An anonymous reader quotes a report from Gizmodo: Clearview AI, the shady face recognition firm which claims to have landed contracts with federal, state, and local cops across the country, has landed a roughly $50,000 deal with the U.S. military for augmented reality glasses. First flagged by Tech Inquiry's Jack Poulson, Air Force procurement documents show that it awarded a $49,847 contract to Clearview AI for the purposes of "protecting airfields with augmented reality facial recognition; glasses." The contract is designated as part of the Small Business Innovation Research (SBIR) program, meaning that Clearview's contract is to determine for the Air Force whether such applications are feasible. Bryan Ripple, a media lead at the Air Force Research Laboratory Public Affairs, told Gizmodo via email that Clearview will conduct a three-month study under which "no glasses or units are being delivered under contract," nor are any prototypes. Clearview, he wrote, stipulated "that security personnel are vulnerable while their hands are occupied with scanners and ID cards" and AR goggles would allow them to "remain hands-free and ready during this timeframe." "Clearview AI's Augmented Reality (AR) Glasses perform facial recognition scanning to vet backgrounds and restrict unauthorized individuals from entering bases and flightlines," Ripple wrote. "This 100% hands-free identity verification wearable device allows Defenders to keep their weapons at the ready, increase standoff and social distance, and confirm authorized base access using rapid and accurate facial biometrics while keeping threats distant. The results are improved safety at entry control points and for bases, faster identity verification without manual ID card checks, and cost savings by replacing the need for large permanent camera installations." In a promotional document shared by the Air Force, Clearview argued that in the time it takes to scan an ID card at the entry point to a military facility, "A criminal or terrorist can pull a gun, knife, or weapon during this brief but critical moment, kill the Defender, and access the base." They argued the AR glasses would increase "standoff distance," save guards time while vetting high volumes of traffic and allow them to maintain distance from anyone contagious with diseases.

Read more of this story at Slashdot.

A new study out of Germany says having traffic lights use AI technology may keep traffic flowing faster and smoother. Jalopnik reports: One of the partners in the study with an aggressively German name -- the Fraunhofer Institute for Optronics, System Technologies and Image Exploitation -- recently installed high-resolution cameras and radar sensors at a busy intersection with a traffic light in the city of Lemgo, according to New Atlas. The setup recorded the number of vehicles waiting for the light to change, the amount of time each of them had to wait and the average speed a vehicle drove through the intersection. Science wizardry was then used to train a machine-learning based computer algorithm. It experimented with different light-changing patterns. They would continuously adapt to real time traffic conditions and see which ones worked best to keep wait times down. According to the simulations, the best artificial intelligence patterns could improve traffic flow by 10 to 15 percent. That may not sound like a ton, but add up all the time you spend white-knuckled at a long traffic light, and chop 15 percent off. Not too bad. The algorithm will be used to run the traffic lights at actual intersections in Germany for the next few months, and can only get better. The study is also looking to find ways to reduce waiting times at crosswalks for pedestrians. They're using LiDAR sensors among other things to assess the walking speed of pedestrians to make sure they have enough time to cross before the light turns on them.

Read more of this story at Slashdot.

DeepMind has created an AI system named AlphaCode that it says "writes computer programs at a competitive level." From a report: The Alphabet subsidiary tested its system against coding challenges used in human competitions and found that its program achieved an "estimated rank" placing it within the top 54 percent of human coders. The result is a significant step forward for autonomous coding, says DeepMind, though AlphaCode's skills are not necessarily representative of the sort of programming tasks faced by the average coder. Oriol Vinyals, principal research scientist at DeepMind, told The Verge over email that the research was still in the early stages but that the results brought the company closer to creating a flexible problem-solving AI -- a program that can autonomously tackle coding challenges that are currently the domain of humans only. "In the longer-term, we're excited by [AlphaCode's] potential for helping programmers and non-programmers write code, improving productivity or creating new ways of making software," said Vinyals.

Read more of this story at Slashdot.

Long-time Slashdot reader Hope Thelps brings news about the future of self-driving cars. "The law commisions of England and Wales and of Scotland (statutory bodies which keep the laws in those countries under review) are recommending a shift in accident liability away from 'drivers' when autonomous cars become a reality." The BBC reports: Human drivers should not be legally accountable for road safety in the era of autonomous cars, a report says. In these cars, the driver should be redefined as a "user-in-charge", with very different legal responsibilities, according to the law commissions for England and Wales, and Scotland. If anything goes wrong, the company behind the driving system would be responsible, rather than the driver.... In the interim, carmakers must be extremely clear about the difference between self-drive and driver-assist features. There should be no sliding scale of driverless capabilities — a car is either autonomous or not.... Transport Minister Trudy Harrison said the government would "fully consider" the recommendations. The Scottish and Welsh governments will also decide whether to introduce legislation. The BBC also summarized some of the reports other recommendations: Data to understand fault and liability following a collision must be accessible Sanctions for carmakers who fail to reveal how their systems work

Read more of this story at Slashdot.

"Focus on the horse race and the flashy news and you'll miss the real stories," argues Mike Loukides, the content strategy VP at O'Reilly Media. So instead he shares trends observed on O'Reilly's learning platform in the first nine months of 2021: While new technologies may appear on the scene suddenly, the long, slow process of making things that work rarely attracts as much attention. We start with an explosion of fantastic achievements that seem like science fiction — imagine, GPT-3 can write stories! — but that burst of activity is followed by the process of putting that science fiction into production, of turning it into real products that work reliably, consistently, and fairly. AI is making that transition now; we can see it in our data. But what other transitions are in progress...? Important signals often appear in technologies that have been fairly stable. For example, interest in security, after being steady for a few years, has suddenly jumped up, partly due to some spectacular ransomware attacks. What's important for us isn't the newsworthy attacks but the concomitant surge of interest in security practices — in protecting personal and corporate assets against criminal attackers. That surge is belated but healthy.... Usage of content about ransomware has almost tripled (270% increase). Content about privacy is up 90%; threat modeling is up 58%; identity is up 50%; application security is up 45%; malware is up 34%; and zero trust is up 23%. Safety of the supply chain isn't yet appearing as a security topic, but usage of content about supply chain management has seen a healthy 30% increase.... Another important sign is that usage of content about compliance and governance was significantly up (30% and 35%, respectively). This kind of content is frequently a hard sell to a technical audience, but that may be changing.... This increase points to a growing sense that the technology industry has gotten a regulatory free ride and that free ride is coming to an end. Whether it's stockholders, users, or government agencies who demand accountability, enterprises will be held accountable. Our data shows that they're getting the message. According to a study by UC Berkeley's School of Information, cybersecurity salaries have crept slightly ahead of programmer salaries in most states, suggesting increased demand for security professionals. And an increase in demand suggests the need for training materials to prepare people to supply that demand. We saw that play out on our platform.... C++ has grown significantly (13%) in the past year, with usage that is roughly twice C's. (Usage of content about C is essentially flat, down 3%.) We know that C++ dominates game programming, but we suspect that it's also coming to dominate embedded systems, which is really just a more formal way to say "internet of things." We also suspect (but don't know) that C++ is becoming more widely used to develop microservices. On the other hand, while C has traditionally been the language of tool developers (all of the Unix and Linux utilities are written in C), that role may have moved on to newer languages like Go and Rust. Go and Rust continue to grow. Usage of content about Go is up 23% since last year, and Rust is up 31%. This growth continues a trend that we noticed last year, when Go was up 16% and Rust was up 94%.... Both Rust and Go are here to stay. Rust reflects significantly new ways of thinking about memory management and concurrency. And in addition to providing a clean and relatively simple model for concurrency, Go represents a turn from languages that have become increasingly complex with every new release. Other highlights from their report: "Quantum computing remains a topic of interest. Units viewed is still small, but year-over-year growth is 39%. That's not bad for a technology that, honestly, hasn't been invented yet...." "Whether it's the future of finance or history's biggest Ponzi scheme, use of content about cryptocurrency is up 271%, with content about the cryptocurrencies Bitcoin and Ethereum (ether) up 166% and 185% respectively...." "Use of JavaScript content on our platform is surprisingly low — though use of content on TypeScript (a version of JavaScript with optional static typing) is up.... Even with 19% growth, TypeScript has a ways to go before it catches up; TypeScript content usage is roughly a quarter of JavaScript's..." "Python, Java, and JavaScript are still the leaders, with Java up 4%, Python down 6%, and JavaScript down 3%...." "Finally, look at the units viewed for Linux: it's second only to Kubernetes. While down very slightly in 2021, we don't believe that's significant. Linux has long been the most widely used server operating system, and it's not ceding that top spot soon."

Read more of this story at Slashdot.

Slashdot reader sciencehabit quotes Science magazine: Imagine using any object around you—a frying pan, a glass paperweight—as the central processor in a neural network, a type of artificial intelligence that loosely mimics the brain to perform complex tasks. That's the promise of new research that, in theory, could be used to recognize images or speech faster and more efficiently than computer programs that rely on silicon microchips. To demonstrate the concept, the researchers built neural networks in three types of physical systems, which each contained up to five processing layers. In each layer of a mechanical system, they used a speaker to vibrate a small metal plate and recorded its output using a microphone. In an optical system, they passed light through crystals. And in an analog-electronic system, they ran current through tiny circuits. In each case, the researchers encoded input data, such as unlabeled images, in sound, light, or voltage. For each processing layer, they also encoded numerical parameters telling the physical system how to manipulate the data. To train the system, they adjusted the parameters to reduce errors between the system's predicted image labels and the actual labels. In one task, they trained the systems, which they call physical neural networks (PNNs), to recognize handwritten digits. In another, the PNNs recognized seven vowel sounds. Accuracy on these tasks ranged from 87% to 97%, they report in this week's issue of Nature. In the future, researchers might tune a system not by digitally tweaking its input parameters, but by adjusting the physical objects—warping the metal plate, say. The team is most excited about PNNs' potential as smart sensors that can perform computation on the fly. A microscope's optics might help detect cancerous cells before the light even hits a digital sensor, or a smartphone's microphone membrane might listen for wake words. These "are applications in which you really don't think about them as performing a machine-learning computation," they say, but instead as being "functional machines."

Read more of this story at Slashdot.

An anonymous reader quotes a report from The Wall Street Journal: Meta said Monday that its research team built a new artificial intelligence supercomputer that the company maintains will soon be the fastest in the world. The supercomputer, the AI Research SuperCluster, was the result of nearly two years of work, often conducted remotely during the height of the pandemic, and led by the Facebook parent's AI and infrastructure teams. Several hundred people, including researchers from partners Nvidia, Penguin Computing and Pure Storage, were involved in the project, the company said. Meta, which announced the news in a blog post Monday, said its research team currently is using the supercomputer to train AI models in natural-language processing and computer vision for research. The aim is to boost capabilities to one day train models with more than a trillion parameters on data sets as large as an exabyte, which is roughly equivalent to 36,000 years of high-quality video. "The experiences we're building for the metaverse require enormous compute powerand RSC will enable new AI models that can learn from trillions of examples, understand hundreds of languages, and more," Meta CEO Mark Zuckerberg said in a statement provided to The Wall Street Journal. Meta's AI supercomputer houses 6,080 Nvidia graphics-processing units, putting it fifth among the fastest supercomputers in the world, according to Meta. By mid-summer, when the AI Research SuperCluster is fully built, it will house some 16,000 GPUs, becoming the fastest AI supercomputer in the world, Meta said. The company declined to comment on the location of the facility or the cost. [...] Eventually the supercomputer will help Meta's researchers build AI models that can work across hundreds of languages, analyze text, images and video together and develop augmented reality tools, the company said. The technology also will help Meta more easily identify harmful content and will aim to help Meta researchers develop artificial-intelligence models that think like the human brain and support rich, multidimensional experiences in the metaverse. "In the metaverse, it's one hundred percent of the time, a 3-D multi-sensorial experience, and you need to create artificial-intelligence agents in that environment that are relevant to you," said Jerome Pesenti, vice president of AI at Meta.

Read more of this story at Slashdot.

Slashdot reader DevNull127 shares this transcript of James Cameron's new interview with the BBC — which they've titled "The Danger of Deepfakes." "Almost everything we create seems to go wrong at some point," James Cameron says... James Cameron: Almost everything we create seems to go wrong at some point. I've worked at the cutting edge of visual effects, and our goal has been progressively to get more and more photo-real. And so every time we improve these tools, we're actually in a sense building a toolset to create fake media — and we're seeing it happening now. Right now the tools are — the people just playing around on apps aren't that great. But over time, those limitations will go away. Things that you see and fully believe you're seeing could be faked. This is the great problem with us relying on video. The news cycles happen so fast, and people respond so quickly, you could have a major incident take place between the interval between when the deepfake drops and when it's exposed as a fake. We've seen situations — you know, Arab Spring being a classic example — where with social media, the uprising was practically overnight. You have to really emphasize critical thinking. Where did you hear that? You know, we have all these search tools available, but people don't use them. Understand your source. Investigate your source. Is your source credible? But we also shouldn't be prone to this ridiculous conspiracy paranoia. People in the science community don't just go, 'Oh that's great!' when some scientist, you know, publishes their results. No, you go in for this big period of peer review. It's got to be vetted and checked. And the more radical a finding, the more peer review there is. So good peer-reviewed science can't lie. But people's minds, for some reason, will go to the sexier, more thriller-movie interpretation of reality than the obvious one. I always use Occam's razor — you know, Occam's razor's a great philosophical tool. It says the simplest explanation is the likeliest. And conspiracy theories are all too complicated. People aren't that good, human systems aren't that good, people can't keep a secret to save their lives, and most people in positions of power are bumbling stooges. The fact that we think that they could realistically pull off these — these complex plots? I don't buy any of that crap! Bill Gates is not really trying to microchip you with the flu vaccine! [Laughs] You know, look, I'm always skeptical of new technology, and we all should be. Every single advancement in technology that's ever been created has been weaponized. I say this to AI scientists all the time, and they go, 'No, no, no, we've got this under control.' You know, 'We just give the AIs the right goals...' So who's deciding what those goals are? The people that put up the money for the research, right? Which are all either big business or defense. So you're going to teach these new sentient entities to be either greedy or murderous. If Skynet wanted to take over and wipe us out, it would actually look a lot like what's going on right now. It's not going to have to — like, wipe out the entire, you know, biosphere and environment with nuclear weapons to do it. It's going to be so much easier and less energy required to just turn our minds against ourselves. All Skynet would have to do is just deepfake a bunch of people, pit them against each other, stir up a lot of foment, and just run this giant deepfake on humanity. I mean, I could be a projection of an AI right now.

Read more of this story at Slashdot.

An anonymous reader quotes a report from TechCrunch: Meta (AKA Facebook) researchers are working on [...] an AI that can learn capably on its own whether it does so in spoken, written or visual materials. The traditional way of training an AI model to correctly interpret something is to give it lots and lots (like millions) of labeled examples. A picture of a cat with the cat part labeled, a conversation with the speakers and words transcribed, etc. But that approach is no longer in vogue as researchers found that it was no longer feasible to manually create databases of the sizes needed to train next-gen AIs. Who wants to label 50 million cat pictures? Okay, a few people probably -- but who wants to label 50 million pictures of common fruits and vegetables? Currently some of the most promising AI systems are what are called self-supervised: models that can work from large quantities of unlabeled data, like books or video of people interacting, and build their own structured understanding of what the rules are of the system. For instance, by reading a thousand books it will learn the relative positions of words and ideas about grammatical structure without anyone telling it what objects or articles or commas are -- it got it by drawing inferences from lots of examples. This feels intuitively more like how people learn, which is part of why researchers like it. But the models still tend to be single-modal, and all the work you do to set up a semi-supervised learning system for speech recognition won't apply at all to image analysis -- they're simply too different. That's where Facebook/Meta's latest research, the catchily named data2vec, comes in. The idea for data2vec was to build an AI framework that would learn in a more abstract way, meaning that starting from scratch, you could give it books to read or images to scan or speech to sound out, and after a bit of training it would learn any of those things. It's a bit like starting with a single seed, but depending on what plant food you give it, it grows into an daffodil, pansy or tulip. Testing data2vec after letting it train on various data corpi showed that it was competitive with and even outperformed similarly sized dedicated models for that modality. (That is to say, if the models are all limited to being 100 megabytes, data2vec did better -- specialized models would probably still outperform it as they grow.)

Read more of this story at Slashdot.

Mustafa Suleyman, a pioneer in the field of artificial intelligence, is leaving Google to join the venture capital firm Greylock Partners. From a report: The departure of Mr. Suleyman, who was Google's vice president of product management and policy for artificial intelligence, closes a tumultuous tenure at the company. He joined Google in 2014 when the search giant acquired DeepMind, a cutting-edge artificial intelligence research lab, in a deal valued at $650 million. The deal demonstrated the value of companies that specialized in "deep learning," a form of artificial intelligence that became more important in the early part of the last decade. In just a few years, DeepMind had hired many of the leading researchers in the field. Mr. Suleyman, known to friends and colleagues as Moose, was not an A.I. researcher by training. But he led the company into an important area of research: health care. He also became a key voice in DeepMind's efforts to ensure that its technologies would not be used for military applications, which led to a clash with Google when the company joined a flagship A.I. project with the Defense Department. (Google eventually pulled out of the project.)

Read more of this story at Slashdot.

Good poker players have always known that they need to maintain a balance between bluffing and playing it straight. Now they can do so perfectly. From a report: One of the earliest and most devoted adopters of what has come to be known as "game theory optimal" poker is Seth Davies's friend and poker mentor, Jason Koon. On the second day of the three-day Super High Roller tournament, I visited Koon at his multimillion-dollar house, located in a gated community inside a larger gated community next to a Jack Nicklaus-designed golf course. On Day 1, Koon paid $250,000 to play the Super High Roller, then a second $250,000 after he was knocked out four hours in, but again he lost all his chips. "Welcome to the world of nosebleed tourneys," he texted me afterward. "Just have to play your best -- it evens out." For Koon, evening out has taken the form of more than $30 million in in-person tournament winnings (and, he says, at least as much from high-stakes cash games in Las Vegas and Macau, the Asian gambling mecca). Koon began playing poker seriously in 2006 while rehabbing an injury at West Virginia Wesleyan College, where he was a sprinter on the track team. He made a good living from cards, but he struggled to win consistently in the highest-stakes games. "I was a pretty mediocre player pre-solver," he says, "but the second solvers came out, I just buried myself in this thing, and I started to improve like rapidly, rapidly, rapidly, rapidly." In a home office decorated mostly with trophies from poker tournaments he has won, Koon turned to his computer and pulled up a hand on PioSOLVER. After specifying the size of the players' chip stacks and the range of hands they would play from their particular seats at the table, he entered a random three-card flop that both players would see. A 13-by-13 grid illustrated all the possible hands one of the players could hold. Koon hovered his mouse over the square for an ace and queen of different suits. The solver indicated that Koon should check 39 percent of the time; make a bet equivalent to 30 percent the size of the pot 51 percent of the time; and bet 70 percent of the pot the rest of the time. This von Neumann-esque mixed strategy would simultaneously maximize his profit and disguise the strength of his hand. Thanks to tools like PioSOLVER, Koon has remade his approach to the game, learning what size bets work best in different situations. Sometimes tiny ones, one-fifth or even one-tenth the size of the pot, are ideal; other times, giant bets two or three times the size of the pot are correct. And, while good poker players have always known that they need to maintain a balance between bluffing and playing it straight, solvers define the precise frequency with which Koon should employ one tactic or the other and identify the (sometimes surprising) best and worst hands to bluff with, depending on the cards in play.

Read more of this story at Slashdot.

Nvidia's latest game-ready driver includes a tool that could let you improve the image quality of games that your graphics card can easily run, alongside optimizations for the new God of War PC port. The Verge reports: The tech is called Deep Learning Dynamic Super Resolution, or DLDSR, and Nvidia says you can use it to make "most games" look sharper by running them at a higher resolution than your monitor natively supports. DLDSR builds on Nvidia's Dynamic Super Resolution tech, which has been around for years. Essentially, regular old DSR renders a game at a higher resolution than your monitor can handle and then downscales it to your monitor's native resolution. This leads to an image with better sharpness but usually comes with a dip in performance (you are asking your GPU to do more work, after all). So, for instance, if you had a graphics card capable of running a game at 4K but only had a 1440p monitor, you could use DSR to get a boost in clarity. DLDSR takes the same concept and incorporates AI that can also work to enhance the image. According to Nvidia, this means you can upscale less (and therefore lose less performance) while still getting similar image quality improvements. In real numbers, Nvidia claims you'll get image quality similar to running at four times the resolution using DSR with only 2.25 times the resolution with DLDSR. Nvidia gives an example using 2017's Prey: Digital Deluxe running on a 1080p monitor: 4x DSR runs at 108 FPS, while 2.25x DLDSR is getting 143 FPS, only two frames per second slower than running at native 1080p.

Read more of this story at Slashdot.

During the past 20 years, deep learning has come to dominate artificial intelligence research and applications through a series of useful commercial applications. But underneath the dazzle are some deep-rooted problems that threaten the technology's ascension. IEEE Spectrum: The inability of a typical deep learning program to perform well on more than one task, for example, severely limits application of the technology to specific tasks in rigidly controlled environments. More seriously, it has been claimed that deep learning is untrustworthy because it is not explainable -- and unsuitable for some applications because it can experience catastrophic forgetting. Said more plainly, if the algorithm does work, it may be impossible to fully understand why. And while the tool is slowly learning a new database, an arbitrary part of its learned memories can suddenly collapse. It might therefore be risky to use deep learning on any life-or-death application, such as a medical one. Now, in a new book, IEEE Fellow Stephen Grossberg argues that an entirely different approach is needed. Conscious Mind, Resonant Brain: How Each Brain Makes a Mind describes an alternative model for both biological and artificial intelligence based on cognitive and neural research Grossberg has been conducting for decades. He calls his model Adaptive Resonance Theory (ART). Grossberg -- an endowed professor of cognitive and neural systems, and of mathematics and statistics, psychological and brain sciences, and biomedical engineering at Boston University -- based ART on his theories about how the brain processes information. "Our brains learn to recognize and predict objects and events in a changing world that is filled with unexpected events," he says. Based on that dynamic, ART uses supervised and unsupervised learning methods to solve such problems as pattern recognition and prediction. Algorithms using the theory have been included in large-scale applications such as classifying sonar and radar signals, detecting sleep apnea, recommending movies, and computer-vision-based driver-assistance software. [...] One of the problems faced by classical AI, he says, is that it often built its models on how the brain might work, using concepts and operations that could be derived from introspection and common sense. "Such an approach assumes that you can introspect internal states of the brain with concepts and words people use to describe objects and actions in their daily lives," he writes. "It is an appealing approach, but its results were all too often insufficient to build a model of how the biological brain really works." The problem with today's AI, he says, is that it tries to imitate the results of brain processing instead of probing the mechanisms that give rise to the results. People's behaviors adapt to new situations and sensations "on the fly," Grossberg says, thanks to specialized circuits in the brain. People can learn from new situations, he adds, and unexpected events are integrated into their collected knowledge and expectations about the world.

Read more of this story at Slashdot.

silverjacket shares a report from Science.org: [A]n AI has shown it can tag people based on their chess-playing behavior, an advance in the field of "stylometrics" that could help computers be better chess teachers or more humanlike in their game play. Alarmingly, the system could also be used to help identify and track people who think their online behavior is anonymous. [...] To design and train their AI, the researchers tapped an ample resource: more than 50 million human games played on the Lichess website. They collected games by players who had played at least 1000 times and sampled sequences of up to 32 moves from those games. They coded each move and fed them into a neural network that represented each game as a point in multidimensional space, so that each player's games formed a cluster of points. The network was trained to maximize the density of each player's cluster and the distance between those of different players. That required the system to recognize what was distinctive about each player's style. The researchers tested the system by seeing how well it distinguished one player from another. They gave the system 100 games from each of about 3000 known players, and 100 fresh games from a mystery player. To make the task harder, they hid the first 15 moves of each game. The system looked for the best match and identified the mystery player 86% of the time, the researchers reported last month at the Conference on Neural Information Processing Systems (NeurIPS). "We didn't quite believe the results," says Reid McIlroy-Young, a student in Anderson's lab and the paper's primary author. A non-AI method was only 28% accurate. [...] The researchers are aware of the privacy risks posed by the system, which could be used to unmask anonymous chess players online. With tweaks, McIlroy-Young says, it could do the same for poker. And in theory, they say, given the right data sets, such systems could identify people based on the quirks of their driving or the timing and location of their cellphone use.

Read more of this story at Slashdot.

"Who needs Terminators when you have precision clickbait and ultra-deepfakes?" asks IEEE Spectrum: Hollywood's worst-case scenario involving artificial intelligence (AI) is familiar as a blockbuster sci-fi film: Machines acquire humanlike intelligence, achieving sentience, and inevitably turn into evil overlords that attempt to destroy the human race. This narrative capitalizes on our innate fear of technology, a reflection of the profound change that often accompanies new technological developments. However, as Malcolm Murdock, machine-learning engineer and author of the 2019 novel The Quantum Price, puts it, "AI doesn't have to be sentient to kill us all. There are plenty of other scenarios that will wipe us out before sentient AI becomes a problem." Their article presents six real-world AI worst-case scenarios that "could simply happen by default, unfolding organically — that is, if nothing is done to stop them." It includes the possibility of deepfakes and large-scale disinformation, as well as AI-enabled "predictive control" that ultimately robs us of our free will. But it also presents an alternative worst-case scenario: that "we become so scared of the power of this tremendous technology that we resist harnessing it for the actual good it can do in the world." Thanks to Slashdot reader schwit1 for sharing the article.

Read more of this story at Slashdot.

When it comes to forecasting the elements, many seem ready to welcome the machine. But humans still outperform the algorithms -- especially in bad conditions. From a report: [...] Similarly, research published by NOAA Weather Prediction Service director David Novak and his colleagues show that while human forecasters may not be able to "beat" the models on your typical sunny, fair-weather day, they still produce more accurate predictions than the algorithm-crunchers in bad weather. Over the two decades of information Novak's team studied, humans were 20 to 40 percent more accurate at forecasting near-future precipitation than the Global Forecast System (GFS) and the North American Mesoscale Forecast System (NAM), the most commonly used national models. Humans also made statistically significant improvements to temperature forecasting over both model's guidance. "Oftentimes, we find that in the bigger events is when the forecasters can make some value-added improvements to the automated guidance," says Novak. Particularly in adverse conditions, great improvements to the model's forecast were usually due to human augmentation, he adds. This is even more true for local, severe events like thunderstorms and tornadoes, which rely on split-second decision-making in order to save lives. As forecasters become more familiar with a particular model, they begin to notice its biases and failings, Novak adds. Just like the model learns from us, we learn from the model.

Read more of this story at Slashdot.

The latest painting to receive the reconstructed-by-AI treatment is Gustav Klimt's 1900 painting "Philosophy". The Washington Post reports: For decades, only black-and-white photographs of "Philosophy" existed. Now, thanks to artificial intelligence, we can see the work in full color. But does the re-creation really look like the original? Does it even look like a Klimt? The new version, created by Google Arts and Culture using machine learning, shows a very different Klimt than you'd expect if you're familiar with "The Kiss" or "Portrait of Adele Bloch-Bauer I...." "I don't know any better than Google what those paintings really look like, but I don't think that they looked like that," says Jane Kallir, longtime director of the Galerie St. Etienne in New York, which gave Klimt his first shows in the United States. "These things look like cartoons. They don't look like Klimt paintings. "It's like people who try to clone their dogs. You can do it, but it's not the same dog." The paintings are one of several recent attempts to use artificial intelligence to re-create lost art. The Rijksmuseum in Amsterdam used AI to reconstruct missing panels from the edges of Rembrandt's famous "Night Watch" and, over the summer, temporarily installed them alongside the real thing. A pair of researchers in the United Kingdom, who call themselves Oxia Palus, say they've rebuilt a Picasso nude that was hidden beneath "The Blind Man's Meal," using 3-D printing and AI. In October, an orchestra in Bonn, Germany, "played" Beethoven's 10th and unfinished symphony in full. The version was written by an algorithm. George Cann, co-founder of Oxia Palus, posits that artificial intelligence "could give us this parallel alternative universe of art that we never really quite had." It's an alluring idea. Peek beneath a Picasso at an earlier painting under the surface layer and it's like you're peering into the artist's mind, eavesdropping on thoughts from a century ago. See a painting that was lost to catastrophe come back to life and it's like you've traveled back in time, reversed fate. But if any of this re-created universe of lost art, like "Philosophy," is inaccurate, the AI creators might not be resurrecting history but inadvertently rewriting it.... [F]or Kallir, there is little of Klimt in what she calls the "gaudy" re-creations, adding that the paintings would have been more subdued, with smoother transitions from one color to the next. "If you've got a decent eye, and you look at the black-and-white reproductions and compare them to other paintings that were done around the same time, you can probably get a better idea of what they really look like," she says.

Read more of this story at Slashdot.

The Washington Post says that when it comes to making New Year's resolutions, the Smithsonian has a better idea. "What if instead of relying on our own resolutions we asked an AI what it thinks we should do?" Starting this weekend, the "Futures" exhibit both online and at its Arts and Industries Building offers a "Resolutions Generator," an AI that makes suggestions on what commitments we should undertake for 2022.... It sounds like a slightly weird idea, and I'd be lying if I said it didn't turn up some weird results. "Change my name to one of my favorite shapes," it suggests, or "Every Friday for a year I will wear a different hat." And, "Every time I hear bells for a month, I will paint a potato." Designed by AI researcher-writer Janelle Shane, the generator's odd results are deliberate; she purposely trained the AI (the powerful GPT-3) with some of the wackier resolutions humans have put online, then set its parameters wide. "We wanted the AI to come up with the kind of interesting resolutions we're not thinking of," Shane said. "We wanted whimsy," added Rachel Goslins, the director of the Arts and Industries Building, "with a little bit of real." Okay, so probably not many people will really "Go into a library, climb up onto a shelf, yell down 'I am a giant giraffe!'" But it's a lot easier than trying to lose those 15 pounds. And this way you end up in a library. Plus they have a point. The truth is by accessing the collective corpus of human resolutions, AI might conceive of ideas that our pale human pea brains cannot... [T]here are growing piles of evidence that deploying AI that can think faster and even differently will pay dividends in the real world. A Stanford study last month concluded that AI sped up discoveries on coronavirus antiviral drugs by as much as a month, potentially saving lives. Canadian researchers in September found that AI made consistently better choices than doctors in treating behavioral problems. Even a button-down institution like Deloitte has a staffer who has persuasively argued that we should use AI, not humans, to update government regulations. The exhibit's AI also generated these New Year's resolutions: "Treat every dog I meet like a celebrity." "Every time I see a mirror I will remember that it is the gateway to another dimension." The AI researcher behind the project also generated Slashdot headlines back in 2017, using 162,000 headlines from the site's first 20 years. Some of my favorites: More Pong Users for Kernel Project Red Hat Releases Linux Games And Moon Why Open Source Power Man Sues Java Microsoft Releases New Months Ask Slashdot: Do We Want To Be the Computers?

Read more of this story at Slashdot.

The Next Web's Tristan Greene combed through a recently published "position paper" detailing China's views on military AI regulation and found that it "makes absolutely no mention of restricting the use of machines capable of choosing and firing on targets autonomously." From the report: Per the paper: "In terms of law and ethics, countries need to uphold the common values of humanity, put people's well-being front and center, follow the principle of AI for good, and observe national or regional ethical norms in the development, deployment and use of relevant weapon systems." Neither the US or the PRC has any laws, rules, or regulations currently restricting the development or use of military LAWs. The paper's rhetoric may be empty, but there's still a lot we can glean from its contents. Research analyst Megha Pardhi, writing for the Asia Times, recently opined it was intended to signal that China's seeking to "be seen as a responsible state," and that it may be concerned over its progress in the field relative to other superpowers. According to Pardhi: "Beijing is likely talking about regulation out of fear either that it cannot catch up with others or that it is not confident of its capabilities. Meanwhile, formulating a few commonly agreeable rules on weaponization of AI would be prudent." "Despite the fact that neither the colonel's article nor the PRC's position paper mention LAWs directly, it's apparent that what they don't say is what's really at the heart of the issue," concludes Greene. "The global community has every reason to believe, and fear, that both China and the US are actively developing LAWS."

Read more of this story at Slashdot.

Amazon has updated its Alexa voice assistant after it "challenged" a 10-year-old girl to touch a coin to the prongs of a half-inserted plug. From a report: The suggestion came after the girl asked Alexa for a "challenge to do". "Plug in a phone charger about halfway into a wall outlet, then touch a penny to the exposed prongs," the smart speaker said. Amazon said it fixed the error as soon as the company became aware of it. The girl's mother, Kristin Livdahl, described the incident on Twitter. She said: "We were doing some physical challenges, like laying down and rolling over holding a shoe on your foot, from a [physical education] teacher on YouTube earlier. Bad weather outside. She just wanted another one." That's when the Echo speaker suggested partaking in the challenge that it had "found on the web". The dangerous activity, known as "the penny challenge", began circulating on TikTok and other social media websites about a year ago.

Read more of this story at Slashdot.