayers. From public feedback on the subject of AI, we've heard that it's been difficult for non-technical people to know who to believe, what AI products (if any) to use, and whether we should fear for our lives or our jobs.

Meanwhile, in keeping with a much-lamented trend of 2022, machine learning research has not slowed down over the past year. On X, former Biden administration tech advisor Suresh Venkatasubramanian wrote, "How do people manage to keep track of ML papers? This is not a request for support in my current state of bewilderment—I'm genuinely asking what strategies seem to work to read (or "read") what appear to be 100s of papers per day."

To wrap up the year with a tidy bow, here's a look back at the 10 biggest AI news stories of 2023. It was very hard to choose only 10 (in fact, we originally only intended to do seven), but since we're not ChatGPT generating reams of text without limit, we have to stop somewhere.

Bing Chat “loses its mind”

Aurich Lawson | Getty Images

In February, Microsoft unveiled Bing Chat, a chatbot built into its languishing Bing search engine website. Microsoft created the chatbot using a more raw form of OpenAI's GPT-4 language model but didn't tell everyone it was GPT-4 at first. Since Microsoft used a less conditioned version of GPT-4 than the one that would be released in March, the launch was rough. The chatbot assumed a temperamental personality that could easily turn on users and attack them, tell people it was in love with them, seemingly worry about its fate, and lose its cool when confronted with an article we wrote about revealing its system prompt.

Aside from the relatively raw nature of the AI model Microsoft was using, at fault was a system where very long conversations would push the conditioning system prompt outside of its context window (like a form of short-term memory), allowing all hell to break loose through jailbreaks that people documented on Reddit. At one point, Bing Chat called me "the culprit and the enemy" for revealing some of its weaknesses. Some people thought Bing Chat was sentient, despite AI experts' assurances to the contrary. It was a disaster in the press, but Microsoft didn't flinch, and it ultimately reigned in some of Bing Chat's wild proclivities and opened the bot widely to the public. Today, Bing Chat is now known as Microsoft Copilot, and it's baked into Windows.

US Copyright Office says no to AI copyright authors

An AI-generated image that won a prize at the colourado State Fair in 2022, later denied US

copyright registration.

Jason M. Allen

In February, the US Copyright Office issued a key ruling on AI-generated art, revoking the copyright previously granted to the AI-assisted comic book "Zarya of the Dawn" in September 2022. The decision, influenced by the revelation that the images were created using the AI-powered Midjourney image generator, stated that only the text and arrangement of images and text by Kashtanova were eligible for copyright protection. It was the first hint that AI-generated imagery without human-authored elements could not be copyrighted in the United States.

This stance was further cemented in August when a US federal judge ruled that art created solely by AI cannot be copyrighted. In September, the US Copyright Office rejected the registration for an AI-generated image that wona colourado State Fair art contest in 2022. As it stands now, it appears that purely AI-generated art (without substantial human authorship) is in the public domain in the United States. This stance could be further clarified or changed in the future by judicial rulings or legislation.

The rise of Meta’s LLaMA and its open weights direction

Getty Images | Benj Edwards

On February 24, Meta released LLaMA, a family of large language models available in different sizes (parameter counts) that kick-started an open-weights large language model (LLM) movement. People soon took things into their own hands when they leaked LLaMA's weights—crucial neutral network files that had previously only been provided to academics—onto BitTorrent. Soon, researchers began fine-tuning LLaMA and building off of it, competing over who could build the most capable model that could run locally on non-data-center computers. In tandem, Meta's Yann LeCun quickly became a vocal proponent of open AI models.

In July, Meta Launched Llama 2, an even more capable LLM, and this time, they let everyone have the weights. Code Llama followed in August, fine-tuned for coding tasks. But Meta wasn't alone in releasing "open" AI models: You've probably also heard of Dolly, Falcon 180B, Mistral 7B, and a few others, all of which continued the tradition of releasing weights so others could fine-tune them for performance improvements. And in early December, Mixtral 8x7B reportedly matched GPT-3.5 in capability, which was a landmark achievement for a relatively small and fast AI language model. Clearly, companies with closed approaches such as OpenAI (ironically), Google, and Anthropic are going to have a run for their money in the coming year.

GPT-4 launches and scares the world for months

A screenshot of GPT-4's introduction to ChatGPT Plus customers from March 14, 2023.

Benj Edwards / Ars Technica

On March 14, OpenAI released its GPT-4 large language model with claims that it "exhibit[ed] human-level performance on various professional and academic benchmarks" and a specification document (model card) that described attempts by researchers to get a raw version of GPT-4 to play out AI takeover scenarios. That set the doom ball rolling. On March 29, the Future of Life Institute published an open letter signed by Elon Musk calling for a six-month pause in the development of AI models more powerful than GPT-4. That same day, Time published an editorial by LessWrong founder Eliezer Yudkowsky advocating that countries should be willing to "destroy a rogue datacenter by airstrike" if they are seen building up a GPU cluster that could train a dangerous AI model because otherwise, "literally everyone on Earth will die" at the hands of a superhuman AI entity.

It was hype factor 11, and the doom kept rolling. In April, President Biden gave brief remarks about the risks of AI. Later that month, a trio of US congressmen announced legislation that proposed keeping AI from ever being able to launch nuclear weapons. In May, Geoffrey Hinton resigned from Google so he could "speak freely" about potential risks posed by AI. On May 4, Biden met with tech CEOs about AI at the White House. OpenAI CEO Sam Altman began a worldwide tour, including a stop at the US Senate, to warn about the dangers of AI and advocate for regulation. And to cap it all off, OpenAI executives signed a brief statement warning that AI could extinguish humanity. Eventually, the fear and hype began to settle down, but there's still a contingent of people (many linked to Effective Altruism) who are convinced that a theoretical superhuman AI is an existential threat to all of humanity, bringing a bubbling undercurrent of anxiety to every AI advancement.

AI art generators remain controversial but continue to grow in capability

An example of lighting and skin effects in the AI image-generator Midjourney v5.

Julie W. Design

2023 was a big year for leaps in capability from image synthesis models. In March, Midjourney achieved a notable jump in the photorealism of its AI-generated images with version 5 of its AI image synthesis model, rendering convincing people with five-fingered hands. Throughout the year, Midjourney reliably invoked disgust from critics of AI artwork, but it also inspired experimentation (and a bit of deception) from people who embraced the technology. And the pace of change didn't stop, with v5.1 coming in May and v5.2 launching in June, each adding new features and detail. Today, Midjourney is testing a standalone interface that does not require Discord to function, and Midjourney v6 is expected to launch sometime in late December.

Also in March, we saw the launch of Adobe Firefly, an AI image generator that Adobe says is trained solely on public domain works and images found in its Adobe Stock archive. And by late May, Adobe had integrated the technology into a beta version of its flagship Photoshop image editor with Generative Fill. And OpenAI's DALL-E 3 took prompt fidelity to a new level in September, raising interesting implications for artists in the near future.

AI deepfakes have a deeper impact

AI-generated photo faking Donald Trump's possible arrest, created by Eliot

Higgins using Midjourney v5.

@EliotHiggins on Twitter

Throughout 2023, the wider implications of image, audio, and video generators began to take hold. Several controversies emerged, including fairly convincing AI-generated images of Donald Trump getting arrested and the Pope in a puffy jacket in March (But Will Smith eating spaghetti fooled no one). Also that month, news broke about a scam where people were mimicking the voices of people's loved ones using AI and routing it through telephone calls to ask for money.

And despite our feature about people being able to use social media photos to create deepfakes in December 2022, AI image generation tech led to warnings from the FBI about "sextortion" scams using fake video to blackmail people in June. In September, nearly all of the Attorneys General in the United States sent a letter to Congress warning about the potential for AI-generated CSAM. And just about a year after our warning, teens in New Jersey reportedly created AI-generated nudes of classmates in November. Even so, we're only beginning to deal with the fallout of the rapidly advancing capability to replicate any form of recorded media almost effortlessly using AI.

AI writing detectors promise results but don’t work

A viral screenshot from April 2023 showing GPTZero saying, "Your text is likely to be written

entirely by AI" when fed part of the US Constitution.

Ars Technica

The emergence of ChatGPT led to an existential crisis for educators that rolled over into 2023, with teachers and professors worrying about synthetic text replacing human thought in class assignments. Companies quickly emerged to capitalize on these fears, promising tools that would be able to detect AI-written text. We soon began hearing stories of people being falsely accused of using ChatGPT to write their work when, in fact, everything had been human-written.

To date, no AI-writing detector is reliable enough to confirm or deny the existence of AI-generated text in a piece of writing. In July, we wrote a large feature explaining why this is the case, and not long after, OpenAI pulled its own AI-writing detector due to low rates of accuracy. By September, OpenAI stated that AI writing detectors don't work, writing in a FAQ, "While some (including OpenAI) have released tools that purport to detect AI-generated content, none have proven to reliably distinguish between AI-generated and human-generated content." Since then, the furor over AI detection has died down somewhat, but commercial tools claiming to detect AI-written work are still out there.

AI-generated “hallucinations” go mainstream

Aurich Lawson | Getty Images

In 2023, the concept of AI "hallucinations"—the propensity for some AI models to convincingly make stuff up—went mainstream thanks to large language models dominating the AI news this year. Hallucinations resulted in legal trouble: In April, Brian Hood sued OpenAI for defamation when ChatGPT falsely claimed that Hood had been convicted for a foreign bribery scandal (later settled). And in May, a lawyer who cited fake cases confabulated by ChatGPT got caught and later fined by a judge.

In April, we wrote a big feature about why this happens, but it didn't stop companies from releasing LLMs that confabulate anyway. In fact, Microsoft built one directly into Windows 11. By the end of the year, two dictionaries, Cambridge and Dictionary.com, named "hallucinate" their word of the year. We still prefer "confabulate," of course—and that AI-related definition wound up in the Cambridge Dictionary as well.

Google’s Bard “dances” to counter Microsoft and ChatGPT

Google

When ChatGPT launched in late November 2022, its immediate popularity caught everyone off guard, including OpenAI. As people began to murmur that ChatGPT could replace web searches, Google jumped into action in January 2023, hoping to counter this apparent threat to its search dominance. When Bing Chat launched in February, Microsoft CEO Satya Nadella said in an interview, "I want people to know that we made [Google] dance." It worked.

Google announced Bard in a botched demo in early February, then it launched Bard as a closed test in March, with a wide release in May. The company spent the rest of the year playing catch-up to OpenAI and Microsoft with revisions to Bard, the PaLM 2 language model in May, and Gemini in early December. The dance isn't over yet, but Microsoft definitely has Google's attention.

OpenAI fires Sam Altman (and he returns)

OpenAI CEO Sam Altman speaks during the OpenAI DevDay event on November 6, 2023, in San Francisco.

Getty Images | Ars Technica

On November 17, I was at a regular checkup talking to my doctor when my phone began repeatedly buzzing, almost nonstop. I apologized to the doctor and checked to see what was going on. OpenAI's nonprofit board of directors dropped a bombshell: It was firing its CEO Sam Altman, and everyone (my wife, mom, friends, and co-workers) was telling me about it. "It's a crazy world," the doctor said as I rushed home to write about the event, which Ars' Kyle Orland had already begun to cover in my stead. Confusing everyone, OpenAI's board did not disclose the exact reason for the surprise firing, only saying that Altman was "not consistently candid in his communications with the board."

Over that weekend, more details emerged, including the resignation of President Greg Brockman in solidarity and the role of OpenAI's Chief Scientist Ilya Sutskever in the firing process. Key investor Microsoft was furious, and Altman was soon in talks to return. He, along with 700-plus OpenAI employees, threatened to join Microsoft if the original team was not reinstated. It later emerged that Altman's handling of his attempted removal of board member Helen Toner largely precipitated the firing. Altman was officially back as CEO two weeks later, and the company claimed it was more united than ever. But the chaotic episode left lingering questions about the company's future and the safety of relying on a potentially unstable company (with an unusual nonprofit/for-profit structure) to responsibly develop what many assume will be world-changing technology.

The tech keeps rolling along

Even though we just covered ten major storylines related to AI from 2023, it feels like they barely scratch the surface of such a busy year. In June, I wrote about buying the last in-print encyclopedia, which gives me nostalgic comfort in the age of AI hallucination that we talked about above. And we've covered a bevy of interesting AI-generated visual stories, including AI-generated QR codes, geometric spirals, and mind-warping beer commercials.

All the while, perceived market leader OpenAI has never sat still technologically, releasing a ChatGPT app in May and introducing image recognition capability to ChatGPT Plus in September. GPT-4 Turbo and GPTs (custom roles for AI assistants) followed in November, and the year ended with GPT-5 development apparently underway. The story of Google's Gemini is still unfolding as well.

It's been a busy and impactful year, and we'd like to thank everyone for sticking with Ars Technica as we've wrestled with how to most effectively cover this rapidly evolving field. We appreciate your comments, support, and feedback. Our prediction for 2024? Buckle up.

Source

Microsoft Weekly: Copilot is the new Start button, Windows 11 23H2 vs 22H2, and Edge cringe

Sun, 17 Dec 2023 16:32:47 +0000

In this episode of Microsoft Weekly, we look at brazen Microsoft Edge behavior (again), final Windows 11 preview builds in 2023, performance comparisons between Windows 11 version 22H2 and 23H2, even more deprecated features in Windows, updates for Office apps, rumors about an ad-supported Xbox Cloud Gaming tier, app updates, and more.

Table of contents:

Windows 11 news
Updates are available
Gaming news
A blast from Microsoft's past
Random fact about Microsoft

Windows 11

Here we talk about everything happening around Microsoft's latest operating system in the Stable channel and preview builds: new features, removed features, controversies, bugs, interesting findings, and more. And of course, you may find a word or two about older but still supported versions.

The December 2023 Patch Tuesday updates are now available for download for those using supported Windows 10 (22H2) and Windows 11 (22H2 and 23H2) versions. Interestingly, the last Patch Tuesday update of the year contains nothing but security updates.

A little more than a month ago, Microsoft released Windows 11 version 23H2. As usual, some users are reluctant to upgrade, fearing potential performance downgrades or slowdowns. To refute those fears (or prove them), we published a detailed performance comparison between Windows 11 version 22H2 and version 23H2, running on the latest hardware from Intel and NVIDIA.

The list of deprecated features in Windows 10 and 11 received an addition: the legacy console mode is no longer in development. This utility for running outdated console apps that do not support the latest versions will soon be removed from future Windows updates. However, users will be able to restore it using features-on-demand.

During a talk at Snapdragon Summit 2023, Satya Nadella spoke about AI in Windows, claiming that Copilot is the new Start menu and the "orchestrator of all your app experiences." This sparked a heated discussion in the comment section, with users expressing their fear of Microsoft turning half-baked Copilot into the centerpiece of the next-generation Windows. With reports of Microsoft planning a big Windows client release in 2024 with a major focus on AI, those fears are not unfounded.

Besides lackluster experience on Windows, Copilot is struggling to provide accurate answers to election questions, as revealed by two nonprofit groups tracking how AI impacts people and society. It is a reminder that modern AI is still in its early infancy.

Windows Insider Program

Here are the new preview builds Microsoft released for testing this week:

Canary Channel	Dev Channel	Beta Channel	RP Channel
26016 with Windows Protected Print Mode, taskbar improvements, and bug fixes.	23606 with deprecated Windows Speech Recognition, File Explorer fixes, and more.	22635.2915 with more language support for Voice Access, improved screen casting, new features for Windows 365 and more.	Nothing new this week.

In addition to things mentioned in its changelog, build 26016 contains a hidden new section for the Settings app. It is called "AI Components," and it is supposed to let users view and uninstall additional AI-powered parts of Windows. The appearance of this section coincides with the recent report about Microsoft planning to release a big new Windows client with a major focus on AI somewhere in mid-2024.

Shortly after releasing builds 26016 and 23606, Microsoft confirmed that the "Phone Link" settings page was renamed to "Mobile Devices." The name is everything new, but Microsoft promises to add more changes in future updates.

Reminder: Builds above are the final preview updates in 2023. The next Windows 11 build will arrive once the holiday season is over. Come back soon!

Another reminder: You can always check out the latest Windows 10 and 11 preview and stable builds on Neowin's home page in the "Builds" widgets.

Updates are available

This section covers software, firmware, and other notable updates (released and coming soon) delivering new features, security fixes, improvements, patches, and more from Microsoft and third parties.

The noradsanta.org website received a big update with new languages and hosting in six Azure Regions worldwide. This is not the first time Microsoft has helped NORAD track Santa as he travels across the globe. This year's Azure updates will allow users to load the website faster, especially during peak loads.

However, the holiday season is still several days from now, so many people still have spreadsheets and presentations to do before they get their much-deserved rest. And if you happen to own one of Meta's VR headsets, you can now write Word documents, do spreadsheets, and create PowerPoint presentations without taking it off. Microsoft's signature Office apps are now available on Quest headsets.

Other updates for Office apps, some of which are only available to Microsoft 365 Insiders, include the following:

New features for the Shifts app in Microsoft Teams.
Improvements for the Catch Up feature in Word, Excel, and PowerPoint.
Video caption and subtitle support for PowerPoint on the web.
New file-sharing features are coming soon to the Microsoft 365 app on Windows and the web.
The School Connection feature is now available for Microsoft Teams on smartphones.

As a reminder, Microsoft will soon stop supporting its old messaging infrastructure, resulting in outdated Teams apps experiencing service disruptions, message delays, and more.

Microsoft might soon change what happens when you close the last tab in the Edge browser. The most recent preview builds contain a feature that "spawns" a new tab page when you close all tabs. You can check out this article to learn how to enable and test that change.

Microsoft revealed that starting with Edge 120, developers can utilize Copilot in DevTools to get help with warnings, errors, and other hard-to-understand messages. After enabling a flag, the browser will display a Copilot icon next to each warning, giving you a quick and easy way to troubleshoot a hiccup. Besides, you can highlight a code snippet and ask Copilot to explain it.

Moving from positive changes to somewhat head-scratching stories, here is Edge's official account posting cringe: in a now-deleted post on X, someone behind the official Microsoft Edge account audaciously replied to a user complaining about the missing "Never" button on a prompt trying to upsell Bing.

If you feel like such brazen conduct is too much and think it is time to ditch Microsoft Edge, check out Vivaldi and its latest update. Version 6.5 is now available for download with a new session panel, full history sync, workspace rules, and more.

If you use Avira to keep your system safe from all sorts of cyber nastiness, you might be exposed to a bug causing system freezes at launch. Multiple online reports emerged with complaints of 100% CPU and RAM usage on Windows 10 and 11 after installing the latest Avira update. Yikes!

Other notable updates released this week include the following:

Microsoft Edge 121 is now available in the Beta Channel.
Microsoft Dev Home Preview 0.8 is out with accessibility patches, UI tweaks, and bug fixes.
PowerToys 0.76.2 is out with fixes for WPF app crashes and the WebView2 handler.
Microsoft Store version 22311 is out with a single fix for missing items in the Library section.
Clipchamp will soon get an auto-compose feature and a new content library tab with access to free and premium content.
Arc is finally sending out invites for beta testing the browser on Windows.
Bing Image Creator (powered by DALLE-E) will soon be able to store generated pictures for longer—90 days instead of the current 50-day limit.

Here are the new drivers released this week:

NVIDIA Game-Ready 546.33 WHQL with The Finals and Fortnite Chapter 5 support.
Intel Arc 31.0.101.5081 and 5122 WHQL with support for the latest Core Ultra processors announced this week.

And here is our weekly Microsoft 365 Roadmap series, where we track the latest upcoming features for Microsoft's productivity apps, such as Copilot in the classic Outlook and channel improvements in Teams.

On the gaming side

Learn about upcoming game releases, Xbox rumors, new hardware, software updates, freebies, deals, discounts and more.

E3 (short for Electronic Entertainment Expo) is officially dead. ESA's official X account announced that the organization decided to pull the plug on the event after serving as a central showcase for the video game industry for more than two decades. Sadly, E3 did not survive the pandemic, and it lost its relevance in the post-pandemic world. Rest in peace!

Baldur's Gate 3, which is finally available on Xbox, has received its first patch with stability improvements. Note, however, that you must update your Xbox to resolve the potential problems with saves disappearing after a crash. Navigate to your Xbox settings and install available updates before jumping back into one of the most critically acclaimed games of 2023. Also, do not expect the game to hit the Xbox Game Pass any time soon: Larian confirmed it has no plans to add Baldur's Gate 3 to Microsoft's gaming service.

Cities Skylines II, another popular game released in 2023, received another big patch with two new maps, budget fixes, performance improvements, and more. This is the last update for the game this year, so expect the next one after the holiday season.

Fallout 4's next-generation update is also expected in 2024. Earlier this week, Bethesda said developers need more time to finish the work. Besides announcing the delay, the studio shared more information about Starfield updates scheduled for 2024. Expect the recently released game to receive city maps, new ways to travel, FSR 3 support, and many more.

Minecraft fans, especially those playing on modern Xbox consoles, will soon get access to a new mob and 4K support. The latest preview update introduces armadillos and high-resolution support on Xbox Series X|S.

As usual, each week, more games become available in NVIDIA's GeForce NOW streaming service. This week's additions include Forza Horizon 4 and Forza Horizon 5, Minecraft Dungeons, Stellaris Nexus, and more. If you do not have a compatible computer or console to enjoy those games, feel free to sign up for NVIDIA GeForce NOW. It's not free, of course, but you get the idea.

Xbox Cloud Gaming, another cloud streaming service, is now available on Meta's Quest 2, 3, and Pro VR headsets. You can tell Xbox finally has some sort of virtual reality support. Naturally, we are talking not about traditional VR experience but about playing games on a big screen inside a virtual room.

If you think that Xbox Cloud Gaming is too expensive, good news: Microsoft might introduce a free ad-supported tier, offering, for example, two hours of gameplay in exchange for watching a 30-second ad.

A blast from Microsoft's past

John Callaham's weekly "Look back" series provides throwbacks into the past, detailing the company's products, partnerships, mishaps, and successes from years ago.

This week's look-back is all about the failed gaming promises. Check out this article where we remember some of Microsoft's games announced at E3 that never made it to the public. Those include Matter, Scalebound, and others.

Random fact about Microsoft

And here is a randomly selected piece of trivia about the company, Windows, and other Microsoft-made things.

Did you know that Microsoft once warned users about Windows 98 and 95 hanging after 49.7 days of continuous use without restarts due to their timing algorithms? However, you would be very lucky to get your Windows 95 machine run that long without a single restart or crash, so not that many users experienced the odd time-related bug.

Source

Homeworld 3 is getting a skirmish mode for its release in March 2024

Sat, 16 Dec 2023 18:01:55 +0000

A few weeks ago, developer Blackbird Interactive announced that the long-in-development sci-fi RTS sequel Homeworld 3 has a definite release date of March 8, 2024. This week, Blackbird posted up a new development update on the game, which confirmed it would include an online multiplayer Skirmish gameplay mode.

The dev update offers the first info on how Skirmish mode will work in Homeworld 3:

For the uninitiated, Skirmish Mode is where you can go head to head against friends, strangers, and the computer in classic deathmatch-style brawls. In Homeworld 3, Skirmish Mode supports up to a max of six players and you can set teams freely. Engage in a chaotic free-for-all, square off in 2v2v2 or 3v3 — or go with some other permutation like 4v2 or 5v1. You can also substitute in computer players with differing difficulty levels or adjust settings like how many resources players start with or whether the game ends when your mothership is destroyed.

The game will be released with eight dedicated Skirmish mode maps, but since Homeworld 3 will also be released with mod support, you can expect to see a lot more Skirmish maps to become available via the in-game mod browser shortly after launch.

Blackbird also says that Skirmish mode will include a way for players to access and use the ships in Homeworld 3's new faction, The Incarnate. The dev update states:

We’ve been tight-lipped about this enigmatic armada because of their role in Homeworld 3’s story, but what I will say is that Incarnate fleets present a unique departure from the strategies and strengths of the Hiigaran navy

The dev update also offered up more info on the game's new co-op War Games mode. Blackbird has already revealed Homeworld 3's PC hardware requirements with a variety of settings.

Homeworld 3 will be released on Steam and the Epic Games Store You can also preorder the game's Collector's Edition on Amazon. Priced at $174.99, it includes, among other things, a 14-inch model of the game's main Mothership, along with smaller models of the game's Hiigaran Destroyer, Torpedo Frigate, and three Recon ships in formation.

As an Amazon Associate I earn from qualifying purchases.

Source

A quick look back at some Microsoft games announced at E3 that were never released

Sat, 16 Dec 2023 17:58:41 +0000

A few days ago, The Entertainment Software Association made the sad but also expected announcement that there would never be another Electronic Entertainment Expo trade show. It's a quiet end to the trade show that, for a long time, was the single biggest annual event in the video and PC game industry.

Earlier this year, I wrote about my personal memories of attending E3 shows in Los Angeles (and, briefly, in Atlanta). It was great fun to walk on the exhibit floor and see all the huge displays that publishers and developers created to help promote their games.

However, an even bigger highlight for me was attending the E3 press conferences that were usually held before the show itself started. Microsoft had its share of E3 press conferences, and they frequently included surprise celebrity guests like Steven Spielberg, Sir Paul McCartney, Ringo Starr, and, at the last Microsoft E3 press conference in 2019, Keanu Reeves promoting his appearance in Cyberpunk 2077.

Of course, Microsoft also used its E3 press conferences to show off some upcoming games. However, over the years, games that were first revealed at those events ended up not being released. Here's a quick look back at some of those games that we never got to play.

Untitled Black Tusk game - E3 2013

In November 2012, Microsoft announced it had formed a new game developer in Vancouver called Black Tusk Studios. At the time, Microsoft said the team was working on "Microsoft's next big entertainment franchise", an all-new original game that it hoped would be as successful as its Halo series.

During Microsoft's E3 2013 press event, the company showed off a teaser trailer for Black Tusk's game, which did not have a title. The trailer showed what looked like a New Year's Eve celebration in Singapore. The trailer cut to a guy repelling from a tall building and then taking out an unfortunate guard from behind. He then brought out what looked like a high-tech pistol.

The trailer was brief, but it did look like it had some potential. It definitely had some similarities to Ubisoft's Splinter Cell franchise.

However, that trailer was all we would ever see of this game. In 2014, Microsoft announced it had acquired the Gears of War franchise from Epic Games. In 2015, it revealed that Black Tusk Games would get a name change to The Coalition and that it would be in charge of making new Gears games for Microsoft, dashing the hopes of an all-new original game franchise from the studio.

Matter - E3 2012

Microsoft's E3 2012 press event showed off a teaser trailer for a new game from developer Blind Wink that was planned to be published by the company for the Xbox 360 console and its Kinect motion sensor accessory. It was called Matter. The trailer showed what looked like a Portal-like sci-fi puzzle experience, with a silver ball moving through ever-changing levels and being attacked by square-shaped enemies.

The other thing that got some people's attention was that Matter had some kind of involvement from film director Gore Verbinski, best known as the director of The Ring and the first three films in the Pirates of the Caribbean series.

However, that brief trailer reveal of Matter was all we ever saw of the game. In March 2013, less than a year after it was revealed, Microsoft confirmed that Matter was canceled. There was no official explanation of what happened with the game's development.

Scalebound - E3 2014

Microsoft's E3 2014 press event included the debut trailer for Scalebound, a highly ambitious fantasy action RPG from Japan-based developer PlatinumGames. Unlike the other two games on this list, Microsoft showed off playable versions of Scalebound at other game events after it debuted at E3.

This was one of Microsoft's most anticipated games for a while, especially since it promised gamers that you could control dragons. However, delays in the game's development started creeping up, and in January 2017, Microsoft announced it had decided to cancel Scalebound.

There have been rumors of some bad blood between Microsoft and PlatinumGames over the development of Scalebound that eventually led to its cancelation. In 2022, the studio's president, Atsushi Inaba, said he would like to revive the game development with Microsoft. However, that seems to be an unlikely prospect.

Once again, we said farewell to E3. It brought lots of joy to gamers and also showed off a few games that never made it to our consoles and PCs. It's truly the end of an era in gaming.

Source

Nvidia To Stop Producing GTX 1600 Graphics Cards In 24Q1

Sat, 16 Dec 2023 07:58:38 +0000

The production stop means that the GTX 1660, GTX 1650 and GTX 1630 series stop becoming available. The GTX 1600 series is cheaper, hence more affordable.

More than 4 years ago, Nvidia released the GTX 1600 graphics cards. Officially known as GeForce 16 series, the architecture used inside the GPU isn’t much different from the RTX 2000.

However, the GTX 1600 graphics cards lack something which is more in the news these days. It lacked hardware support for ray tracing and DLSS upscaling.

Due to this, the GTX 1600 series graphics were cheaper and became more affordable for budget gamers. In fact, so famous is the GTX 1600 that as per Steam Hardware Survey, the GTX 1650 remains the second most used graphics card out there even now.

The whole GTX series line-up includes:

GeForce GTX 1630
GeForce GTX 1650
GeForce GTX 1650 Super
GeForce GTX 1660
GeForce GTX 1660 Super
GeForce GTX 1660 Ti

Nvidia has discontinued the GTX 1660 series long ago. Now it seems that now Nvidia wants to stop the production of the GTX 1650 and GTX 1630 series too.

GTX 1600 To Be Discontinued

ITHome (via @harukaze5719) and VideoCardz reports that starting the first quarter of next year, that is, Q1 2024, Nvidia will stop the production of the GeForce GTX 1600 series.

This means that Nvidia will stop supplying the GTX 1600 series GPUs to the graphics cards brands and makers.

If that is not enough, the rumor states that there will be no cheap graphics card available to fill this price gap.

For the record, the cheapest graphics card available in the RTX 3000 series is the RTX 3050 and Nvidia wants to stop producing the 8GB version of it and replace it with a 6GB version instead.

Just a quick search reveals that while the GTX 1650 sells for around $170-$180, the RTX 3050 is available for $220-$260. This means there’s a lot of room in the sub-$200 graphics cards, which Nvidia seems to be discontinuing.

Yes, likes of AMD is going to release Ryzen 8000 series with a powerful built-in GPU, especially in the G version of processors. However, all this means cheap dedicated graphics cards will not be as commonly available as they have been before. Especially for the budget users.

Source

RTX 4080, RTX 4070 Ti, RTX 4070 Super Release Dates Leak

Sat, 16 Dec 2023 07:57:14 +0000

Nvidia GeForce RTX 4000 Super graphics cards are expected to start selling from January 2024. It includes RTX 4080 Super, RTX 4070 Ti Super & RTX 4070 Super.

About a year ago, Nvidia released the RTX 40 series of graphics cards. While the RTX 4090 was praised for its unmatchable performance, all the other graphics cards in the series received mixed reviews.

The reasons are simple, when compared to the RTX 3000 graphics cards, the RTX 4000 graphics cards offer unexciting uplift in performance, while being expensive at the same time.

With Nvidia unlikely to release the RTX 50 series anytime soon, Nvidia wants to do something to increase the sales of the RTX 40. This means making Super versions of the RTX 40 series graphics cards.

RTX 40 Super Release Dates Leak

ITHome reports (via VideoCardz) that Nvidia is going to announce the RTX 40 Super series of graphics cards on 8 January 2024. Interestingly, it’s the same day that Intel is going to release the non-K version of Raptor Lake Refresh CPUs. It should be called a day of refreshes.

Anyway, the RTX 40 Super series is going to have three new graphics cards at launch. The RTX 4080 Super, RTX 4070 Ti Super & RTX 4070 Super. While the RTX 4070 Super will release on 17th January, the RTX 4070 Ti Super will come on January 24 and RTX 4080 Super will be available on January 31.

Another interesting thing is what they are replacing as per the reports. While RTX 4080 Super and RTX 4070 Ti Super will replace their non-Super versions, the RTX 4070 Super and RTX 4070 will coexist.

RTX 40 Super Specs

The RTX 4080 Super is likely to be priced at $999. Which is $200 cheaper than what RTX 4080 was at launch. It’s likely going to use the AD103-400 GPU. The RTX 4080 Super will come with 10240 CUDA cores. It will come with 256-bit memory bus width with 16GB VRAM. The VRAM itself will be based on GDDR6X running at 24Gbps. The TDP will be 320W. It’s likely going to be 6%-9% faster than RTX 4080.

The RTX 4070 Ti Super is likely to be priced around $799-849, which makes it the same or more expensive than the RTX 4070 Ti at launch. It’s going to use either AD103-275 or AD102-175 GPU. The RTX 4070 Ti Super will come with 8448 CUDA cores. It will come with 256-bit memory bus. It will come with 16GB VRAM, which is 4GB more than RTX 4070 Ti. The VRAM will be running at 21-22.4Gbps. The RTX 4070 Ti Super is said to be 14%-22% faster than the RTX 4070 Ti. The performance is said to be similar to the RTX 4080.

The RTX 4070 Super is likely to be priced around $599-$649, which makes it the same or more expensive than the RTX 4070 at launch. It’s going to use either AD104-350 or AD103-175 GPU. The RTX 4070 Super will come with 7168 CUDA cores. It will come with 192-bit memory bus. In terms of VRAM, it comes with 12GB GDDR6X with a speed of 21Gbps. The performance is said to be similar to the RTX 4070 Ti.

Conclusion

The RTX 4000 graphics cards have been unexciting. So the Super refresh graphics cards are much welcomed. Especially if they can provide better performance at the same price. What’s also good is that it’s going to release sooner, rather than later.

What we wish to see, however, is Nvidia also releasing a Super version of RTX 4060 too. Not forget, they should release an affordable RTX 4050 so that budget gamers can buy the RTX 40 series too.

Source

Star Citizen Alpha 3.22 update brings 15 new Derelict Settlements, an alien ship, and more

Fri, 15 Dec 2023 18:10:20 +0000

Cloud Imperium Games is finishing off 2023 with a bang. The next major Alpha update for the crowd-funded sci-fi RPG Star Citizen is now available. Dubbed Wrecks to Richest, Alpha 3.22 update brings a large selection of new locations to visit, upgraded ship salvaging gameplay, brand-new hairstyles for players, an alien ship modified for humans, and more.

The amount of Derelict Settlements players can run into, while exploring or as part of quests, have been increased by 15. Spread across the Hurston and microTech planets, these settlements can house neutral NPCs and shops, while others may be abandoned or packed with hostile gangs.

Salvaging aspects of Star Citizen have been getting continuous features in recent patches, and the latest update is no different. Drake Vulture and Aegis Reclaimer salvage ship owners can now break apart and disintegrate abandoned ships to sell off the resulting raw materials.

Looking at vehicles, the new Aopoa San'tok.yāi is a spaceship made by the Star Citizen universe's Xi’an alien race. The acrobatic ship has been modified for human use. At the same time, The X1 drops in as an open-canopy ground vehicle, which is essentially a hoverbike capable of high speeds.

Cloud Imperium Games recently gave a deep look at the state of Squadron 42, the upcoming single-player campaign of Star Citizen. While a release date is not yet attached to the now "feature complete" project, hairstyles and related tech from the campaign are being ported over to the multiplayer portion. 20 hairstyles from Squadron 42 can now be used by players and NPCs, with more coming soon.

Other changes of this update include Arena Commander content for FPS and flight modes, personal cargo containers for players to store items, and holiday promotions. The complete patch notes for Star Citizen Alpha 3.22 can be read here.

Source

Microsoft is exploring specialized liquids for cooling its AI chips, amid local water supply concerns

Fri, 15 Dec 2023 18:06:44 +0000

Microsoft is looking into a new technique that will allow it to use specialized liquids to cool its AI chips, rather than fans and excessive amounts of water.

What you need to know

Microsoft is potentially looking into a new technique that will allow it to use a "specialized liquid" to cool its AI chips.
The company unveiled an advanced AI chip in November dubbed Maia 100 chip, which ships with a cold plate with cool water tucked beneath it to prevent it from frying.
This is amid local water supply concerns, as the technology consumes up to one water bottle for cooling per query.

Generative AI is probably one of the most discussed topics this year across the board in the tech world. Many organizations have shifted focus to the tech in the past few months, with Microsoft at the forefront after its multi-billion dollar investment, which extended its partnership with OpenAI.

Admittedly, Microsoft has achieved incredible milestones by integrating AI across most of its products and services. Barely a month after the company debuted Copilot (formerly Bing Chat) Bing crossed 100 million daily active users.

While all this is impressive, running an AI-powered chatbot is no easy feat. We already know OpenAI spends up to $700,000 on a daily basis to keep ChatGPT running. Not forgetting the local water supply concerns, as Copilot and ChatGPT consume an entire water bottle for cooling per query.

Microsoft addresses local water supply concerns with specialized liquids for cooling its AI chips

(Image credit: Windows Central)

With this in mind, both OpenAI and Microsoft have heavily invested in fans as well as a steady water supply to ensure that operations run seamlessly. Without an adequate supply of water and fans, the AI chips will simply fry due to the enormous amount of heat produced throughout the entire process.

This is both a daunting and expensive venture, which prevents companies like Microsoft from realizing the full potential of AI. According to a spot by Bloomberg, Microsoft is well on its way to pulling itself out from under this hazardous rug. While making its debut in advanced chip making for AI last month, the company unveiled its brand new Maia 100 chip.

The Maia 100 chip is designed to take on NVIDIA's top-of-the-line products. And unlike previous chips used in AI ventures, Microsoft's flagship AI chip ships with a cold plate. As the name suggests, it's designed to ensure that the semiconductor remains cool at all times. The cold plate features a cold fluid tucked at the bottom of the chip to prevent it from overheating.

Per Bloomberg's report, this could just be the starting point, as Microsoft could potentially explore full immersion cooling if everything goes according to plan. The servers will essentially be immersed in a "specialized liquid" to ensure that operations run smoothly without any fear that the chips might overheat.

Do you think the specialized liquid approach for cooling AI chips is a permanent solution for Microsoft? Share your thoughts in the comments.

Source

Jeff Bezos plays down AI dangers and says one trillion humans could live in huge cylindrical space stations

Fri, 15 Dec 2023 17:07:55 +0000

People living in space could visit Earth on vacation, billionaire says

Artificial intelligence is more likely to save humanity than to destroy it, Jeff Bezos said recently. The billionaire also said he would like to see the human population grow to one trillion, with most people living in huge cylindrical space stations.

In an interview with podcaster Lex Fridman, the Amazon AMZN, +1.01% founder and former CEO rejected the idea that humans should colonize other planets, saying he believes building space colonies is the only way to achieve such population growth.

“I would love to see a trillion humans living in the solar system. If we had a trillion humans, we would have, at any given time, 1,000 Mozarts and 1,000 Einsteins,” he said. “The only way to get to that vision is with giant space stations. The planetary surfaces are just way too small.”

Bezos, who has a net worth of $172 billion, said that if people lived on O’Neill space colonies near Earth, built using raw materials from the moon and objects in the asteroid belt, they could visit our current planet on vacation.

The concept of O’Neill colonies was developed by science-fiction writer Gerard K. O’Neill as a solution to the problem of livable environments in space. The space stations, designed as two cylinders that rotate around an axis, would offer an artificial Earth-like environment and use rotation to simulate gravity.

Bezos’s vision is in opposition to that put forward by Tesla TSLA, +0.53% CEO Elon Musk, who is currently the richest person in the world. Musk has said that he hopes that humans will become a “multiplanetary species” and that he aims to colonize Mars via his company SpaceX.

In Bezos’s vision, space colonies would help support a population that is 125 times the size of the Earth’s current population.

He said people would be free to choose whether or not to live in space, but those who opted for the O’Neill colonies would “’be able to use much more energy and much more material resources in space than they would be able to use on Earth.”

Bezos said people living in space would still have the opportunity to travel to Earth on vacation, in the “same way that you might go to Yellowstone National Park.”

He acknowledged that he “won’t live long enough to see the fruits” of his efforts to colonize space, saying that the personal rewards of his work with space company Blue Origin “come from building a road to space.”

In the interview, Bezos also put forward an optimistic vision of the future of artificial intelligence, despite warning it has the potential to be “incredibly destructive.”

AI has the potential to save humans from going extinct, and people who are “overly concerned” about the dangers of the technology “may be missing part of the equation,” he argued.

“Even in the face of all this uncertainty, my own view is that these powerful tools are much more likely to help us and save us, even, than they are to, on balance, hurt us and destroy us,” Bezos said.

AI has the potential to help humanity develop “better medicines and better tools to develop more technologies,” which could ensure its long-term survival, he said.

Musk, in contrast, has repeatedly raised concerns about the dangers of AI and has said it poses a risk to humanity.

Despite their differences, Bezos said he thinks Musk “must be a very capable leader” in view of his successes with SpaceX and Tesla.

“I don’t really know Elon very well. I know his public persona, but I also know you can’t know anyone by their public persona. It’s impossible. You may think you do, but I guarantee you don’t,” Bezos said.

Bezos also warned about the dangers of nuclear weapons and climate change. “We need to start training ourselves to think longer term,” he said.

He talked about his childhood, saying that working on his grandfather’s ranch in Texas helped him develop a “problem-solving mentality.”

Between the ages of 4 and 16, Bezos spent summers on the ranch in order, he said, to give his mother — who was 17 years old when Bezos was born — a break. He did a variety of chores while taking daily breaks with his grandfather to watch the soap opera “Days of Our Lives.”

Bezos said the resourcefulness he developed on the ranch helped him in his path to becoming an inventor, adding that he hopes what he creates will be taken for granted in the future.

“That’s an inventor’s greatest dream, is that their inventions are so successful that they are one day taken for granted. Nobody thinks of Amazon as an invention anymore,” Bezos said.

“Nobody thinks of customer reviews as an invention. We pioneered customer reviews, but now they’re so commonplace. Same thing with one-click shopping and so on. But that’s a compliment,” he said. “You invent something that’s so used, so beneficially used by so many people, that they take it for granted.”

The entrepreneur also said that while he was at Princeton University, an encounter with a fellow student from Sri Lanka convinced him not to pursue a career as a theoretical physicist. Bezos realized, he said, that “your brain has to be wired in a certain way.”

He recalled how the student was able to solve a “difficult partial differential equations problem” — one that Bezos and a fellow student had been working on for three hours without making any headway whatsoever — in seconds.

Source

DeepMind’s AI finds new solution to decades-old math puzzle — outsmarting humans

Fri, 15 Dec 2023 16:26:01 +0000

Researchers claim it is the first time an LLM has made a novel scientific discovery

DeepMind has used a large language model (LLM) to generate a novel solution to one of humanity’s toughest math problems — in a breakthrough that could herald a new era in AI development.

The model, known as FunSearch, discovered a solution to the so-called “cap set puzzle.” The decades-old math conundrum essentially comes down to how many dots you can joint down on a page while drawing lines between them, without three of them ever forming a straight line.

If that gave you a migraine, don’t worry. What’s important to note is that the problem has never been solved, and researchers have only ever found solutions for small dimensions. Until now.

FunSearch successfully discovered new constructions for large cap sets that far exceeded the best-known ones. While the LLM didn’t solve the cap set problem once and for all (contrary to some of the news headlines swirling around), it did find facts new to science.

“To the best of our knowledge, this shows the first scientific discovery – a new piece of verifiable knowledge about a notorious scientific problem — using an LLM,” wrote the researchers in a paper published in Nature this week.

In previous experiments, researchers have used large language models to solve maths problems with known solutions.

FunSearch works by combining a pre-trained LLM, in this case a version of Google’s PaLM 2, with an automated “evaluator.” This fact-checker guards against the production of false information.

LLMs have been shown to regularly produce so-called “hallucinations” — basically when they just make shit up and present it as fact. This has, naturally, limited their usefulness in making verifiable scientific discoveries. However, researchers at the London-based lab claim that the use of an in-built fact-checker makes FunSearch different.

FunSearch engages in a continuous back-and-forth dance between the LLM and the evaluator. This process transforms initial solutions into new knowledge.

What also makes the tool quite promising for scientists is that it outputs programs that reveal how its solutions are constructed, rather than just what the solutions are.

“We hope this can inspire further insights in the scientists who use FunSearch, driving a virtuous cycle of improvement and discovery,” said the researchers.

Source

Intel's CEO swipes at NVIDIA's AI strategy, calling it "shallow and small"

Fri, 15 Dec 2023 16:23:22 +0000

Intel believes that inference, not learning, is the key to winning the AI war.

Intel CEO Pat Gelsinger criticizes NVIDIA's AI technology, calling it "shallow and small" compared to Intel's own.

NVIDIA focuses on training AI models from scratch, while Intel focuses on adapting existing models to new situations.

Gelsinger believes that the inference market is where the game will be at, and that big players like Google and OpenAI are moving in that direction.

We're still recovering from all the cool Meteor Lake laptop announcements made at Intel's "AI Everywhere" event, but Intel's CEO is still fired up. In a recent statement, Intel CEO Pat Gelsinger launched an attack on NVIDIA's AI technology, calling its field of expertise "shallow and small" compared to his own.

Intel's big swing at NVIDIA's AI technology

As reported by Tom's Hardware, Pat Gelsinger made a statement at NASDAQ criticizing his technological rival's methodology in the AI market. Intel and NVIDIA have been locked in an AI arms war for a while now, as both try to produce the hardware needed to power a new tech scene driven by artificial intelligence.

Right now, both sides are taking different approaches to AI processing. NVIDIA's CUDA focuses more on training artificial intelligence from the ground up, giving models data and letting it learn from what it's given. It's the AI equivalent of someone going through education to get the degrees they need to do a job right.

Intel, however, focuses more on "inference." This is when an existing AI model adapts and learns from a situation it has never seen before. This is the AI equivalent of taking someone who has the degrees needed to do a job, giving them a relevant task where they have zero experience, and monitoring how they apply their knowledge to the situation.

As both sides race to become the leader in AI processing, Pat Gelsinger believes that NVIDIA's approach is short-lived:

"We think of the CUDA moat as shallow and small. Because the industry is motivated to bring a broader set of technologies for broad training, innovation, data science, et cetera."[...]"As inferencing occurs, hey, once you've trained the model… There is no CUDA dependency, it's all about, can you run that model well?"

Pat Gelsinger continued, stating that that Intel will still compete on the training level, but believes that "fundamentally, the inference market is where the game will be at." He also claims that "the entire industry is motivated to eliminate the CUDA market," stating that big players in the AI market like Google and OpenAI are moving toward inference.

Regardless of which side will win the AI war, it's going to be a wild few months as these two tech giants clash. We'll have to wait and see which strategy will outlive the other.

Source

Microsoft unveils Phi-2, a small language model that packs power

Fri, 15 Dec 2023 16:13:38 +0000

Even though Google's Gemini Nano has only been out for a week, it already has some competition.

When you think of language models in relation to generative artificial intelligence (AI), the first term that probably comes to mind is large language model (LLM). These LLMs power most popular chatbots, such as ChatGPT, Bard, and Copilot. However, Microsoft's new language model is here to show that small language models (SLMs) have great promise in the generative AI space, too.

On Wednesday, Microsoft released Phi-2, a small language model capable of common-sense reasoning and language understanding, and it's now available in the Azure AI Studio model catalog.

Don't let the word small fool you, though. Phi-2 packs 2.7 billion parameters in its model, which is a big jump from Phi-1.5, which had 1.3 billion parameters.

Despite its compactness, Phi-2 showcased "state-of-the-art performance" among language models with less than 13 billion parameters, and it even outperformed models up to 25 times larger on complex benchmarks, according to Microsoft.

Phi-2 outperformed models -- including Meta's Llama-2, Mistral, and even Google's Gemini Nano 2, which is the smallest version of Google's most capable LLM, Gemini -- on several different benchmarks, as seen below.

Microsoft

Phi-2's performance results are congruent with Microsoft's goal with Phi of developing an SLM with emergent capabilities and performance comparable to models on a much larger scale.

"A question remains whether such emergent abilities can be achieved at a smaller scale using strategic choices for training, e.g., data selection," said Microsoft.

"Our line of work with the Phi models aims to answer this question by training SLMs that achieve performance on par with models of much larger scale (yet still far from the frontier models)."

When training Phi-2, Microsoft was very selective about the data used. The company first used what it calls "text-book quality" data. Microsoft then augmented the language model database by adding carefully selected web data, which was filtered on educational value and content quality.

So, why is Microsoft focused on SLMs?

SLMs are a cost-effective alternative to LLMs. Smaller models are also useful when they are being used for for a task that isn't demanding enough to require the power of an LLM.

Furthermore, the computational power required to run SLMs is much less than LLMs. This reduced requirement means users don't necessarily have to invest in expensive GPUs to power their data-processing requirements.

Source

The world's first human brain-scale supercomputer will go live next year

Fri, 15 Dec 2023 15:41:16 +0000

The supercomputer, named DeepSouth, is being developed by Western Sydney University in Australia.

When it goes online next year, it will be capable of 228 trillion synaptic operations per second.

It could one day help create a cyborg brain vastly more powerful than our own.

Our brains are remarkably energy efficient.

Using just 20 watts of power, the human brain is capable of processing the equivalent of an exaflop — or a billion-billion mathematical operations per second.

Now, researchers in Australia are building what will be the world's first supercomputer that can simulate networks at this scale.

The supercomputer, known as DeepSouth, is being developed by Western Sydney University.

When it goes online next year, it will be capable of 228 trillion synaptic operations per second, which rivals the estimated rate of operations in the human brain.

The hope is to better understand how brains can use such little power to process huge amounts of information.

If researchers can work this out, they could someday create a cyborg brain vastly more powerful than our own. The work could also revolutionize our understanding of how our brains work.

"Progress in our understanding of how brains compute using neurons is hampered by our inability to simulate brain-like networks at scale," said André van Schaik, a director at Western Sydney University's International Centre for Neuromorphic Systems.

"Simulating spiking neural networks on standard computers using Graphics Processing Units and multicore Central Processing Units is just too slow and power intensive," he added. "Our system will change that."

Ralph Etienne-Cummings at Johns Hopkins University, Baltimore, who is not involved in the work, told New Scientist that DeepSouth will be a game changer for the study of neuroscience.

"If you are trying to understand the brain this will be the hardware to do it on," he said.

Etienne-Cummings said that there will be two main types of researchers who will be interested in the technology — those studying neuroscience, and those who want to prototype new engineering solutions in the AI space.

DeepSouth is just one of many research projects aiming to create a machine that will rival the human brain.

Other researchers are trying to tackle the same problem by creating "biological computers" powered by actual brain cells.

Source

Intel launches its first Core Ultra processors with a big focus on AI

Thu, 14 Dec 2023 17:57:17 +0000

During its AI Everywhere event, Intel unveiled the new Core Ultra lineup for laptops and tablets. It consists of eleven models with different form factors and price tags. The new CPU family is based on the Meteor Lake architecture, and it promises significant performance and efficiency improvements, notably better built-in graphics, and a big focus on artificial intelligence.

The launch of Intel Core Ultra represents the unmatched scale and speed at which Intel is enabling AI on the PC. By 2028, AI PCs will comprise 80% of the PC market and together with our vast ecosystem of hardware and software partners, Intel is best positioned to deliver this next generation of computing.

Intel Core Ultra is built using Intel's 4 process technology (formerly referred to as 7nm) and Foveros 3D packaging. According to Intel, the Core Ultra lineup marks the largest architectural shift in 40 years of chipmaking. The redesigned P-core architecture delivers better performance thanks to improved instructions-per-cycle (IPC), and new E-cores enable improved energy efficiency.

The company boasts that the new Core Ultra chips offer up to 79% lower power consumption in the same 28W envelope compared to AMD's Ryzen 7 7840U when idle on Windows desktop. In other areas, customers can also expect improvements, such as 48% better power performance when streaming Netflix, 44% when playing local 4K video, or 7% when browsing the web.

The lineup offers up to 16 cores/22 threads clocked at up to 5.1GHz with support for up to 64GB of LPDDR5x RAM or 96GB DDR5 memory. Like other newest Intel processors, the Core Ultra lineup supports Wi-Fi 6E and discrete Wi-Fi 7, Thunderbolt 5, and Bluetooth 5.4 with LE Audio.

Intel also improved upon its built-in graphics, equipping new processors with up to eight Xe cores with DX12 Ultimate and XeSS support. The company claims customers can expect double the graphics performance over previous-gen processors with iGPUs. Other features include hardware-accelerated ray tracing, AV1 encode/decode, and HDMI 2.1 and DisplayPort 2.1 support.

Finally, Intel has a big focus on artificial intelligence and its new Neural Processing Unit called Intel AI Boost, promising "longer-running AI workloads" at low power in conjunction with CPU and GPU for better performance and up to 2.5x better efficiency. You can expect these chips to play nice with Microsoft's upcoming next-generation Windows client scheduled for the second half of 2024.

Intel Core Ultra H and U-Series are available on the market right now. The Intel Core Ultra 9 185H with 45W TDP will arrive in the first quarter of 2024 alongside ultra-low-powered 9W models. Intel says its partners prepared over 230 PC designs.

Source

Doom’s creators reminisce about “as close to a perfect game as anything we made”

Thu, 14 Dec 2023 17:53:42 +0000

In 30th anniversary stream, Carmack and Romero recall a game dev "perfect storm."

The archived hour-long chat is a must-watch for any long-time Doom fan.

While Doom can sometimes feel like an overnight smash success, the seminal first-person shooter was far from the first game created by id co-founders John Carmack and John Romero. Now, in a rare joint interview that was livestreamed during last weekend's 30th-anniversary celebration, the pair waxed philosophical about how Doom struck a perfect balance between technology and simplicity that they hadn't been able to capture previously and have struggled to recapture since.

Carmack said that Doom-precursor Wolfenstein 3D, for instance, "was done under these extreme, extraordinary design constraints" because of the technology available at the time. "There just wasn't that much we could do."

Wolfenstein 3D's grid-based mapping led to a lot of boring rectangular rooms connected by long corridors.

Steam

One of the biggest constraints in Wolfenstein 3D was a grid-based mapping system that forced walls to be at 90-degree angles, leading to a lot of large, rectangular rooms connected by long corridors. "Making the levels for the original Wolfenstein had to be the most boring level design job ever because it was so simple," Romero said. "Even [2D platformer Commander Keen] was more rewarding to make levels for."

By the time work started on Doom, Carmack said it was obvious that "the next step in graphics was going to be to get away from block levels." The simple addition of angled walls let Doom hit "this really sweet spot," Carmack said, allowing designers to "create an unlimited number of things" while still making sure that "everybody could draw in this 2D view... Lots of people could make levels in that."

Doom's support for angled walls and variable heights added a huge amount of design variation

while still keeping things relatively simple to edit.

ZDoom forums / CBM

Romero expanded on the idea, saying that working on top of the Doom engine was, at the time, "the easiest way to make something that looks great. If you want to get anything that looks better than this, you're talking 10 times the work."

Was Quake too complex?

Then came Quake, with a full 3D design that Carmack admitted was "more ambitious" and "did not reach all of its goals." When it came to modding and designing new levels, Carmack lamented how, with Quake, "a lot of potentially great game designers just hit their limit as far as compositional aesthetic in terms of what something is going to look like."

"When you have the ability to do a full six degree-of-freedom modeling, you not only have to be a game designer, you have to be an architect, a modeler working through your composition," Carmack continued. "[Doom] helped you along by keeping you from wasting time doing some crazy things that you would have had to be a master of a different craft to pull off... Going to full 3D made this something that not everybody does on a lark, but something you set out time, almost set a career arc to make mods for newer games."

Quake's full 3D maps required much more design skill to build good-looking levels.

You Got Red On You

Romero reminisced during the chat about "the Everest of Quake" and "the insane amount of technology we took on" during its development. "Even adding QuakeC on top of [a] client/server [architecture] on top of full 3D, it was so much tech. It was a whole new engine, it wasn't Doom at all. It was all brand new."

Looking back, Carmack allowed that "there's a couple different steps we could have taken [with Quake], and we probably did not pick the optimal direction, but we kept wanting to make this, just throw everything at it and say, 'If you can think of something that's going to be better, we should just strive our hardest to do that.' When Doom came together, it was just a perfect storm of 'everything went right.'.. [it was] as close to a perfect game as anything we made."

Squashing the drama

During the roughly hour-long moderated livestream interview, Romero and Carmack generally avoided rehashing the drama of their storied professional breakup in the '90s. Playing off the moderator's mention of comparisons between id's Carmack/Romero and the Beatles' Lennon/McCartney, Carmack talked about the two Johns being treated as some of the first "rock stars" of the game development world and teased that Romero "leaned into it a little bit."

Carmack remembered Doom's heyday as a time when "the world was fascinated with 'video games are no longer this Atari 2600 that you plug into your TV.' It suddenly looks cool enough for actual mainstream press to talk about. And any time the mainstream press is going to turn the 'eye of Sauron' on you, there's going to be a focus on the interpersonal drama and figuring out everything else they can do to make an interesting or exciting story about it. It was almost inevitable."

John Carmack (left) and John Romero (second from right) pose with their id Software colleagues in the early '90s.

John Romero

Romero, on the other hand, said the Lennon/McCartney comparison resonated for him because "I think we both spoke the same language because I also programmed. If you have a designer that can speak code, it's a lot easier to know what to compromise on rather than coming up on a very unrealistic idea for a design thing. A lot of times, the designers don't know what can be done, or even a simple feature could be really complex, but change a couple of words and we can do that right now."

The beauty of “dumb” enemies

These days, Carmack spends most of his professional time in search of so-called artificial general intelligence. But when it came to designing Doom 30 years ago, Carmack says he found making enemy AI dumber actually made for a better gaming experience.

"We could have made the enemies [in Doom] a lot smarter, even with the constraints we had back then," Carmack said. But creating super-smart enemies "is not as good of an idea as people might think," he added, because "the world has to feel like it revolves around the player for the types of games we were doing."

"There's something to be said for exploring a game that exists with or without you, but the enemies [in Doom] are mostly there to have an effect on the players, not to efficiently do something and not to go about their lives," Carmack said. "For the types of games we were doing, you don't really want particularly smart enemies. You don't want them to do pathologically dumb-looking things that break the player's immersion in something. But you do not really want them picking a hiding spot [and] flanking the player from behind. That's not the enjoyable thing."

While some players might enjoy facing realistic, super-intelligent enemies, Carmack said he found that most players "want to go in and Rambo their way through everything, and that doesn't take a whole lot of intelligence in the enemies."

Fighting one super-smart enemy is a lot less fun then mowing down a room full of dumb demons.

Romero agreed during the livestream that "having more of something dumb [on-screen] is better than having one super-smart AI," a lesson that he urged developers to heed today. "People are still trying to push the envelope on having smarter enemies. Well, people are like, 'That's too smart,' people want to turn it off. So I think we made the right decision in Doom with having enemies that moved how they did, which was typically pretty slow, but we had huge numbers of them, and that was more interesting because you had to be more strategic."

But Carmack suggested that the state of modern AI technology could soon change that calculus for certain types of games. "We are approaching a new era with the way artificial intelligence stuff is going right now," he said, "where the world of actual deep characters is going to be very different than it has been in the past years, where you can create something today that does have entire deep backstories that will, for certain types of games, be very transformative... AI stuff is definitely the future, but back then, I think we did make the right calls [for Doom]."

While Doom owes a lot of its longevity to its "sweet spot" simplicity, Romero also pointed to the 1997 release of the game's source code as something that has "really made the game live [because] now it can be kept current on everything, it can now run on everything."

But Carmack lamented that Doom's openness to modification hasn't been embraced by the wider industry. "You don't see game companies release source code or even modding tools," he said. "Reverse engineers have better tools... but it's really not with the support of the studios, and that's really a shame."

Listing image by Id

Source

Intel To Release 14-Gen Non-K CPUs In January Next Year

Thu, 14 Dec 2023 03:08:09 +0000

In addition to the release dates for the non-K 14-Gen Intel i9-14900, i7-14700, i5-14600, i5-14500. The details of i3-14100 and i5-14400 processors leak too.

About a year ago, Intel released the Raptor Lake series of the processors. After that, everyone thought that Intel will succeed it with Meteor Lake series, which uses an entirely new architecture.

However, to everyone’s surprise, Intel decided to not release Meteor Lake for desktops and reserve it only for notebooks and other mobile PCs.

Instead, Intel decided that for desktops, it will release a refresh for the Raptor Lake series in the 14-gen version of its CPUs. This refresh, in simple words, is nothing but the same Raptor Lake series processors with minor improvements in the clock rates.

About a month ago, Intel released the K series version of unlocked Intel 14-Gen Raptor Lake Refresh processors. This included the Core i9-14900K/F, i7-14700K/F and i5-14600K/F processors. For the record, K series processors allow overclocking, while non-K processors don’t.

However, non-K processors in the series wasn’t released or announced. It looks like that’s going to change now.

Intel 14-Gen Non-K CPUs To Release In January

Well known Twitter (now X) based leaker ECSM_Official has leaked (via VideoCardz) the release dates for the 14-Gen Non-K Intel Raptor Lake Refresh CPUs.

The tweet reveals that these CPUs will be released on the 8th of January 2024. Which is just a day before a major tech show starts.

The non-K CPUs that are likely to be in the Intel 14-Gen Raptor Lake Refresh series are going to be:

Intel Core i9-14900 / F
Intel Core i7-14700 / F
Intel Core i5-14600 / F
Intel Core i5-14500 / F
Intel Core i5-14400 / F
Intel Core i3-14100 / F

The F versions, which is an additional option, come without any built-in iGPU inside the CPU and require a dedicated graphics to run the computer.

i5-14400 And i3-14100 Specs Leak

In addition to the release dates, ECSM_Official in his tweet and followup tweet has also leaked the specs for the i5-14400 And i3-14100 CPUs.

As expected, the screenshots show that the Intel i5-14400 will come with 6 P-Cores and 4 E-Cores equaling 16 threads, running at 4.1GHz and 3.5GHz respectively. With the package power reaching 85W in the test.

Meanwhile, the screenshots also detail the specs for Intel i3-14100. It comes just 4-P Cores and no E-Cores, with 8 total threads. It seems to be running at 4.5GHz with the CPU package power reaching just 68W.

Interestingly, the i3-14100 seems to be based on the H0 silicon, which basically means it’s nothing but the same Alder Lake generation CPU being refreshed and sold as 14-Gen Raptor Lake Refresh.

The tweet gives screenshots of synthetic benchmarks for both the processors, so those who want to check them should. However, synthetic benchmarks only give some idea of real world usage, the best are gaming benchmarks.

The reviews, we expect, would come on or around the release dates. Only then will we actually come to know. They will tell us how faster, if at all, the 14-Gen Raptor Lake Refresh CPUs are going to be when compared to previous-gen Raptor Lake CPUs.

Source

Microsoft ranked the BEST-MANAGED company for 4 years consecutively

Thu, 14 Dec 2023 03:05:40 +0000

Microsoft is at the top of the food chain, having ranked as the best-managed company of 2023, ahead of Apple and NVIDIA. It scored highly across all categories except customer satisfaction.

What you need to know

Microsoft has been ranked as the top best-managed company in 2023 by researchers at the Drucker Institute.
Apple and NVIDIA followed closely in second and third place respectively.
Growth, employee retention, and customer satisfaction were among the main metrics and categories used while ranking.
Microsoft scored significantly high across most of the categories except customer satisfaction where it came in at 148th place.

We're edging closer to the end of the year, as is the norm. It's usually around this time that we see how sought-after tech firms performed throughout the year in terms of profits plowed back, growth, employee retention, and more.

In September, TIME magazine ranked Microsoft as the best company in the world based on employee satisfaction and other metrics. As it turns out, this isn't the only award Microsoft is bagging this year. According to a spot by The Wall Street Journal, Microsoft has now been ranked as the best-managed company in 2023 by researchers at the Drucker Institute.

Apple came in second place, followed closely by NVIDIA, which isn't entirely surprising. Its revenue for Q3 2023 stands at $18.12 billion, and its profits are $10.42 billion, which makes it the most profitable semiconductor chip brand ahead of Intel and Samsung.

The researchers used various metrics and data when compiling the list, including job satisfaction, supply-chain management, shareholder returns, innovation, and more. Microsoft excelled extremely well across most categories, except customer satisfaction, which came in at 148th place.

While Microsoft's deep-dive into AI after its multi-billion dollar investment in the technology seems to be doing the trick as far as innovation is concerned, the move to "evolve" its Microsoft Rewards program with a cool-down period isn't doing it any favors in the customer satisfaction front.

Microsoft's unbeatable winning streak

Microsoft CEO Satya Nadella. (Image credit: Daniel Rubino)

This is the fourth year in a row that Microsoft claims the top spot as the best-managed company. While the secret to the company's immense success might remain a secret, Drucker Institute's Executive Director, Michael Kelly, shared the following sentiments:

"Microsoft and other top-performing companies who have sustained themselves at the top of the rankings look to have developed a virtuous cycle that continuously reinforces their successes. This allows these companies to secure the resources and capital needed to innovate and secure talent, which in turn allows them to achieve greater and greater successes."

Over the years, Microsoft's CEO, Satya Nadella, has mentioned the importance of empathy and its role in determining employee satisfaction. He terms it the most challenging skill to learn, yet vital to run any business successfully.

What do you think is Microsoft's secret formula for running a successful business? Share your thoughts in the comments.

Source

Turing test on steroids: Chatbot Arena crowdsources ratings for 45 AI models

Thu, 14 Dec 2023 03:04:03 +0000

Over 130K blind ratings show ChatGPT-4 Turbo outclassing the competition.

As the AI landscape has expanded to include dozens of distinct large language models (LLMs), debates over which model provides the "best" answers for any given prompt have also proliferated (Ars has even delved into these kinds of debates a few times in recent months). For those looking for a more rigorous way of comparing various models, the folks over at the Large Model Systems Organization (LMSys) have set up Chatbot Arena, a platform for generating Elo-style rankings for LLMs based on a crowdsourced blind-testing website.

Chatbot Arena users can enter any prompt they can think of into the site's form to see side-by-side responses from two randomly selected models. The identity of each model is initially hidden, and results are voided if the model reveals its identity in the response itself.

The user then gets to pick which model provided what they judge to be the "better" result, with additional options for a "tie" or "both are bad." Only after providing a pairwise ranking does the user get to see which models they were judging, though a separate "side-by-side" section of the site lets users pick two specific models to compare (without the ability to contribute a vote on the result).

A blind test on our old favorite "Who invented video games?" prompt. Note that Model B goes on

for much longer if you scroll (and mistakenly says Nintendo and Atari were making video games

in the '60s)

https://chat.lmsys.org/?arena

Since its public launch back in May, LMSys says it has gathered over 130,000 blind pairwise ratings across 45 different models (as of early December). Those numbers seem poised to increase quickly after a recent positive review from OpenAI's Andrej Karpathy that has already led to what LMSys describes as "a super stress test" for its servers.

Chatbot Arena's thousands of pairwise ratings are crunched through a Bradley-Terry model, which uses random sampling to generate an Elo-style rating estimating which model is most likely to win in direct competition against any other. Interested parties can also dig into the raw data of tens of thousands of human prompt/response ratings for themselves or examine more detailed statistics, such as direct pairwise win rates between models and confidence interval ranges for those Elo estimates.

And the winner is...

Chatbot Arena's latest public leaderboard update shows a few proprietary models easily beating out a wide range of open-source alternatives. OpenAI's ChatGPT-4 Turbo leads the pack by a wide margin, with only an older GPT-4 model ("0314," which was discontinued in June) coming anywhere close on the ratings scale. But even months-old, defunct versions of GPT-3.5 Turbo outrank the highest-rated open-source models available in Chatbot Arena's testbed.

Chatbot Arena's current leaderboards show GPT-4 Turbo and other proprietary models are still the best in class.

Chatbot Arena

Anthropic's proprietary Claude models also feature highly in Chatbot Arena's top rankings. Oddly enough, though, the site's blind human testing tends to rank the older Claude-1 slightly higher than the subsequent releases of Claude-2.0 and Claude-2.1.

Among the tested non-proprietary models, the Llama-based Tulu 2 and 01.ai's Yi get rankings that are comparable to some older GPT-3.5 implementations. Past that, there's a slow but steady decline until you get to models like Dolly and StableLM at the bottom of the pack (amid older versions of many models that have more recent, higher-ranking updates on Chatbot Arena's charts).

This kind of ranking system has its flaws, of course. Humans may be ill-equipped to accurately rank chatbot responses that sound plausible but hide harmful hallucinations of incorrect information, for instance. Chatbot Arena users may also naturally gravitate towards certain types of prompts that favor certain types of models; LMSys's own LLM-assisted analysis of user-submitted Arena prompts finds requests for role-playing professional personas, writing stories, and "exploring ethical dilemmas and societal norms" among the most popular categories.

To balance out these potential human biases, LMSys has also developed a completely automated ranking system called LLM Judge, which uses LLM models themselves to rank the quality of responses from other LLMs into an "MT-Bench" score. Those rankings are also compared to a standardized MMLU evaluation, which ranks models on a variety of common tasks.

The most common types of prompts users submit to Chatbot Arena's system.

LMSys

LMSys's academic paper on the subject finds that "strong LLM judges like GPT-4 can match both controlled and crowdsourced human preferences well, achieving over 80% agreement, the same level of agreement between humans." From those results, the organization suggests that having LLMs rank other LLMs provides "a scalable and explainable way to approximate human preferences, which are otherwise very expensive to obtain."

Indeed, comparing the different ranking methods on Chatbot Arena's leaderboards finds broadly similar standings. There are some differences, though; MT-Bench ranks UC Berkeley's Starling model as better than some versions of ChatGPT and Claude, while the MMLU tests rank the Yi model alongside the best proprietary models.

While LMSys promises real-time leaderboard updates will be coming soon, the current leaderboards only reflect models added last month. More recently, LMSys revealed in a social media post that the new "mixture of experts" Mixtral model has shown strong (but not proprietary-level) results in early blind trials. We can't wait to see how models like Google's Gemini or even Elon Musk's Grok fare in future direct competition.

Source

Elon Musk’s X ad revenue reportedly fell $1.5B this year amid boycotts

Thu, 14 Dec 2023 03:02:17 +0000

"We are not Twitter any longer," X exec said.

It's hard to know exactly how dire the financial situation is at Elon Musk's X (formerly Twitter). However, insider sources recently revealed to Bloomberg that the social media platform expects to end 2023 with "roughly" $2.5 billion in advertising revenue.

That's "a significant slump from prior years," sources said. It's also about half a billion short of the $3 billion that X executives expected to make in ad sales in 2023, one source said.

Last year, Twitter raked in more than $1 billion in ad revenue per quarter, sources said. But in each of the first three quarters of 2023, X only managed to generate "a little more than $600 million" in ad revenue. Now, the most recent advertiser fallout over antisemitic content on X—estimated in November as triggering a sudden $75 million loss—is still casting a shadow on what could become an even more dismal fourth quarter.

Sources told Bloomberg that advertising earnings comprise 70–75 percent of X's revenue. This suggests that X's total earnings in 2023 will be "roughly $3.4 billion," boosted by subscriptions and data licensing deals, Bloomberg reported.

Joe Benarroch, head of X business operations, told Bloomberg that its report "presents an incomplete view of our entire business, as the sources you’re relying on for information are not providing accurate and comprehensive details."

Benarroch also made it clear that X is no longer interested in being compared to Twitter. According to Bloomberg, Benarroch said that X is an “evolving NEW global business with multiple revenue streams. We are not Twitter any longer and not measuring ourselves by old Twitter metrics—both in revenue and user metrics.”

X did not immediately respond to Ars' request to comment.

Musk beefing with Disney on X

After Musk boosted an antisemitic post on X, he apologized, but he never removed his controversial post and continued antagonizing advertisers that he claimed were "going to kill the company."

Among the major brands pausing advertising on X is Disney, which seems to have particularly offended Musk. He's spent the past week targeting Disney CEO Bob Iger in a series of X posts, calling out Disney for boycotting X. Musk appears particularly frustrated that Disney is advertising on Meta platforms after New Mexico Attorney General Raúl Torrez filed a lawsuit alleging that Facebook and Instagram are "prime locations for predators to trade child pornography and solicit minors for sex."

Last December, Musk was similarly beefing with Apple through a Twitter tirade that Musk admitted later was due to his own "misunderstanding" about Apple's decision-making. That beef didn't seem to impact any of X's other advertiser relationships long-term, but the latest advertiser boycott does not appear that it will be resolved quickly. Musk's continued inflammatory statements about advertisers and the platform's recent decision to reinstate the X account of Sandy Hook shooting-denier Alex Jones could keep advertisers off X. That, as Musk has warned, risks causing serious long-term damage to X as a viable platform.

Although Twitter/X stopped publicly reporting financial data when Musk took over Twitter in 2022, Musk has been transparent about advertiser revenue being down by "roughly 50 percent" throughout 2023. In September, Musk confirmed that things were getting worse, reporting that ad sales were down by 60 percent and blaming groups monitoring hate speech for spooking advertisers off with their #StopToxicTwitter campaign.

Musk has always said that his goal for X is to reduce reliance on advertising by pivoting to subscriptions he hoped would one day comprise half of X's earnings. But X subscriptions are not as popular as planned, sources told Bloomberg. So far, X only has "just over 1 million paying subscribers"—out of hundreds of millions of X users—and X's subscription revenue is "less than $120 million annually," sources said.

X has rarely been profitable, but under Musk, its financial growth so far appears to have been significantly set back. Before Musk took control, Twitter earned more than $5 billion in 2021, Bloomberg reported, and that year Jack Dorsey "set a public goal to reach $7.5 billion in revenue by the end of 2023." That projection makes the current $3.4 billion estimate for 2023 appear especially stark.

As many big brands remain hesitant to advertise on X, the platform is offering deals hoping to attract smaller brands to invest more in the platform.

Generating more ad revenue is crucial not just to meet X's financial goals in 2024 but also to hit Musk's projection of 1 billion monthly X users in 2024. Musk's plan to attract more users largely depends on content creators lured to increase engagement on X by lucrative ad revenue sharing recently launched by the platform.

Creators have already voiced concerns that the current boycott will impact their ability to profit on the platform, and so far, Musk has only said that there's "not much we can do if advertisers boycott or reduce spend on our platform."

Source

Tesla unveils its latest humanoid robot, Optimus Gen 2, in demo video

Wed, 13 Dec 2023 19:24:55 +0000

Tesla says its new prototype is 30% faster, 10 kg lighter, and has sensors on all fingers.

The Tesla Optimus Gen 2 robot.

Tesla | Benj Edwards

On Tuesday, Tesla released a demo video showing the latest version of its prototype humanoid robot, Optimus Gen 2. Over one year after Tesla's first public Optimus robot demonstration, which showcased shaky robots that waved and slumped over, things have apparently progressed quite a bit, assuming that the video accurately reflects the technology.

"Everything in this video is real, no CGI. All real time, nothing sped up. Incredible hardware improvements from the team," wrote Tesla Senior Staff Software Engineer Julian Ibarz on X.

After a recent episode where Google fudged an AI demonstration for the sake of marketing hype, it's best to take Tesla's claims with a grain of salt until they are independently verified in practical, real-world demonstrations. With that dose of skepticism in mind, let's take a look at what Tesla is promising in this non-production prototype robot.

The promotional video opens with a shot of "Bumblebee," Tesla's first humanoid prototype from September 2022. Then it shows "Optimus - Gen 1" from March 2023. And after some teasing shots of a different robot, the video puts text on the screen that reads, "Introducing Optimus - Gen 2. December 2023." As the robot performs various tasks, such as walking slowly, crouching, manipulating eggs without breaking them, and waving its arms, it lists out these features of Gen 2:

Tesla-designed actuators and sensors
2-DoF (degrees of freedom) actuated neck
Actuators-integrated electronics and harnessing
30 percent walk speed boost
Foot force/torque sensing
Articulated toe sections
Human foot geometry
10 kg total weight reduction
Improved balance and full-body control
Faster, 11-DoF brand-new hands
Tactile sensing on all fingers
Delicate objects manipulation (holding an egg)

As previously mentioned, Optimus Gen 2 is not designed for production or sale. As a prototype, it represents a potential waypoint on a much longer journey to a more functional human robot that Tesla CEO Elon Musk says will be able to do anything that humans don't want to do. It's designed to be a similar shape and size as humans so it can replace human labor in a seamless way. Whatever the application, simply swap out a person and insert an Optimus.

Given difficulties in engineering, if and when that human labor replacement will actually happen still remains to be seen, but if what's shown in the video is true, it's looking like Tesla is making significant process toward its goal.

Source

Starfield is getting city maps, new ways to travel, FSR 3, and more features in 2024

Wed, 13 Dec 2023 03:34:32 +0000

Bethesda has been pushing out updates to Starfield at a steady pace since the RPG's launch in September, but they have mostly included bug fixes and small changes. While the final update of 2023 has already landed for the Xbox title, the studio has now given a rundown of what it's working on to ship in 2024.

The information drop arrived via a Reddit post by an official Bethesda Game Studios account. It said that in early next year, a larger game update would arrive with fixes for "in-progress" stuck quest. "Though we fixed several quest issues from occurring, in-progress quest fixes are much harder to fix and we’ve built a new system to correct those without you having to roll back your save," explains the studio.

The title received DLSS support with frame generation (DLSS 3) in November, and soon, those looking forward to AMD FidelityFX Super Resolution 3 (FSR 3) and Intel XeSS will have their day too. The same early 2024 update will have these tech for use on almost any graphics card.

New game features are also incoming sometime in 2024. This will include maps that actually provide useful information when exploring cities as well as the promised editor tools on PC and in-game support for installing mods.

Bethesda also teased another interesting addition, saying "all new ways of traveling" are coming to Starfield with an update. It's unclear if this means players are finally getting vehicles to explore planets faster, reduced loading screens to offer more seamless transitions, or perhaps even alien mounts.

As for when can players expect these features, Bethesda said this:

"These will be rolling out with a regular cadence of fixes and updates we expect to have roughly every six weeks. If something can be done in a smaller hotfix in between (like the asteroid), and we feel it’s safe, we’ll get one of those out as well. Safe is the key here. We do take a lot of time to test even the smallest change in a game this large and dynamic."

The studio staff will be going on holiday before returning to development in 2024. Starfield is also slated to receive story expansions, though a release time frame has not been announced yet.

Source

Epic Games wins in court as Jury says Google Play Store is an illegal monopoly

Tue, 12 Dec 2023 18:01:48 +0000

Epic Games has emerged as the winner of its three-year legal fight against Google over its running of the Play Store. As reported by The Verge, the jury in the month-long trial in San Francisco, only a few hours after it started deliberations earlier today, has ruled that the Google Play Store, and its billing service, is an illegal monopoly.

In the jury's verdict, it agreed with Epic's argument that Google not only has a monopoly on apps for Android via the Google Play Store, but the company also engaged in anti-competitive practices by paying developers to keep that monopoly in place.

Epic Games posted its reaction to the jury's verdict on the case on its blog, stating:

Today’s verdict is a win for all app developers and consumers around the world. It proves that Google’s app store practices are illegal and they abuse their monopoly to extract exorbitant fees, stifle competition and reduce innovation.

Over the course of the trial we saw evidence that Google was willing to pay billions of dollars to stifle alternative app stores by paying developers to abandon their own store efforts and direct distribution plans, and offering highly lucrative agreements with device manufacturers in exchange for excluding competing app stores.

The Verge also received a statement on today's verdict from Wilson White, Google VP, Government Affairs & Public Policy, which indicated the company does plan to appeal the ruling:

We plan to challenge the verdict. Android and Google Play provide more choice and openness than any other major mobile platform. The trial made clear that we compete fiercely with Apple and its App Store, as well as app stores on Android devices and gaming consoles. We will continue to defend the Android business model and remain deeply committed to our users, partners, and the broader Android ecosystem.

Epic Games first filed a lawsuit against Google in August 2020. This happened after Epic launched its own in-game monetary system for its hit game Fortnite, which went against the policies of both the Google Play Store. Epic filed a similar lawsuit with Apple in 2020. However, the subsequent trial and appeals ended up with Apple winning most of the case.

In the Epic vs Google trial, Epic was not seeking any monetary damages if it won, but it does want Google to completely free up Android to allow it to support full game app stores, with their own billing practices. Those requests will be taken up by the judge in the case in early January 2024.

Source

E3 is officially no more as the ESA shuts down the long running video game trade show

Tue, 12 Dec 2023 17:56:09 +0000

Although it's not exactly shocking news, today the Entertainment Software Association confirmed that it will no longer organize and hold the Electronic Entertainment Expo trade show, better known as E3. The official announcement was made on the ESA's X (formerly Twitter) account.

In an interview with the Washington Post, ESA president Stanley Pierre-Louis went into more detail about the decision to end E3. He stated:

"There were fans who were invited to attend in the later years, but it really was about a marketing and business model for the industry and being able to provide the world with information about new products,” he said. “Companies now have access to consumers and to business relations through a variety of means, including their own individual showcases.”

The decision to hold a trade show devoted to the video game industry resulted in the first E3 in 1995 in Los Angeles. With a few exceptions over the past few decades, E3 was always held at the LA Convention Center, where game publishers big and small exhibited new and upcoming console and PC games to the press and retail partners, and later to the general public.

The last physical E3 show to be held was in 2019. The COVID-19 pandemic caused the show to be completely canceled in 2020 and 2022, with an all digital event held in 2021. The ESA announced plans to bring back the show as a physical event to the LA Convention Center for June 2023. However, after many of the major game publishers announced they would not attend the show, the ESA canceled those plans in March 2023.

The big question now is: What, if anything, will replace E3 for game announcements in June? It's possible that publishers like Microsoft, Nintendo, Sony and others will hold their own separate digital showcases during that period. There's also the Summer Games Fest, which has been a big digital showcase of upcoming games for the last few years. The next edition will be held sometime in June 2024.

Source

The latest Starfield update is a small one but it fixes an annoying issue with "space rocks"

Tue, 12 Dec 2023 04:21:57 +0000

Bethesda Game Studios is rolling out a small update today for its recent space-based RPG Starfield. The update, which was previously made available in beta form for people who purchased the Steam version of the game, is a small patch, but it does fix a rather annoying bug that many players have encountered while piloting their ship in the game.

The 1.8.88 update for Starfield should keep outside debris from getting stuck on player ships, which has been labeled as "pet rocks" by some players. Here is the full changelog for the patch:

FIXES AND IMPROVEMENTS

GAMEPLAY:

Addressed an issue that would cause space matter to become stuck to player's ship during space travel. Loading a save will now remove the space clingon. Please note: This fix should address any space matter being stuck in your travels, but not in instances where player ships have New Atlantis attached. A fix for that will be released in a later update.

OUTPOST:

Fixed an issue that prevents random guns from spawning in a newly created Weapon Case after loading a save.

SAVE/LOAD:

[MSS/Xbox] Fixed an issue where players could experience crashes while saving during long playthrough without going through the Unity.

Starfield's latest major update on November 20 added official support for NVIDIA's DLSS support technology for PC players who have that company's GPUs. We are still waiting for the promised addition of AMD's FidelityFX Super Resolution 3 support. The game has supported AMD's FSR 2 upscaling tech since it was first released in September.

If you have yet to get Starfield, it's available on Xbox Series X|S consoles, along with PC via the Microsoft Store and Steam. It's available on Xbox Game Pass, PC Game Pass, and Xbox Game Pass Ultimate as well, which includes cloud gaming support. The game is also available to play for owners and PC Game Pass subscribers via NVIDIA GeForce Now.

Source

AI companion robot helps some seniors fight loneliness, but others hate it

Tue, 12 Dec 2023 04:20:47 +0000

There's limited evidence for health benefits so far; early work suggests no one-size-fits-all.

ElliQ, an AI companion robot from Intuition Robotics.

ElliQ

Some seniors in New York are successfully combating their loneliness with an AI-powered companion robot named ElliQ—while others called the "proactive" device a nag and joked about taking an ax to it.

The home assistant robot, made by Israel-based Intuition Robotics, is offered to New York seniors through a special program through the state's Office for the Aging (NYSOFA). Over the past year, NYSOFA has partnered with Intuition Robotics to bring ElliQ to over 800 seniors struggling with loneliness. In a report last week, officials said they had given out hundreds and had only 150 available devices.

ElliQ includes a tablet and a two-piece lamp-like robot with a head that lights up and rotates to face a speaker. Marketed as powered by "Cognitive AI technology," it proactively engages in conversations with users, giving them reminders and prompts, such as asking them how they're doing, telling them it's time to check their blood pressure or take their medicine, and asking if they want to have a video call with family. Speaking with a female voice, the robot is designed to hold human-like conversations, engage in small talk, express empathy, and share humor. It can provide learning and wellness programs, such as audiobooks and relaxation exercises.

Interest in using social robots, such as ElliQ, for elder care has been growing for years, but the field still lacks solid evidence that the devices can significantly improve health, well-being, and depression. Systemic reviews in 2018 found the technology had potential, but studies lacked statistical significance and rigorous design.

The program in New York adds to the buzz but doesn't offer the high-quality study design that could yield definitive answers. In August, the state released a report on an unspecified number of ElliQ users, which indicated that the device was helpful. Specifically, 59 percent of users reported the device was "very helpful" at reducing loneliness, while 37 percent reported it was "helpful" and only 4 percent reported it as "unhelpful." Engagement with the device declined over time, with users initially interacting with ElliQ an average of 62 times a day in the first 15 days of use, which fell to 21 times a day between 60 and 90 days and 33 times a day after 180.

Mixed feelings

"We had high hopes for the efficacy of ElliQ, but the results that we’re seeing are truly exceeding our expectations," Greg Olsen, director of the New York State Office for the Aging, said in a statement at the time of the report's release. "The data speaks for itself, and the stories that we’re hearing from case managers and clients around the state have been nothing short of unbelievable."

But other recent data on the potential for companion robots to reduce loneliness has indicated that there's no one-size-fits-all approach. There are a lot of factors that can influence how individuals perceive such a device. A 2021 qualitative study evaluated the responses from 16 seniors who were asked for feedback on three types of robot companions, including ElliQ. The results were mixed for the proactive robot. While some felt the occasional chattiness of ElliQ would be comforting during an otherwise solitary day, others felt it was intrusive and "nagging." Some felt the device's tone was "rude."

"I don’t know whether that would drive me mental if it kept interrupting me and telling me what to do ... I might want to get an ax and cut it up," one study participant said.

How welcoming a person might be to an assertive AI-assistant like ElliQ may link with a person's general preferences regarding human company, the authors suggested. Those who value their space and autonomy may be less open to such as device compared with more gregarious seniors.

While some participants said ElliQ's reminders could be useful, others expressed a deep concern that an overreliance on technology for everyday tasks—like paying bills, taking medications, or turning lights off—could hasten the decline of cognitive and physical abilities. Study participants also raised concerns regarding the inauthenticity of a relationship with a nonhuman, a loss of dignity, and a lack of control. Some disliked that ElliQ couldn't be fully controlled by the user and was so assertive, which some perceived as pushy. Some worried about feeling embarrassed about being seen interacting with a robot companion. A 2022 study also explored the issue of stigma, with participants expressing that the use of such devices could reinforce stereotypes of aging, including isolation and dependency.

While researchers continue to explore the potential use and design of AI-powered companion robots, anecdotes from New York's program suggest the tools are clearly helpful for some. One New Yorker named Priscilla told CBS News she found ElliQ helpful.

"She keeps me company. I get depressed real easy. She's always there. I don't care what time of day, if I just need somebody to talk to me," Priscilla said. "I think I said that's the biggest thing, to hear another voice when you're lonely."

Source