Hard disk drives (HDDs) are seven times cheaper than solid disks (SSDs), and have plenty more growth and development in them, says Toshiba.

The comments from Rainer Kaese, senior manager, HDD Business Development at Toshiba Electronics Europe, contradict claims from the SSD industry that hard drives will be obsolete in five years.

Kaese told Blocks and Files: “HDDs maintain a gap in cost per capacity with flash storage of around a factor of seven… HDDs may reach 40 or even 50 terabytes without approaching comparable costs with flash storage.”

As long as more density can be put on the platters, a hard drive costs the same amount to make, he argues, so cost per TB will go down as capacity goes up. Kaese says that cheap storage will fill automatically.

HDDs are potentially better for the environment, as they are more recyclable, forming part of a circular economy. They are largely made of aluminum and copper while SSDs contain more PCBs, chips, and plastic.

Aaron Klotz at Tom's Hardware says Kaese is exaggerating the price difference: "The cost of SSDs today is nowhere near the '7X' multiplier Kaese suggested. Most SSD prices currently are around three to four times more expensive than hard drives. It's possible that some enterprise-grade SSDs cost seven times as much as their hard drive counterparts, but that is not the norm."

In May, Pure Storage's VP of R&D Shawn Rosemarin from Pure Storage told Tom's Hardware that hard drives could go extinct in 2028, as manufacturers would simply cease because the power demands of HDDs would no longer be justifiable.

Klotz doesn't believe that either: "As far as we can tell, nothing is stopping hard drives from becoming more advanced and capable as long as manufacturers invest R&D dollars."

Source

The world’s biggest computing companies and a raft of well-funded startups all agree: The future of computing is manipulating data with quantum mechanics. Over the past decade, governments, private companies, and venture capitalists have collectively invested billions of dollars into quantum computing, which aims to solve problems using a new type of logic enabled by harnessing quantum properties such as superposition and entanglement rather than ordinary 1’s and 0’s. Yet despite some prototypes capable of elementary operations, the hardware isn’t reliable enough to be practically useful.

Researchers from Google and Colorado-based startup Quantinuum independently announced results this year that advanced a long-sought idea tipped to solve quantum hardware’s flakiness. Both teams demonstrated a mechanism needed for a component called a topological qubit, which should offer a way to hold onto and manipulate information encoded into quantum states more robustly than current hardware designs. But did the two companies actually make the long-sought component? It depends who you ask.

Physicists developed the design of the topological qubit to reduce computational errors, enabling more complex algorithms and opening the door to the technology’s projected moneymaking applications, from drug discovery to financial modeling to more efficient AI. “This could well be a transistor moment for the quantum computing industry,” said Quantinuum founder Ilyas Khan in the company’s announcement. “We have used a quantum computer as the machine tool for building topological qubits.”

Google performed a similar demonstration to Quantinuum’s, using the company’s own quantum computer, but they refuse to label it in the same way. “We did not realize a topological qubit,” write Google researchers Trond Andersen and Yuri Lensky in an email to WIRED. To add to the multiplicity of interpretations of the experiments, one of Google’s collaborators holds a different view.

The disagreement over what exactly is happening inside some of the leading prototype quantum computers points to the hurdles faced by the nascent industry. Researchers have developed elegant designs and exciting potential applications for the budding machines, but they struggle to execute them in practice.

Under pressure to demonstrate progress to their financial backers, and in need of significant funding to continue developing the technology, some in the quantum computing community have a tendency of announcing milestones of ambiguous significance—triggering pushback from some disgruntled physicists. Others take the opposite tack and cautiously qualify their achievements in technical jargon. (Google has been accused in the past of spreading hype, being among the first to announce a disputed quantum computing milestone and kicking off the trend that led up to Quantinuum’s topological qubit claim.)

Physicist Sergey Frolov of the University of Pittsburgh, who was not involved in the work, has despaired of the hype and says it’s clear to him that neither Quantinuum nor Google has created a topological qubit. “It does not take quantum computing to the next level,” he says. While Google and Quantinuum demonstrated that their hardware can exhibit several hallmark features of a topological qubit, the components are too fragile to fulfill their intended role in quantum computing—to hold onto and manipulate information robustly. “To me, a topological qubit has protection,” Frolov says. “This doesn’t.”

Bold Claims

Google’s project to explore the underpinning of one of the holy grails of quantum computing emerged from a collaboration with a team at Cornell University led by physicist Eun-Ah Kim. They published the results in the scientific journal Nature in May.

Topological qubits store and work with digital information in quirky minuscule objects known as non-Abelian anyons, which emerge from the collective behavior of electrons or other particles confined to a flat surface. Unlike most particles that physicists and engineers work with, non-Abelian anyons are not defined by the material they’re made of but by how they behave owing to their geometric properties. The designation “anyon” is similar to “wave,” in that a wave is defined by its behavior and can consist of water, air, a metal guitar string, or a variety of materials. Researchers have proposed making non-Abelian anyons out of clusters of electrons, ions, neutral atoms, and superconducting circuits.

Most crucially, non-Abelian anyons retain a sort of “memory” of their past movement that can be used to represent binary data. At its simplest, a topological qubit is a system that encodes data into the properties of pairs of non-Abelian anyons by physically swapping them around with each other in space. These changes are robust to outside disturbances such as tiny vibrations or fluctuations in temperature that can overwhelm other qubits—resilience that is rooted in math from the field of topology, the study of spatial relationships and geometry that hold true even when shapes are distorted. This makes non-Abelian anyons desirable components for holding and manipulating information in quantum computers.

Kim’s team made non-Abelian anyons using 25 superconducting circuits that make up one of Google’s quantum computers. They showed that after moving the non-Abelian anyons around, they did retain a memory of their past motion. Kim says that leaves no room for dispute, as that “memory” is the device’s signature design feature. “We made a topological qubit,” she says.

Andersen and Lensky of Google disagree. They do not think the experiment demonstrates a topological qubit, because the object cannot reliably manipulate information to achieve practical quantum computing. “It is repeatedly stated explicitly in the manuscript that error correction must be included to achieve topological protection and that this would need to be done in future work,” they write to WIRED.

When WIRED spoke with Tony Uttley, the president and COO of Quantinuum, after the company’s own announcement in May, he was steadfast. “We created a topological qubit,” he said. (Uttley said last month that he was leaving the company.) The company’s experiments made non-Abelian anyons out of 27 ions of the metal ytterbium, suspended in electromagnetic fields. The team manipulated the ions to form non-Abelian anyons in a racetrack-shaped trap, and similar to the Google experiment, they demonstrated that the anyons could “remember” how they had moved.

Quantinuum published its results in a preprint study on arXiv without peer review two days before Nature published Kim’s paper.

Room for Improvement

Ultimately, no one agrees whether the two demonstrations have created topological qubits because they haven’t agreed on what a topological qubit is—even if there is widespread agreement that such a thing is highly desirable. Consequently, Google and Quantinuum can perform similar experiments with similar results but end up with two very different stories to tell.

Regardless, Frolov at the University of Pittsburgh says that neither demonstration appears to have brought the field closer to the true technological purpose of a topological qubit. While Google and Quantinuum appear to have created and manipulated non-Abelian anyons, the underlying systems and materials used were too fragile for practical use.

David Pekker, another physicist at Pittsburgh, who previously used an IBM quantum computer to simulate the manipulation of non-Abelian anyons, says that the Google and Quantinuum projects don’t showcase any quantum advantage in computational power. The experiments don’t shift the field of quantum computing from where it has been for a while: Working on systems that are too small-scale to yet compete with existing computers. “My iPhone can simulate 27 qubits with higher fidelity than the Google machine can do with actual qubits,” Pekker says.

Still, technological breakthroughs sometimes grow from incremental progress. Delivering a practical topological qubit will require all kinds of studies—large and small—of non-Abelian anyons and the math underpinning their quirky behavior. Along the way, the quantum computing industry’s interest is helping further some fundamental questions in physics.

Nobel Laureate and theoretical physicist Frank Wilczek first proposed the existence of anyons, of which non-Abelian anyons are a specific type, in 1982, outside of the context of technology. Anyons, if they existed, constituted a fundamentally new category of matter, separate from previously recognized particles such as electrons and photons. Wilczek’s idea expanded the types of objects nature would allow, and for years physicists tried to conclusively show that non-Abelian anyons existed. While physicists previously found evidence pointing to their existence, Quantinuum’s and Google’s work are the first ever to demonstrate their signature feature, their “memory” of their movement.

In fact, Kim says she wanted to do this experiment to study anyons themselves, not because she was striving to build a better quantum computer. “My personal motivation was pure science, outside of any hype or any computing motivations,” she says.

As a theoretical physicist, Kim has thought deeply about quantum particles such as anyons to generate testable hypotheses about the strange ways they could behave in certain materials. Her work is motivated by a curiosity about how matter works. But for a while, it remained too difficult to manipulate materials in the laboratory with the level of control she needed. “I had kind of given up because the gap between the real world and where theorists comfortably live seemed too far,” she says.

Google’s quantum computer gave Kim a means to create anyons and test her hypotheses. The nascent quantum computing industry may have to wait a while longer until it gets practical topological qubits to start building with, but projects like Kim’s and Quantinuum’s are helping to deepen physicists’ understanding of non-Abelian anyons.

Source

When Microsoft released Windows 11, it did so with a requirement for computers to have TPM 2.0 support for added security but many pre-Windows 11 computers don’t support this so cannot upgrade to Windows 11.

With the requirement to obtain new hardware, the analyst firm Canalys has estimated that as many as 240 million PCs will be thrown away - even if they otherwise work perfectly fine. This is not a great revelation for Microsoft’s green credentials.

If your computer doesn’t support TPM 2.0, there are a few things you can do to help minimize the landfill issue. The first option is to switch to Linux Mint, Ubuntu, or some other Linux distribution.

Linux works very well out of the box on most computers and it isn’t too difficult to write the ISO image to a USB and install it on your PC, the process should take around 30 minutes. With much of our computing being done online, the lack of Windows software for Linux is not too much of an issue now.

If you absolutely need to use Windows and plan to landfill your old computer then take out anything that can be removed such as hard drives and memory (on laptops). Hard drives can be placed in external hard drive cases available on sites like Amazon, giving you a new external hard drive, and the RAM can be used with other devices to add more memory if they need it. You could also sell these parts.

If you are planning to buy a new PC, have a look at Neowin’s deals page to get big discounts on PCs and other products, this helps to keep Neowin free of charge for readers.

Source: Reuters

Source

Source

The Steam cart update brings three new features. The new private purchasing option is available from the cart itself, letting customers mark the games as private before they even enter the library. This will stop the purchases from appearing in user profiles and even hide them from friends when playing.

Next, buying gifts for friends usually means making a new cart and going through the purchasing process for each and every Steam account. The new Inline gifting feature streamlines this to make purchasing multiple gifts a single action. Gift messaging has been changed to "speed up the checkout process" as well.

Lastly, beta users will find that their Steam cart is now shared between all their devices that they have signed into.

The private games function also extends to already purchased items:

Starting today you’ll be able to mark specific games as private and they’ll disappear from anywhere they’d be viewed by someone other than you. That includes: your ownership, in-game status, playtime, and activity in that game. This additional control allows you to keep most of your Steam Library visible to your friends, so they can see what you are playing and join in, yet also keep a few of those games just to yourself.

Marking a game as Private will do the following:

• The game will no longer appear on your Steam Community profile games list, or in the recently played games section.
• Your status on the Friends List of other users or in chats will not change to "In Game".
• Your activity in the game (like achievement unlocks or "Played for the first time") will not appear in the Activity feed of people on your Steam Friends List.
• When a Steam Friend is browsing the Steam Store, your profile will not appear in the "Friends who own this game" section.
• If a Steam Friend attempts to buy a game as a gift that you have marked private, you will not show as already owning the game.
• Showcases on your Steam Community Profile will not show the game or achievements for the game (such as the "Rarest Achievements" showcase).
• You will not receive Steam Trading Cards for the game while it is marked as Private.
• If you have changed the "Game File Transfer over Local Network" setting to allow other users on the same network to download games, they will not be able to transfer games that have been marked as Private.

Games can be marked as private from the shopping cart, the games list straight from a web browser, or via the Steam client by heading to Game page > Settings > Manage.

To enter the Steam client beta and use the new features, head to the Interface tab in Steam Settings use the Client Beta Participation dropdown. Valve is currently testing an upgraded Steam Workshop too, bringing popular modding features like load orders and dependency displays.

Source

Despite having no idea about its specs, Neowin was quite impressed by what we saw after having a hands-on experience with the headset.

Microsoft had big ambitions with MR, and in a way, it is sort of similar to how the company is now pumped for Copilot on the next-gen Windows, so much so that Microsoft CEO Satya Nadella feels like it's just as immense and significant as the Start button itself. At the time, the HoloLens and its accompanying platform framework was called the "Windows Holographic," though in 2017, this was renamed to "Windows Mixed Reality."

Although it may be a bit difficult to relate to it now, Microsoft was not too wrong in its outlook as VR and AR seemed like the next big thing at that time. It was only natural for the tech giant to get on the train as it did not want to be left behind since almost everyone else was doing it too. It is fair to say though that the industry as a whole, including Microsoft, probably overestimated the potential of MR and as such, the company was not able to grow the business beyond a few niche markets here and there.

Perhaps it could be a reason why Microsoft has now announced that it is deprecating the Windows Mixed Reality platform. On a webpage outlining the deprecated Windows client features, Microsoft has added a section for Windows Mixed Reality, where it notes that it's killing off Windows Mixed Reality alongside the Mixed Reality Portal app as well as Windows MR for SteamVR and Steam VR beta.

It writes:

Windows Mixed Reality is deprecated and will be removed in a future release of Windows. This deprecation includes the Mixed Reality Portal app, and Windows Mixed Reality for SteamVR and Steam VR Beta.

In case you want a quick recap about the Windows Mixed Reality platform, you can read this article Neowin's John Callaham recently did in their "Look back" column.

Source

(TNS) — Pitching New Jersey as the East Coast’s answer to Silicon Valley in a rapidly growing and heavily scrutinized technology, Gov. Phil Murphy and other leaders announced Monday the state will establish a formal hub for artificial intelligence research and development at Princeton University.

The project is a partnership between the state government, the Ivy League school, and private companies to explore and harness the technology, which has become an increasingly bigger part of the world in recent years, sparking both wonder and worry.

Murphy predicted AI technology could define “the next decade or more” across the globe, and now both the state and Princeton are “poised to shape the future of this revolutionary field and unleash a new century of game-changing discovery.”

“There is no better place to be a leader in AI than our great Garden State,” the governor added during at a ceremony at the university announcing the plans. “After all, innovation is in our state’s DNA.”

Officials said the center, in collaboration with the New Jersey Economic Development Authority, will conduct research, examine how to use AI in an ethical manner, and train state employees in the technology. Researchers, industry leaders, and startup companies will all be involved, officials said.

Murphy said it will be “a new home for the world’s boldest and brightest to pioneer breakthroughs for the betterment of humankind” — possibly, he noted, by helping develop cures for diseases and methods to combat climate change, as well as boosting the economy with new businesses and jobs.

Officials also announced Princeton and the state will co-host an AI conference at the school on April 11 to discuss the “most pressing AI issues of the day,” University President Christopher Eisgruber said.

But officials did not provide a specific timeline for the center’s rollout. Not did they say how much taxpayer money will be used for the project.

AI has been both praised for the transformational way it generates text, images, audio, video, and other art on command, as well as criticized for making it easier for students, companies, and news organizations to use tools like ChatGPT to produce automated work. There have been concerns about data security and privacy, as well.

Monday’s announcement comes two months after Murphy signed an executive order in October establishing a state government task force to examine the possible benefits and pitfalls of AI and how it can be used responsibly in New Jersey.

Beth Noveck, the state Office of Innovation Chief Innovation Officer, said AI “offers a powerful set of tools” and noted the state government is already utilizing it to bolster public policy and benefits. That includes using it to give business owners customized advice and gather comments about unemployment.

All of this is fitting, officials said, because New Jersey is home to decades of innovation, such as Thomas Edison’s light bulb developments, while Princeton, one of the world’s most renowned schools, is the alma mater of generations of scientific leaders, such as Alan Turing, considered the father of computers.

“Much of the modern world was born right here in New Jersey,” Murphy said.

Now, the governor noted, the state faces “stiffer competition” in northern California’s Silicon Valley, Europe, and East Asia. But, he said, he would still stake New Jersey against any of them.

“Only here in New Jersey will you find that golden combination of talent, location, and a thriving innovation economy,” Murphy said. “And only here will you find Princeton University.”

Princeton Provost Jennifer Rexford said there is already a “palpable:” excitement about AI the school’s campus, saying they often don’t have rooms big enough to host everyone trying to attend events related to the topic.

“We see it in students who flock en masse to any course related to AI on both the undergraduate and at the graduate level” Rexford said. “And you see it in the dissertation work of our Ph.D. students, in the senior theses of our undergrads who are using or advancing AI in the work that they’re doing.”

Murphy also noted Bell Labs, Nokia’s research arm, announced last week it will will leave its historic headquarters in Murray Hill over the next five years to relocate to a new tech hub being built in New Brunswick. That’s about a 20-minute drive from Princeton.

“Now more than ever, Route 1 will be the corridor of innovation,” Murphy said.

“In other words, buckle up,” he added. “Because with the help of everyone here today, New Jersey is going to be at the forefront of the AI revolution.”

©2023 Advance Local Media LLC. Distributed by Tribune Content Agency, LLC.

Source

Artificial intelligence programs easily and consistently outplay human competitors in cognitively intensive games like chess, poker, and Go—but it’s much harder for robots to beat their biological rivals in games requiring physical dexterity. That performance gap appears to be shortening, however, starting with a classic children’s puzzle game.

Researchers at Switzerland’s ETH Zurich recently unveiled CyberRunner, their new robotic system that leveraged precise physical controls, visual learning, and AI training reinforcement in order to learn how to play Labyrinth faster than a human.

Labyrinth and its many variants generally consist of a box topped with a flat wooden plane that tilts across an x and y axis using external control knobs. Atop the board is a maze featuring numerous gaps. The goal is to move a marble or a metal ball from start to finish without it falling into one of those holes. It can be a… frustrating game, to say the least. But with ample practice and patience, players can generally learn to steady their controls enough to steer their marble through to safety in a relatively short timespan.

CyberRunner, in contrast, reportedly mastered the dexterity required to complete the game in barely 5 hours. Not only that, but researchers claim it can now complete the maze in just under 14.5 seconds—over 6 percent faster than the existing human record.

The key to CyberRunner’s newfound maze expertise is a combination of real-time reinforcement learning and visual input from overhead cameras. Hours’ worth of trial-and-error Labyrinth runs are stored in CyberRunner’s memory, allowing it learn step-by-step how to best navigate the marble successfully along its route.

“Importantly, the robot does not stop playing to learn; the algorithm runs concurrently with the robot playing the game,” reads the project’s description. “As a result, the robot keeps getting better, run after run.”

CyberRunner not only learned the fastest way to beat the game—but it did so by finding faults in the maze design itself. Over the course of testing possible pathways, the AI program uncovered shortcuts allowing it to shave off time from its runs. Basically, CyberRunner created its own Labyrinth cheat codes by finding shortcuts that sidestep the maze’s marked pathways.

CyberRunner’s designers have made the project completely open-source, with an aim for other researchers around the world to utilize and improve upon the program’s capabilities.

“Prior to CyberRunner, only organizations with large budgets and custom-made experimental infrastructure could perform research in this area,” project collaborator and ETH Zurich professor Raffaello D’Andrea said in a statement this week. “Now, for less than 200 dollars, anyone can engage in cutting-edge AI research. Furthermore, once thousands of CyberRunners are out in the real-world, it will be possible to engage in large-scale experiments, where learning happens in parallel, on a global scale.”

Source

It’s really, really hard to kill a large, beloved social network. But Elon Musk has seemingly been giving it his absolute best shot: Over the past year, Twitter has gotten a new name (X), laid off much of its staff, struggled with outages, brought back banned accounts belonging to Alex Jones and Donald Trump, and lost billions in advertising revenue.

Opportunistic competitors have launched their own Twitter clones, such as Bluesky, Mastodon, and Threads. The hope is to capture fleeing users who want “microblogging”—places where people can shoot off little text posts about what they ate for lunch, their random thoughts about politics or pop culture, or perhaps a few words or sentences of harassment Threads, Meta’s entry which launched in July, seems the most promising, at least in terms of pure scale. Over the summer, it broke the record for fastest app to reach 100 million monthly active users—beating a milestone set by ChatGPT just months earlier—in part because Instagram users were pushed toward it. (Turns out, it’s pretty helpful to launch a new social network on the back of the defining social-media empire of our time.)

But the decline of Twitter, and the race to replace it, is in a sense a sideshow. Analytics experts shared data with me suggesting that the practice of microblogging, while never quite dominant, is only becoming more niche. In the era of TikTok, the act of posting your two cents in two sentences for strangers to consume is starting to feel more and more unnatural. The lasting social-media imprint of 2023 may not be the self-immolation of Twitter but rather that short-form videos—on TikTok, Instagram, and other platforms—have tightened their choke hold on the internet. Text posts as we’ve always known them just can’t keep up.

Social-media companies only tend to sporadically share data about their platforms, and of all the main microblogging sites —X, Threads, Bluesky, and Mastodon—just Bluesky provided a comment for this story. “We’ve grown to 2.6 million users on an invite-only basis in 2023,” BlueSky’s CEO, Jay Gruber, wrote in an email, “and are excited about growth while we open up the network more broadly next year.” So I reached out to outside companies that track social analytics. They told me that these new X competitors haven’t meaningfully chipped away at the site’s dominance. For all of the drama of the past year, X is by far still the predominant network for doing brief text posts. It is still home to more than four times as many monthly active users as Threads, Bluesky, and Mastodon combined, according to numbers shared with me by data.ai, a company that tracks app-store activity. (Data.ai looks only at mobile analytics, so it can’t account for desktop users.)

Mastodon and Bluesky amounted to just “rounding errors, in terms of the number of people engaging,” says Paul Quigley, the CEO of NewsWhip, a social-media-monitoring platform. Threads has not fared much better. Sensor Tower, another analytics firm, estimates that fewer than 1 percent of Threads users opened the app daily last month, compared with 18 percent of Twitter users. And even those who open the app are spending an average of just three minutes a day on it.

That doesn’t mean X is thriving. According to data.ai’s 2024-predictions report, the platform’s daily active users peaked in July 2022, at 316 million, and then dropped under Musk. Based on its data-science algorithms, data.ai predicts that X usership will decline to 250 million in 2024. And data.ai expects microblogging overall to decline alongside X next year, even though these new platforms seem positioned for growth: Threads, after all, just recently launched in Europe and became available as a desktop app, and to join Bluesky, you still need an invite code.

Of course, these are just predictions. Plenty of people do still want platforms for sending off quick thoughts, and perhaps X or any other alternative will gain more users. But the decline of microblogging is part of a larger change in how we consume media. On TikTok and other platforms, short clips are served up by an at-times-magical-seeming algorithm that makes note of our every interest. Text posts don’t have the same appeal. “While platforms like X are likely to maintain a core niche of users, the overall trends show consumers are swapping out text-based social networking apps for photo and video-first platforms,” data.ai noted in their predictions report.

Short-form videos have become an attention vortex. Users are spending an average of 95 minutes a day on TikTok and 61 minutes on Instagram as of this quarter, according to estimates from Sensor Tower. By comparison, they’re estimated to average just 30 minutes on Twitter and three minutes on Threads. People also want companies to shift to video along with them in what is perhaps this the real pivot to video: In a recent survey by Sprout Social, a social-media-analytics tool, 41 percent of consumers said that they want brands to publish more 15- to 30-second videos more than they want any other style of social-media post. Just 10 percent wanted more text-only content.

Maybe this really is the end for the short text post, at least en masse. Or maybe our conception of “microblogging” is due for an update. TikTok videos are perhaps “just a video version of what the original microblogs were doing when they first started coming out in the mid-2000s,” André Brock, a media professor at Georgia Tech who has studied Twitter, told me; they can feel as intimate and authentic as a tweet about having tacos for lunch. Trends such as “men are constantly thinking about the Roman empire” (and the ensuing pushback) could have easily been a viral Twitter or Facebook conversation in a different year. For a while, all of the good Twitter jokes were screenshotted and re-uploaded to Instagram. Now it can feel like all of the good TikToks are downloaded and reposted on Instagram. If the Dress (white-and-gold or black-and-blue?) were to go viral today, it would probably happen in a 30-second video with a narrator and a soundtrack.

But something is left behind when microblogging becomes video. Twitter became an invaluable resource during news moments—part of why journalists flocked to the platform, for better or for worse—allowing people to refresh and instantaneously get real-time updates on election results, or a sports game, or a natural disaster. Movements such as Occupy Wall Street and Black Lives Matter turned to Twitter to organize protests and spread their respective messages.

Some of the news and political content may just as easily move to TikTok: Russia’s war with Ukraine has been widely labeled the “first TikTok War,” as many experienced it for the first time through that lens. Roughly a third of adults under 30 now regularly get their news from TikTok, according to Pew Research. But we don’t yet totally know what it means to have short-form videos, delivered via an algorithmic feed, be the centerpiece of social media. You might log onto TikTok and be shown a video that was posted two weeks ago.

Perhaps the biggest stress test for our short-form-video world has yet to come: the 2024 U.S. presidential election. Elections are where Twitter, and microblogging, have thrived. Meanwhile, in 2020, TikTok was much smaller than what it is now. Starting next year, its true reign might finally begin.

Source

The team, led by Mikhail Lukin, the Joshua and Beth Friedman University Professor in physics and co-director of the Harvard Quantum Initiative, has created the first programmable, logical quantum processor, capable of encoding up to 48 logical qubits, and executing hundreds of logical gate operations, a vast improvement over prior efforts.

Published in Nature, the work was performed in collaboration with Markus Greiner, the George Vasmer Leverett Professor of Physics; colleagues from MIT; and QuEra Computing, a Boston company founded on technology from Harvard labs.

The system is the first demonstration of large-scale algorithm execution on an error-corrected quantum computer, heralding the advent of early fault-tolerant, or reliably uninterrupted, quantum computation.

Lukin described the achievement as a possible inflection point akin to the early days in the field of artificial intelligence: the ideas of quantum error correction and fault tolerance, long theorized, are starting to bear fruit.

"I think this is one of the moments in which it is clear that something very special is coming," Lukin said. "Although there are still challenges ahead, we expect that this new advance will greatly accelerate the progress toward large-scale, useful quantum computers."

Denise Caldwell of the National Science Foundation agrees.

"This breakthrough is a tour de force of quantum engineering and design," said Caldwell, acting assistant director of the Mathematical and Physical Sciences Directorate, which supported the research through NSF's Physics Frontiers Centers and Quantum Leap Challenge Institutes programs.

"The team has not only accelerated the development of quantum information processing by using neutral atoms, but opened a new door to explorations of large-scale logical qubit devices, which could enable transformative benefits for science and society as a whole."

It's been a long, complex path.

In quantum computing, a quantum bit or "qubit" is one unit of information, just like a binary bit in classical computing. For more than two decades, physicists and engineers have shown the world that quantum computing is, in principle, possible by manipulating quantum particles—be they atoms, ions, or photons—to create physical qubits.

But successfully exploiting the weirdness of quantum mechanics for computation is more complicated than simply amassing a large-enough number of qubits, which are inherently unstable and prone to collapse out of their quantum states.

The real coins of the realm are so-called logical qubits: bundles of redundant, error-corrected physical qubits, which can store information for use in a quantum algorithm. Creating logical qubits as controllable units—like classical bits—has been a fundamental obstacle for the field, and it's generally accepted that until quantum computers can run reliably on logical qubits, the technology can't really take off.

To date, the best computing systems have demonstrated one or two logical qubits, and one quantum gate operation—akin to just one unit of code—between them.

The Harvard team's breakthrough builds on several years of work on a quantum computing architecture known as a neutral atom array, pioneered in Lukin's lab. It is now being commercialized by QuEra, which recently entered into a licensing agreement with Harvard's Office of Technology Development for a patent portfolio based on innovations developed by Lukin's group.

The key component of the system is a block of ultra-cold, suspended rubidium atoms, in which the atoms—the system's physical qubits—can move about and be connected into pairs—or "entangled"—mid-computation.

Entangled pairs of atoms form gates, which are units of computing power. Previously, the team had demonstrated low error rates in their entangling operations, proving the reliability of their neutral atom array system.

With their logical quantum processor, the researchers now demonstrate parallel, multiplexed control of an entire patch of logical qubits, using lasers. This result is more efficient and scalable than having to control individual physical qubits.

"We are trying to mark a transition in the field, toward starting to test algorithms with error-corrected qubits instead of physical ones, and enabling a path toward larger devices," said paper first author Dolev Bluvstein, a Griffin School of Arts and Sciences Ph.D. student in Lukin's lab.

The team will continue to work toward demonstrating more types of operations on their 48 logical qubits and to configure their system to run continuously, as opposed to manual cycling as it does now.

Source

Source

What you need to know

• Bethesda Game Studios revealed that Starfield has crossed 13 million players since its launch earlier this year. 
• The team reiterated features that are coming in future updates, such as City Maps, as well as providing more detail on gameplay difficulty tweaks that'll be added. 
• Bethesda Game Studios also confirmed that Shattered Space, the first expansion for Starfield, is set to launch in 2024.

Bethesda Game Studios' science-fiction role-playing game reached a new milestone in time for the holidays. 

The team shared on Wednesday that Starfield crossed 13 million players, a new high after the title saw 10 million players a few weeks after its launch at the start of September earlier in the year. Bethesda Game Studios notes that the average player has put 40 hours into Starfield.

Bethesda Game Studio also reiterated how it's looking to launch updates to the game around every six weeks in 2024, with features on the way like an City Maps, new ways of traveling, and official mod support via a new Creation Kit. The team revealed players will be able to customize their gameplay experience further in the future, adjusting "carry capacity, cargo access distance, ship damage, vendor credits, how you suffer afflictions, new survival mechanics," and more. 

When does Starfield get its first DLC?

While we've known for some time that there would be DLC content released for Starfield beyond the free updates, Bethesda also shared on Wednesday that Shattered Space, the game's first expansion, is meant to launch in 2024. This paid content will include "new story content, new locations, new gear, and much more," while being included at no extra cost for anyone who has bought the Starfield Digital Deluxe Edition. 

Analysis: Success in the stars

These numbers are good to see, and I'm sure it's nice to know your game is popular as everyone winds down for the holidays. I am curious how the average of 40 hours playtime per player is calculated, as I know some people have likely put in hundreds, if not thousands of hours across New Game+ and beyond. 

The cadence of updates being promised for next year is extremely exciting and meaningful for players on Xbox and Windows PC alike, if the team can stick to it. I'm especially excited by Shattered Space launching next and seeing just where this first expansion will explore. Hopefully we also get confirmation of DLC beyond that first paid outing.

Source

The Verge got a hold of an internal Microsoft memo that was sent out today from Microsoft Gaming CEO Phil Spencer that revealed the departure date for Kotick. Spencer added:

Under Bobby’s watch, Activision Blizzard in its many incarnations has been an enduring pillar of video games. Whether it’s Call of Duty, World of Warcraft, Candy Crush Saga or any number of other titles, his teams have created beloved franchises and entertained hundreds of millions of players for decades. I’d like to thank Bobby—for his invaluable contributions to this industry, his partnership in closing the Activision Blizzard acquisition and his collaboration following the close—and I wish him and his family the very best in his next chapter.

Kotick also posted his own farewell message on Activision's blog, stating in part:

Forty years later, as my last day leading this company inches closer, I marvel at how far the talented people at our company have come toward realizing the great potential of games. You have transformed a hobbyist form of entertainment into the world’s most engaging medium. It has been the privilege of my lifetime to work alongside you as we broadened the appeal of games.

Spencer stated in the memo that the leadership team for Activision Publishing, Blizzard, and King will remain and there will be "no changes to the structure of how the studios and business units are run." The leadership for those three aspects of Activision Blizzard will report to Matt Booty, the recently named President of Game Content and Studios at Microsoft.

In a separate memo from Booty, he mentions that Jill Braff is the new head of ZeniMax/Bethesda Studios, and they will continue to operate as "limited integration entities" with Microsoft.

A number of other Activision Blizzard executives will be departing in March 2024, including Thomas Tippl, the Vice Chairman of Activision Blizzard. Others will be leaving in January and December 2023, including Humam Sakhnini, the Vice Chairman of Blizzard and King. Spencer stated:

We have an exciting 2024 lineup of games across Activision, Bethesda, Blizzard, King and Xbox Game Studios, and I know that we all look forward to sharing more details with our player communities when the time is right.

Hopefully one of those announcements will be adding older Activision Blizzard games to Xbox Game Pass.

Source

Source

Apple started adding satellite features to its phones starting with the iPhone 14, allowing people to get help wherever they are outside of traditional cell network coverage. The newer satellite-based roadside assistance feature is just an extension of that, letting you call a tow truck or other resource if something happens to your car during a road trip. Now, Apple is adding Verizon compatibility to the mix.

Currently, Apple's satellite roadside assistance services are provided by AAA, but now, with iOS 17.2, you can also get assistance using Verizon's roadside assistance service. The service has been around for some time, but it has needed cell service to use it. Now it can also use Apple's satellite services, so you can get assistance anywhere within the United States, regardless of whether you happen to have a cell signal or not. Verizon's roadside assistance is available to all Verizon subscribers and it's an additional $4.99 charge on top of your bill, with additional charges varying depending on what you need exactly—towing, for one, starts at $154.

Apple's services are provided by AAA, and likewise, any help resulting from the satellite assistance request needs to be paid to AAA rather than Apple. In practice, this is just an alternative provider of services that you can use instead of AAA for your roadside assistance. It's also one that a lot of you might already have—Verizon has offered roadside assistance and vehicle services for years, and the only truly new thing here is the fact that you can use it now even if you're off the grid. The subscription comes with winching, towing, battery jumpstarts, tire changes, and fuel assistance.

Like other Apple satellite services, this will need at least an iPhone 14 in addition to a Verizon SIM, so if you have an older iPhone, you'll need to upgrade if you want to take advantage of this.

It's certainly good to see that Apple is opening up its Emergency SOS features to more service providers, and hopefully, features like this might come to Android one day. We'll have to wait for that, though.

Source: MacRumors

Source

Bill Gates is still a dominant figure in technology, nearly a half-century after co-founding Microsoft. And when a foundational industry figure speaks--albeit one who shared ties with the now-deceased sex criminal Jeffrey Epstein--it's worth paying attention. A new blog post summarizes Gates's thoughts on "the road ahead" in 2024, an influential commentary with insights on artificial intelligence and innovation.

What makes Gates's thoughts important for tech entrepreneurs is that these posts provide peeks into the mind of someone with a finger on the pulse of bleeding-edge technology who can also still influence the entire industry.

A.I.

Gates thinks that the  A.I. explosion of 2023 is merely the beginning of a tech trend that will shape the near future to an even greater extent. The acronym appears 28 times across the six-page post.

He predicts A.I. will move from its current somewhat geeky corner niche into the limelight.

"In high-income countries like the United States, I would guess that we are 18-24 months away from significant levels of A.I. use by the general population," he writes.

Gates expects A.I. use in less-developed nations to soar, too, with a shorter lag time than had been seen for many other tech innovations. Gates admits that "2023 marked the first time I used artificial intelligence for work and other serious reasons, not just to mess around and create parody song lyrics for my friends." 

For business leaders directly involved in developing A.I. solutions, this will sound familiar, even if they haven't played at being Weird Al Yankovic with a chatbot. Small and midsize businesses are still exploring the many ways A.I. applications might help their companies. If a business isn't already considering incorporating A.I. into customer-facing uses (perhaps as a chatbot) or business decision-making, Gates' enthusiasm may be a nudge to embrace this tech.

Innovation

Though "innovation" runs as a theme through Gates's note, he also ties it to A.I. That's because artificial intelligence is set to "supercharge the innovation pipeline." Citing advances in electric power, cars, planes and the digital realm, Gates says "innovation is the reason our lives have improved so much over the last century." He also blows his own trumpet a little, noting "we are far more productive because of the IT revolution."

Gates, who retired as Microsoft CEO in 2000 to focus on his work with the Gates Foundation, suggests the next step in innovation will be the acceleration of development projects that actually impact people's lives materially, financially and socially. 

The Gates Foundation focuses its work on improving health and reducing poverty globally, so the examples used in Gates's note are tailored toward these topics. The drug discovery process already uses A.I. applications to sift through massive amounts of data almost instantly, slashing new medicines' time to market. Gates muses that A.I. could help treat high-risk pregnancies. He also suggests that A.I. could bring personalized tutors to every learner: "The A.I. education tools being piloted today are mind-blowing because they are tailored to each individual." He also thinks A.I. could make it easy for every health worker to access medical info.

Some of these lessons resonate in the business world too. A.I. can already help companies learn more about their customers, reinforce their supply chains and help process raw data to understand business dynamics better. It can even help founders put together a business plan to start a new enterprise. And its use in content creation for business cases is already well known.

Essentially, Gates's note is a primer for how entrepreneurs should definitely be thinking about how A.I. can help businesses innovate, particularly if it can "get game-changing technologies out to the people who need them faster than ever before."

It's Not All Easy

All of this sounds exciting, and might be a shot in the arm for A.I. startups or leaders planning to use A.I. to boost their business. 

But Gates says figuring out A.I. isn't easy. The world is just starting to get an idea about how A.I. can replace some jobs, and act as a "copilot" for others. And though the tech is exciting, he concedes it's also "confusing." He even admits that though he thought he'd use A.I. to help him sift through "hundreds of pages of briefing materials" to help craft strategy reviews for the Gates Foundation, expecting it could "accurately summarize" the data for him, he ended up doing it all himself.

Source

What you need to know

• Former Microsoft CEO, Bill Gates says 2024 will be the year of AI.
• He says the technology is practically taking over the world, and that it's going to get more intensive as the years go by and more people embrace AI.
• The Gates Foundation is primarily focused on using the tech to address health issues affecting poor countries like AIDS and TB.

There's no doubt that this year has been huge for generative AI, given the numerous inventions and advances that have been made. AI-powered tools like ChatGPT, Microsoft's Copilot (formerly Bing Chat), Bing Image Creator, and more have completely revolutionized how I work (but that's a story for another day).

I believe the same sentiments are reiterated by large masses of people, including Microsoft's former CEO, Bill Gates. While looking back at the year that was, as we forge towards 2024 in the next few days, Gates penned a letter highlighting how AI will shape the future. 

Gates looks into the future through his young granddaughter's eyes. This is because this is essentially the generation that will inherit a "full-fledged" AI-powered world. While many might argue that the technology is still new, we've already seen the input it has on various industries, including computing, medicine, education, and more.

Admittedly, there's a cloud of concern looming over most users, predominantly because of the lack of guardrails alongside privacy issues revolving around the emerging technology. The Biden-Harris administration has already addressed some of these issues through an Executive Order, which has already started taking shape. It recently blocked an NVIDIA shipment headed to China over safety concerns. The US government quickly issued a statement citing that the move was designed to establish control over AI chips, and not to run down the country's economy. However, the country has since turned its eyes to GPU recycling factories for its AI chip needs. 

Bill Gates compares the emergence of AI to when the internet had just become available. He looks back at the trajectory the technology took from the onset, from not knowing how many people have access to everyone having an email address and using the tool to scour the web. 

Bill Gates: The future looks bright for AI

Gates further points out his investment in innovation while working at Microsoft and even now at the Gates Foundation alongside his wife, Melinda. He categorically stated that it's the main reason why a lot of progress has been over the years across all spheres of our lives.

He added that innovation has contributed to the emergence of electricity, cars, planes, and significant advances in medicine over the past few years, further highlighting his favorite innovation story, where the number of children who die before the age of five has reduced by half since 2000. He attributed this success to innovation, as it allowed scientists to come up with efficient and effective vaccines that are safe and affordable for all. 

Bill Gates through the Gates Foundation intends to explore the technology more and optimize its capabilities in a bid to create new medicines by combining large amounts of data and AI tools to hasten the process. 

According to Gates:

"Some companies are already working on cancer drugs developed this way. But a key priority of the Gates Foundation in AI is ensuring these tools also address health issues that disproportionately affect the world's poorest, like AIDS, TB, and malaria."

To this end, the Gate Foundation is exploring multiple AI innovations alongside its partners to leverage these technologies to enhance living conditions across low-income areas.

What are your thoughts on AI? Do you think the guardrails in place are enough to prevent it from spiraling out of control? Let us know in the comments.

Source

Using a type of artificial intelligence known as deep learning, MIT researchers have discovered a class of compounds that can kill a drug-resistant bacterium that causes more than 10,000 deaths in the United States every year.

In a study appearing today in Nature, the researchers showed that these compounds could kill methicillin-resistant Staphylococcus aureus (MRSA) grown in a lab dish and in two mouse models of MRSA infection. The compounds also show very low toxicity against human cells, making them particularly good drug candidates.

A key innovation of the new study is that the researchers were also able to figure out what kinds of information the deep-learning model was using to make its antibiotic potency predictions. This knowledge could help researchers to design additional drugs that might work even better than the ones identified by the model.

“The insight here was that we could see what was being learned by the models to make their predictions that certain molecules would make for good antibiotics. Our work provides a framework that is time-efficient, resource-efficient, and mechanistically insightful, from a chemical-structure standpoint, in ways that we haven’t had to date,” says James Collins, the Termeer Professor of Medical Engineering and Science in MIT’s Institute for Medical Engineering and Science (IMES) and Department of Biological Engineering.

Felix Wong, a postdoc at IMES and the Broad Institute of MIT and Harvard, and Erica Zheng, a former Harvard Medical School graduate student who was advised by Collins, are the lead authors of the study, which is part of the Antibiotics-AI Project at MIT. The mission of this project, led by Collins, is to discover new classes of antibiotics against seven types of deadly bacteria, over seven years.

Explainable predictions

MRSA, which infects more than 80,000 people in the United States every year, often causes skin infections or pneumonia. Severe cases can lead to sepsis, a potentially fatal bloodstream infection.

Over the past several years, Collins and his colleagues in MIT’s Abdul Latif Jameel Clinic for Machine Learning in Health (Jameel Clinic) have begun using deep learning to try to find new antibiotics. Their work has yielded potential drugs against Acinetobacter baumannii, a bacterium that is often found in hospitals, and many other drug-resistant bacteria.

These compounds were identified using deep learning models that can learn to identify chemical structures that are associated with antimicrobial activity. These models then sift through millions of other compounds, generating predictions of which ones may have strong antimicrobial activity.

These types of searches have proven fruitful, but one limitation to this approach is that the models are “black boxes,” meaning that there is no way of knowing what features the model based its predictions on. If scientists knew how the models were making their predictions, it could be easier for them to identify or design additional antibiotics.

“What we set out to do in this study was to open the black box,” Wong says. “These models consist of very large numbers of calculations that mimic neural connections, and no one really knows what's going on underneath the hood.”

First, the researchers trained a deep learning model using substantially expanded datasets. They generated this training data by testing about 39,000 compounds for antibiotic activity against MRSA, and then fed this data, plus information on the chemical structures of the compounds, into the model.

“You can represent basically any molecule as a chemical structure, and also you tell the model if that chemical structure is antibacterial or not,” Wong says. “The model is trained on many examples like this. If you then give it any new molecule, a new arrangement of atoms and bonds, it can tell you a probability that that compound is predicted to be antibacterial.”

To figure out how the model was making its predictions, the researchers adapted an algorithm known as Monte Carlo tree search, which has been used to help make other deep learning models, such as AlphaGo, more explainable. This search algorithm allows the model to generate not only an estimate of each molecule’s antimicrobial activity, but also a prediction for which substructures of the molecule likely account for that activity.

Potent activity

To further narrow down the pool of candidate drugs, the researchers trained three additional deep learning models to predict whether the compounds were toxic to three different types of human cells. By combining this information with the predictions of antimicrobial activity, the researchers discovered compounds that could kill microbes while having minimal adverse effects on the human body.

Using this collection of models, the researchers screened about 12 million compounds, all of which are commercially available. From this collection, the models identified compounds from five different classes, based on chemical substructures within the molecules, that were predicted to be active against MRSA.

The researchers purchased about 280 compounds and tested them against MRSA grown in a lab dish, allowing them to identify two, from the same class, that appeared to be very promising antibiotic candidates. In tests in two mouse models, one of MRSA skin infection and one of MRSA systemic infection, each of those compounds reduced the MRSA population by a factor of 10.

Experiments revealed that the compounds appear to kill bacteria by disrupting their ability to maintain an electrochemical gradient across their cell membranes. This gradient is needed for many critical cell functions, including the ability to produce ATP (molecules that cells use to store energy). An antibiotic candidate that Collins’ lab discovered in 2020, halicin, appears to work by a similar mechanism but is specific to Gram-negative bacteria (bacteria with thin cell walls). MRSA is a Gram-positive bacterium, with thicker cell walls.

“We have pretty strong evidence that this new structural class is active against Gram-positive pathogens by selectively dissipating the proton motive force in bacteria,” Wong says. “The molecules are attacking bacterial cell membranes selectively, in a way that does not incur substantial damage in human cell membranes. Our substantially augmented deep learning approach allowed us to predict this new structural class of antibiotics and enabled the finding that it is not toxic against human cells.”

The researchers have shared their findings with Phare Bio, a nonprofit started by Collins and others as part of the Antibiotics-AI Project. The nonprofit now plans to do more detailed analysis of the chemical properties and potential clinical use of these compounds. Meanwhile, Collins’ lab is working on designing additional drug candidates based on the findings of the new study, as well as using the models to seek compounds that can kill other types of bacteria.

“We are already leveraging similar approaches based on chemical substructures to design compounds de novo, and of course, we can readily adopt this approach out of the box to discover new classes of antibiotics against different pathogens,” Wong says.

In addition to MIT, Harvard, and the Broad Institute, the paper’s contributing institutions are Integrated Biosciences, Inc., the Wyss Institute for Biologically Inspired Engineering, and the Leibniz Institute of Polymer Research in Dresden, Germany. The research was funded by the James S. McDonnell Foundation, the U.S. National Institute of Allergy and Infectious Diseases, the Swiss National Science Foundation, the Banting Fellowships Program, the Volkswagen Foundation, the Defense Threat Reduction Agency, the U.S. National Institutes of Health, and the Broad Institute. The Antibiotics-AI Project is funded by the Audacious Project, Flu Lab, the Sea Grape Foundation, the Wyss Foundation, and an anonymous donor.

Source

The entertainment conglomerate is looking for a strong performance from the PS5, which is in its fourth year on the market, during the holiday shopping season to hit its sales target of a record 25 million units in the current financial year ending March 31.

"Given the momentum we've had in November and a lot of what we're seeing in December, just in general we're feeling very good about sales overall," said Eric Lempel, senior vice president for global marketing, sales and business operations at Sony Interactive Entertainment.

Sony's games business was its largest unit by sales and the second-biggest contributor to operating profit after music in the last financial year.

Management comments in August that PS5 sales had been weaker than expected led to jitters among investors. Sony has said its 25 million unit goal is a "high target and not within easy reach", with the company looking to strike a good balance between sales volumes and profitability.

Reuters Graphics

"We've done some good promotions this year. I will say we've done fewer promotions at this stage of the lifecycle than we ever have in the history of the company," Lempel said in an interview.

PS5 sales were hampered by supply chain snarls earlier in its lifecycle. Those have eased, with it now gaining tailwinds including the Oct. 20 launch of critically acclaimed game "Marvel's Spider-Man 2".

The Japanese company, which is known for single-player games, is making a push into live-service titles, which offer continuous online play.

Reuters Graphics

Sony's Naughty Dog last week said it has cancelled an online game from its "The Last of Us" franchise, saying releasing it would have hit development of single-player titles.

The company's sales push comes as the industry debates the future for consoles as advances in cloud technology promise gaming untethered from bulky hardware.

For now, consoles continue to draw gamers, with Nintendo's (7974.T) ageing Switch getting a boost this year from the launch of games such as its latest "Zelda" title.

Upcoming PlayStation titles include "The Last of Us Part II Remastered", which launches in January, and limited-time exclusive "Final Fantasy VII Rebirth" in February.

"We are still pushing really hard and I think we will have a record-breaking year no matter where we end up," Lempel said.

Source

Stephen Thaler wanted to be granted two patents in the UK for inventions he says were devised by his "creativity machine" called DABUS.

His attempt to register the patents was refused by the UK's Intellectual Property Office (IPO) on the grounds that the inventor must be a human or a company, rather than a machine.

Thaler appealed to the UK's Supreme Court, which on Wednesday unanimously rejected his appeal as under UK patent law "an inventor must be a natural person".

Judge David Kitchin said in the court's written ruling that the case was "not concerned with the broader question whether technical advances generated by machines acting autonomously and powered by AI should be patentable".

Thaler's lawyers said in a statement that the ruling "establishes that UK patent law is currently wholly unsuitable for protecting inventions generated autonomously by AI machines and as a consequence wholly inadequate in supporting any industry that relies on AI in the development of new technologies".

'LEGITIMATE QUESTIONS'

A spokesperson for the IPO welcomed the decision "and the clarification it gives as to the law as it stands in relation to the patenting of creations of artificial intelligence machines".

They added that there are "legitimate questions as to how the patent system and indeed intellectual property more broadly should handle such creations" and the government will keep this area of law under review.

Thaler earlier this year lost a similar bid in the United States, where the Supreme Court declined to hear a challenge to the U.S. Patent and Trademark Office's refusal to issue patents for inventions created by his AI system.

Giles Parsons, a partner at law firm Browne Jacobson, who was not involved in the case, said the UK Supreme Court's ruling was unsurprising.

"This decision will not, at the moment, have a significant effect on the patent system," he said. "That's because, for the time being, AI is a tool, not an agent.

"I do expect that will change in the medium term, but we can deal with that problem as it arises."

Rajvinder Jagdev, an intellectual property partner at Powell Gilbert, said the ruling followed similar decisions by courts in Europe, Australia and the U.S. and has "given certainty that inventors must be a natural person."

But he added: "The judgment does not preclude a person using an AI to devise an invention – in such a scenario, it would be possible to apply for a patent provided that person is identified as the inventor."

In a separate case last month, London's High Court ruled that artificial neural networks can attract patent protection under UK law.

Source

When Elon Musk first took over Twitter, those of us in the tech media had all kinds of theories about how the acquisition might bring about the death of the 17-year-old platform.

Some posited that his inept attempts at cost-cutting would cause irreparable damage to Twitter’s infrastructure or that mass resignations would lead to catastrophic instability. But as is so often the case with Musk, predictions were in vain. Twitter did die this year, but the way it played out was both more boring and more stupid than anyone could have possibly imagined.

Musk killed Twitter by slowly making it useless for those who relied on it for real-time information, by choking off conversations from those not willing to pay, by flooding users’ timelines with spammy blue-check sycophants and renaming the company X. He killed it by re-platforming actual Nazis and far-right trolls and Alex Jones and boosting anti-semitism so loudly the site's largest remaining advertisers and most prominent users abandoned the platform in droves. Though you can still go to www.twitter.com and see a website that vaguely resembles the thing we used to call Twitter, it’s only a dull echo of what it once was.

- via Getty Images

The beginning of the end

While you could argue the death spiral began the second Musk walked into Twitter HQ carrying a sink 14 months ago, the platform we all knew began to die three months later, when Musk abruptly decided to ban third-party client apps from its platform and put the rest of its API behind an outrageously expensive paywall.

Twitter had long been an outlier among its social media peers for having a relatively open platform. It gave researchers tools to access the full history of all public conversations on Twitter. It allowed developers to build their own apps on top of its platform, which fostered a small but robust ecosystem of third-party Twitter clients.

Third-party apps like Tweetbot and Twitterific had a relatively small (but devoted) following, but they also played a significant role in defining the culture of Twitter. In the early days of Twitter, the company didn’t have its own mobile app, so it was third-party developers that set the standard of how the service should look and feel. Third-party apps were often the first to adopt now-expected features like in-line photos and video, and the pull-to-refresh gesture. The apps are also responsible for popularizing the word “tweet” and Twitter’s bird logo.

And while many of these apps had become less prominent in recent years, they were emblematic of the way that Twitter, at its best, empowered its users to shape the platform.

Likewise, having an open and readily-available API meant that Twitter, while not the largest social platform, could play an outsize role in shaping online culture. Because its firehose of data was easily accessible to researchers, the public conversations that happened there fueled studies into everything from global elections to public health.

By closing its API to developers and the research community, Musk made it clear he was not interested in using Twitter for anything that couldn’t make him a buck in the process. Twitter’s data was simply another part of the platform to commodify. Nearly a year later, making Twitter’s API inaccessible to all but those with the deepest pockets may not seem like even the tenth-most consequential change to happen under Musk, but it showed just how willing he was to alienate influential communities on Twitter. It was also a major warning sign of what was to come.

The blue check fiasco

If killing Twitter’s API was a quiet warning sign, the complete destruction of Twitter “verification” was a five-alarm fire. Twitter’s verification system was always flawed, but it hinged on the basic premise that the company had some evidence the accounts it verified belonged to the actual people claiming them and that those were people or organizations of some importance. When Musk rolled out his poorly thought out paid verification scheme last year, it went horribly and predictably wrong almost immediately because he failed to uphold any kind of identity check.

Despite the chaotic initial rollout, verification's now-meaningless status did not become fully apparent until this year. After a wave of thousands of spammers, scammers and Musk sycophants signed up for verification, Twitter began removing “legacy” verification from thousands of accounts.

The algorithmic boost provided to the new paid-for wave of blue checks, combined with the promise of a potential share of ad revenue, has drastically altered the dynamics of conversation on Twitter. Verified accounts are given priority ranking in replies and search results, regardless of the size of their following or their engagement — which has made Twitter even less relevant and useful. And the promise of potential ad revenue has incentivized the worst kind of engagement bait.

The result is that even the most carefully curated timelines have become filled with useless spam. And fraudsters are increasingly using pay-to-play verification to carry out scams targeting people trying to reach legitimate customer service channels.

X marks… the death of Twitter
If you were to look for a singular moment when Twitter died, however, it happened in July, when Musk announced that the company would now be known as X. The company changed its name, logo and everything formerly associated with the bird app.

This was more than an ill-considered rebrand. X, a letter with which Musk has long been fascinated, represented, literally, the end of Twitter. For as much as Musk has said it’s about creating an “everything app,” it’s also about fully severing any ties to the expectations and norms associated with Twitter. Want to break verification? Want to charge new users for the privilege of posting? Want to make news stories unreadable? Want to maliciously slow down links to competitors’ websites? Want to re-platform the most heinous peddlers of hate and conspiracy theories? Those actions may have been at odds with Twitter’s mission, but at X, it’s all just another Tuesday. As CEO Linda Yaccarino told CNBC “the rebrand represented really a liberation from Twitter.”

Justin Sullivan via Getty Images

It’s unclear if Musk will ever succeed at creating anything resembling an “everything app” where users will be able to use X to run their “entire financial world.” So far, users seem to have little interest in the somewhat random assortment of new features that have been introduced, like live shopping and aggregating job listings. What Musk has succeeded at, however, is reshaping the platform in his own image.

But if there was any doubt remaining about whether the platform had a chance, Musk has almost single handedly wiped out what remained of Twitter’s ad business. After boosting an antisemitic conspiracy theory and repeatedly failing to prevent ads from appearing near pro-Nazi content, many of the company’s largest remaining advertisers have halted their spending on the platform.

Musk, naturally, responded by telling advertisers “go fuck yourself,” while speculating that the loss of ad dollars could “kill the company.”

But it’s not just advertisers who have fled an increasingly toxic platform. Many of the biggest and most-followed accounts have stopped posting in recent weeks. X’s infrastructure continues to slowly crumble, with random features constantly breaking. Meanwhile, all this has only strengthened the growing number of X competitors, and especially the Meta-owned Threads app. Threads is surging, landing at number four on Apple’s list of most-downloaded apps of the year, despite a late summer launch. X, which has seen steady declines in traffic and engagement, did not make the list.

Source

Søren Freiesleben has lived in Odense his entire life. He likes the historic Danish city for its size. It’s not too big—just 200,000 people live there—and he never feels like he’s drowning in crowds. So far so normal. But there is something unusual about Odense: Its homes are heated by the social giant Meta.

Since 2020, Meta’s hyperscale data center—spanning 50,000 square meters on an industrial estate on the edge of the city—has been pushing warm air generated by its servers into the district heating network under Odense. That heat is then dispersed through 100,000 households hooked up to the system, with Meta providing enough heat to cover roughly 11,000.

Freiesleben, a local councilor, has a few gripes about sharing such a small city with one of the world’s biggest tech companies. Meta should pay more taxes in Denmark, he argues. Locals have also complained about the bright lights positioned around the data center’s perimeter, he says. But those concerns are trumped by the benefits of the heating system. “It's a really good idea to use the heat that would otherwise just vanish into thin air,” Freiesleben says. Would he accept more data centers in his city if they were also hooked up to the heating system? “The simple answer,” he says, “is yes.”

Odense is the first place in the world where Meta has tried to pipe excess heat from a data center directly into people’s homes—but it’s not the only tech company doing so. In Ireland, an Amazon data center already helps heat TU Dublin, a university, while Microsoft is building what is expected to be the world’s largest data center heating system in Espoo, a city in southern Finland.

It’s a trend that’s expected to grow. The current AI boom has been accompanied by a data center construction rush. Companies like Microsoft, Google, Meta, and Amazon will have to invest about $1 trillion on infrastructure to handle computing demand for AI, according to research firm Dell’Oro Group.

All those data centers produce a lot of heat. If that heat is redirected to local homes, it can help Big Tech firms meet their climate pledges, and governments can say they struck a deal that made Big Tech give something back to the local community.

Meta's data center in Odense, Denmark on August 9, 2023.PHOTOGRAPH: ROBBERT FRANK HAGENS/GETTY IMAGES

“Data centers are like giant refrigerators,” says David Lunts, CEO of Old Oak and Park Royal Development Corporation, a UK property developer. In November, the company announced a new development in London which will be heated by a nearby data center.

A lot of the time, the heat these data centers generate is simply released into the atmosphere. And that, Lunts claims, makes using it to heat nearby buildings a no-brainer. “It’s a win for the data centers because they’re not expelling their surplus heat into the atmosphere, which is not great for their green credentials,” he says. It’s also great for developers, he says, who get a source of “cheap heat.”

The heat pumped out by data centers isn’t hot enough to be plugged into people’s houses directly, and an energy company has to both boost the temperature and handle billing. No one in Odense is getting an energy bill straight from Meta. In Denmark, that process is managed by Fjernvarme Fyn. The utility company captures the heat from Meta at around 80 degrees Fahrenheit (around 26.6 degrees Celsius), before boosting it to the 170 degrees Fahrenheit (76.6 degrees Celsius) needed for district heating, says Palle Grøndahl, Fjernvarme Fyn’s acting head of development.

For Big Tech, there are few better places to experiment with data center heating than in the Nordics. This idea works best when data centers can be connected to preexisting district heating systems, where a group of buildings share a common heating system instead of each having their own.

These communal systems are commonplace in countries like Denmark, Finland, and Sweden—and tech isn’t the first industry to experiment with connecting to them.

For the past 20 years, Patrik Öhlund’s home in the northern Swedish city of Luleå has been partly heated by the waste heat from a nearby steel plant. Now, Öhlund, who is director of Energy Markets at Microsoft, is working on recreating this system in the Finnish city of Espoo. But this time it’s Microsoft that’s being hooked up to the local district heating network as part of a project that will eventually heat 100,000 households. Once completed, it’s expected to be the largest data center heating system in the world.

Microsoft’s project in Espoo will generate slightly hotter water—90 degrees Fahrenheit (32.2 degrees Celsius)—than Meta’s Danish system, partly because the Finnish data center will also have the capacity to power AI systems. Finnish energy company Fortum will then boost the heat to between 180 and 250 degrees Fahrenheit (82.2 and 121.1 degrees Celsius), before it enters people’s homes—which should happen sometime after 2025. Heat extracted from data centers that power AI tends to be hotter because they often have a higher-density setup of server racks, says Tom Glover, head of data center transactions at real estate consultancy JLL. “You're provided with a higher quality of heat, which can be used better within district heating grids,” he adds.

When Microsoft’s Espoo system is switched on, energy prices won’t necessarily be cheaper, according to Teemu Nieminen, who leads the data center heat recovery project for Fortum. Neither company will disclose how much Microsoft is charging for the heat, but they do confirm it’s part of a commercial arrangement. It might not be cheaper, but prices should be more stable, says Nieminen, “compared to fossil fuels, where prices fluctuate very wildly.”

Microsoft also hopes this stability will help make data centers on this scale more welcome in local communities, some of whom take issue with Big Tech sucking up huge amounts of renewable power. “It will keep the prices stable, and with people living nearby knowing this … they are also more positive to our data centers,” says Öhlund.

In Ireland, where data centers have been labeled “energy vampires” by environmentalists and were responsible for 19 percent of the country’s total energy use last year, Amazon is giving its data center’s excess heat away for free to the local university and local government offices, according to John O’Shea, senior energy systems analyst at Codema, Dublin’s energy agency. Amazon’s donation is not an entirely selfless act. “We provide free cooling to them as a byproduct of taking their heat,” O’Shea adds.

So far, the arrangement is working. The project is being expanded and the city is in the process of connecting the Amazon data center to more buildings, as well as 135 apartments currently under construction. But O’Shea is wary of endorsing the data centers. Despite the heating project, there are still concerns in the region of County Dublin, where there are between 65 and 70 data centers, about the amount of energy they use. “The development of data centers themselves is something to be discussed,” O’Shea says. “But if they are being developed, we think it makes sense to use this waste source that is otherwise just being pumped into the air or into waterways.”

There’s a similar attitude in Denmark. Data centers will be responsible for 14 percent of the country’s total energy consumption by 2030, according to a forecast by the Danish Energy Agency. Big Tech data centers might source their energy from renewable sources, but a data center hooked up to local homes still requires the energy of more wind turbines than the homes would if they were heated by the turbines directly, says Henrik Lund, professor of energy planning at Denmark’s Aalborg University. “The data centers themselves are putting pressure on the green transition,” he adds.

Source

Later on, in October, Microsoft seemed to realize its mistake and restored several of the CPUs on the list. Most consumers though weren't affected by this at all as they were Xeon SKUs, and even if a user owned such a Xeon processor, Windows 11 likely continued to work nicely on such PCs.

Fast forward a couple of months, and Microsoft has now added many new Intel x86 processors to the list of supported CPUs on Windows 11 22H2 and 23H2. Towards the end of September, the company began ensuring system requirements compatibility for Windows 11 version 23H2 after confirming earlier that hardware eligibility on 23H2 will remain unchanged from 22H2.

The new Intel processors added to the list mainly consist of 14th Gen Intel desktop and laptop CPUs. On the desktop side of things, Intel released new chips based on Raptor Lake-S refresh, while on the mobile front, the company unveiled new Meteor Lake parts that come with a new naming scheme as it (temporarily) dropped the Core "i" nomenclature in favor of "Ultra". The chips also have what Intel refers to as NPUs (Neural Processing Units) and VPUs (Vision Processing Units). Word on the street is that the next Windows version will be released in June of next year alongside a slew of AI-heavy PCs.

The newly added Intel Meteor Lake (mobile) CPUs are:

• Intel Core 3 processor 100U
• Intel Core 5 processor 120U
• Intel Core 7 processor 150U
• Intel Core Ultra 5 processor 125H
• Intel Core Ultra 5 processor 125U
• Intel Core Ultra 5 processor 134U
• Intel Core Ultra 5 processor 135H
• Intel Core Ultra 5 processor 135U
• Intel Core Ultra 7 processor 155H
• Intel Core Ultra 7 processor 155U
• Intel Core Ultra 7 processor 164U
• Intel Core Ultra 7 processor 165H
• Intel Core Ultra 7 processor 165U
• Intel Core Ultra 9 processor 185H

Meanwhile, the newly added Raptor Lake refresh (desktop and a few notebooks) models are:

• Intel Core i9-14900
• Intel Core i9-14900F
• Intel Core i9-14900HX
• Intel Core i9-14900K
• Intel Core i9-14900KF
• Intel Core i9-14900T
• Intel Core i7-14650HX
• Intel Core i7-14700
• Intel Core i7-14700F
• Intel Core i7-14700HX
• Intel Core i7-14700K
• Intel Core i7-14700KF
• Intel Core i7-14700T

You can view the full list of Intel Windows 11-supported processors on Microsoft's website. Meanwhile, on the AMD side, Microsoft seems to have ignored newly announced Ryzen 8000 series parts for some reason.

Source

The restrictions are set to be enforced by early 2024, the company said.

Aside from the U.S., a slew of groundbreaking elections are expected in 2024, including national elections in India, the world's largest democracy, and South Africa, among others.

The tech giant said it will "work with an increased focus on the role artificial intelligence (AI) might play" as it looks to service voters and campaigns related to these elections.

Facebook-owner Meta (META.O) had also said in November it is barring political campaigns and advertisers in other regulated industries from using its new generative AI advertising products.

Advertisers on Meta will also have to disclose when artificial intelligence (AI) or other digital methods are used to alter or create political, social or election related advertisements on Facebook and Instagram.

On the other hand, Elon Musk's social media platform X, which is being probed by the European Union, said in August it would now allow political advertising in the U.S. from candidates and political parties. It will also expand its safety and elections team ahead of the U.S. election.

All political ads had previously been banned globally on X since 2019.

Governments across the globe have been rallying to regulate AI in light of the threats it poses, such as the spread of misinformation.

Big Tech firms will face new European Union rules to clearly label political advertising on their platforms, who paid for it and how much and which elections are being targeted.

Source

Investigating Generic Artificial Intelligence and Brain Decoding

One state-of-the-art method that scientists are using to visualize perceptual contents based on brain activity is brain decoding. Reconstruction of human visual perception from functional magnetic resonance imaging (fMRI) data has been demonstrated in prior research. Nevertheless, these investigations could only cover narrow topics, like the alphabet.

To get around these restrictions, a group of researchers from Osaka University, another national institute, and the National Institutes of Quantum Science and Technology came up with a system that measures brain activity and uses generative AI to create complicated pictures. They were able to generate realistic mental representations of the landscapes and objects shown to the participants by integrating fMRI data with predictive algorithms.

How to Conduct Research

Experiments were meticulously planned as part of the research. While fMRI measured participants’ brain signals in real time, they were shown 1,200 images of different objects and landscapes. A quantitative analysis was conducted on the correlation between the brain signals and the corresponding images.

Next, the generative AI was given these pictures and data on brain activity, and it learned to make new versions of the objects using the patterns it saw. Using an iterative process, the AI progressively became better at creating realistic mental representations that matched the initial inputs.

The Result: Visualizations of Realms and Things in One’s Head

The research yielded absolutely astounding results. Using participants’ brain activity, the researchers were able to create mental images of landscapes and objects that were vague but recognisable. As an example, they were able to create a leopard image that had features like a mouth, ears, and spots. They were also able to replicate an aeroplane whose wings were illuminated by red lights.

These successes represent a watershed moment in our knowledge of how the brain processes visual information. It paves the way for novel approaches to communication technology, medical breakthroughs, and understanding the complex brain.

Consequences for the Welfare and Medical Sectors

This discovery has far-reaching consequences for healthcare and social welfare. By visualizing brain activity, brain decoding technology has the potential to help people with limited communication skills express themselves. People who suffer from disorders like locked-in syndrome or profound speech impairments may find this immensely helpful.

Additionally, scientists have long struggled to decipher the neural bases of hallucinations and dreams. Researchers can obtain valuable insights into the underlying processes by analyzing and visualizing brain activity related to these phenomena. Conditions like schizophrenia and sleep disorders may be better treated with this new information.

Revealing the Power of Communication Tools

Innovative means of communication are one of the most promising areas for future advancements thanks to this discovery. Let your mind wander to a future where people no longer need to rely on written or spoken language to convey their ideas and opinions. This could become a reality with the help of brain decoding technology.

More effective and efficient expression would be possible if brain activity could be decoded and translated into visual representations or text. It has the potential to greatly improve communication for individuals with disabilities and open up new avenues of seamless interaction in many different types of professional settings.

Deciphering the Brain’s Role in Dreams and Hallucinations

Psychologists and scientists have long been enthralled by hallucinations and dreams. Now that brain decoding technology exists, scientists have a potent instrument at their disposal to investigate these phenomena in greater detail. Hallucinations and dreams are unique cognitive experiences, and scientists can learn more about them by studying brain activity linked with them.

Not only in neuroscience, but also in psychiatry and psychology, this information might have huge ramifications. People who suffer from hallucinations or sleep disorders may find new treatments and interventions as a result.

The Next Steps in Artificial Intelligence and Brain Decoding

An important milestone in our quest to understand the brain has been reached with the effective use of brain decoding and generative AI technologies to generate mental images. Additional groundbreaking discoveries are likely to follow as scientists hone their methods and amass more data.

Artificial intelligence (AI) and brain decoding have uses outside of neuroscience. There is an infinite number of potential applications, ranging from medical breakthroughs to communication tools. The more we learn about the brain, the more opportunities we have to improve our lives and make revolutionary discoveries.

FAQ

WHAT IS BRAIN DECODING, AND HOW DOES IT WORK IN THE CONTEXT OF ARTIFICIAL INTELLIGENCE (AI)?

Brain decoding is a method that involves measuring brain activity and using generative AI to create mental images of objects and landscapes based on that activity. In this process, fMRI data and predictive algorithms are integrated to generate realistic mental representations from brain signals.

WHAT WAS THE OBJECTIVE OF THE RESEARCH ON BRAIN DECODING DESCRIBED IN THE ARTICLE?

The research aimed to expand the scope of brain decoding beyond narrow topics, such as the alphabet, by using generative AI. The goal was to create complicated pictures and mental representations of landscapes and objects based on participants’ brain activity.

HOW WERE THE EXPERIMENTS CONDUCTED IN THE RESEARCH ON BRAIN DECODING?

The experiments involved participants who underwent fMRI scans while being shown 1,200 images of various objects and landscapes. The correlation between brain signals and corresponding images was quantitatively analyzed. Generative AI was then used to create new versions of these objects based on the patterns observed.

WHAT WERE THE RESULTS OF THE RESEARCH ON BRAIN DECODING AND GENERATIVE AI?

The research successfully created mental images of landscapes and objects from participants’ brain activity. These images were vague but recognizable and included examples like a leopard with features like a mouth, ears, and spots, as well as an airplane with illuminated red lights on its wings.

WHAT ARE THE POTENTIAL CONSEQUENCES OF THIS DISCOVERY FOR HEALTHCARE AND SOCIAL WELFARE?

Brain decoding technology has the potential to benefit individuals with limited communication skills, such as those with locked-in syndrome or speech impairments. It can also help researchers better understand conditions like schizophrenia and sleep disorders by visualizing brain activity related to hallucinations and dreams.

HOW MIGHT BRAIN DECODING IMPACT COMMUNICATION TOOLS AND INTERACTIONS IN PROFESSIONAL SETTINGS?

Brain decoding technology could lead to more efficient means of communication, translating brain activity into visual representations or text. This could greatly improve communication for individuals with disabilities and create new possibilities for seamless interaction in various professional settings.

WHAT IMPLICATIONS DOES BRAIN DECODING HAVE FOR THE STUDY OF DREAMS AND HALLUCINATIONS?

Brain decoding provides scientists with a powerful tool to investigate dreams and hallucinations by analyzing brain activity associated with these phenomena. This research can lead to insights in neuroscience, psychiatry, and psychology, potentially improving treatments for related conditions.

WHAT ARE THE NEXT STEPS IN THE DEVELOPMENT OF ARTIFICIAL INTELLIGENCE AND BRAIN DECODING?

The successful use of brain decoding and generative AI to generate mental images represents an important milestone in understanding the brain. Further discoveries and applications are expected as scientists refine their methods and gather more data. The potential applications of AI and brain decoding extend beyond neuroscience and encompass various fields, from medicine to communication tools.

Source