An anonymous reader quotes a report from MIT Technology Review: DeepMind has used its board-game playing AI AlphaZero to discover a faster way to solve a fundamental math problem in computer science, beating a record that has stood for more than 50 years. A year after it took biologists by surprise, AlphaFold has changed how researchers work and set DeepMind on a new course. The problem, matrix multiplication, is a crucial type of calculation at the heart of many different applications, from displaying images on a screen to simulating complex physics. It is also fundamental to machine learning itself. Speeding up this calculation could have a big impact on thousands of everyday computer tasks, cutting costs and saving energy. Despite the calculation's ubiquity, it is still not well understood. A matrix is simply a grid of numbers, representing anything you want. Multiplying two matrices together typically involves multiplying the rows of one with the columns of the other. The basic technique for solving the problem is taught in high school. But things get complicated when you try to find a faster method. This is because there are more ways to multiply two matrices together than there are atoms in the universe (10 to the power of 33, for some of the cases the researchers looked at). The trick was to turn the problem into a kind of three-dimensional board game, called TensorGame. The board represents the multiplication problem to be solved, and each move represents the next step in solving that problem. The series of moves made in a game therefore represents an algorithm. The researchers trained a new version of AlphaZero, called AlphaTensor, to play this game. Instead of learning the best series of moves to make in Go or chess, AlphaTensor learned the best series of steps to make when multiplying matrices. It was rewarded for winning the game in as few moves as possible. [...] The researchers describe their work in a paper published in Nature today. The headline result is that AlphaTensor discovered a way to multiply together two four-by-four matrices that is faster than a method devised in 1969 by the German mathematician Volker Strassen, which nobody had been able to improve on since. The basic high school method takes 64 steps; Strassen's takes 49 steps. AlphaTensor found a way to do it in 47 steps. "Overall, AlphaTensor beat the best existing algorithms for more than 70 different sizes of matrix," concludes the report. "It reduced the number of steps needed to multiply two nine-by-nine matrices from 511 to 498, and the number required for multiplying two 11-by-11 matrices from 919 to 896. In many other cases, AlphaTensor rediscovered the best existing algorithm."

Read more of this story at Slashdot.

Not to be outdone by Meta's Make-A-Video, Google today detailed its work on Imagen Video, an AI system that can generate video clips given a text prompt (e.g., "a teddy bear washing dishes"). From a report:While the results aren't perfect -- the looping clips the system generates tend to have artifacts and noise -- Google claims that Imagen Video is a step toward a system with a "high degree of controllability" and world knowledge, including the ability to generate footage in a range of artistic styles. As my colleague Devin Coldewey noted in his piece about Make-A-Video, text-to-video systems aren't new. Earlier this year, a group of researchers from Tsinghua University and the Beijing Academy of Artificial Intelligence released CogVideo, which can translate text into reasonably-high-fidelity short clips. But Imagen Video appears to be a significant leap over the previous state-of-the-art, showing an aptitude for animating captions that existing systems would have trouble understanding. "It's definitely an improvement," Matthew Guzdial, an assistant professor at the University of Alberta studying AI and machine learning, told TechCrunch via email. "As you can see from the video examples, even though the comms team is selecting the best outputs there's still weird blurriness and artificing. So this definitely is not going to be used directly in animation or TV anytime soon. But it, or something like it, could definitely be embedded in tools to help speed some things up."

Read more of this story at Slashdot.

The Biden administration unveiled a set of far-reaching goals Tuesday aimed at averting harms caused by the rise of artificial intelligence systems, including guidelines for how to protect people's personal data and limit surveillance. From a report: The Blueprint for an AI Bill of Rights notably does not set out specific enforcement actions, but instead is intended as a White House call to action for the U.S. government to safeguard digital and civil rights in an AI-fueled world, officials said. "This is the Biden-Harris administration really saying that we need to work together, not only just across government, but across all sectors, to really put equity at the center and civil rights at the center of the ways that we make and use and govern technologies," said Alondra Nelson, deputy director for science and society at the White House Office of Science and Technology Policy. "We can and should expect better and demand better from our technologies." The office said the white paper represents a major advance in the administration's agenda to hold technology companies accountable, and highlighted various federal agencies' commitments to weighing new rules and studying the specific impacts of AI technologies. The document emerged after a year-long consultation with more than two dozen different departments, and also incorporates feedback from civil society groups, technologists, industry researchers and tech companies including Palantir and Microsoft. It suggests five core principles that the White House says should be built into AI systems to limit the impacts of algorithmic bias, give users control over their data and ensure that automated systems are used safely and transparently.

Read more of this story at Slashdot.

Last week, a number of outlets reported that Bruce Willis sold his face to a deepfake company called Deepcake, allowing a "digital twin" of himself to be created for use on screen. The only problem is that it's apparently not true. According to the BBC, the actor's agent said that he had "no partnership or agreement" with the company and a representative of Deepcake said only Willis had the rights to his face From the report: On 27 September, the Daily Mail reported that a deal had been struck between Willis and Deepcake. "Two-time Emmy winner Bruce Willis can still appear in movies after selling his image rights to Deepcake," the story reads. The story was picked up by the Telegraph and a series of other media outlets. "Bruce Willis has become the first Hollywood star to sell his rights to allow a 'digital twin' of himself to be created for use on screen." said the Telegraph. But that doesn't appear to be the case. What is true is that a deepfake of Bruce Willis was used to create an advert for Megafon, a Russian telecoms company, last year. The tech used in the advert was created by Deepcake, which describes itself as an AI company specializing in deepfakes. Deepcake told the BBC it had worked closely with Willis' team on the advert. "What he definitely did is that he gave us his consent (and a lot of materials) to make his Digital Twin," they said. The company says it has a unique library of high-resolution celebrities, influencers and historical figures. On its website, Deepcake promotes its work with an apparent quote from Mr Willis: "I liked the precision of my character. It's a great opportunity for me to go back in time. "The neural network was trained on content of Die Hard and Fifth Element, so my character is similar to the images of that time." A representative from Deepcake said in a statement: "The wording about rights is wrong... Bruce couldn't sell anyone any rights, they are his by default."

Read more of this story at Slashdot.

The Biden administration is expected to announce new measures to restrict Chinese companies from accessing technologies that enable high-performance computing, The New York Times reported Monday, citing several people familiar with the matter, the latest in a series of moves aimed at hobbling Beijing's ambitions to craft next-generation weapons and automate large-scale surveillance systems. From a report: The measures, which could be announced as soon as this week, would be some of the most significant steps taken by the Biden administration to cut off China's access to advanced semiconductor technology. They would build on a Trump-era rule that struck a blow to the Chinese telecom giant Huawei by prohibiting companies around the world from sending it products made with the use of American technology, machinery or software. A number of Chinese firms, government research labs and other entities are expected to face restrictions similar to Huawei, according to two people with knowledge of the plans. In effect, any firm that uses American-made technologies would be blocked from selling to the Chinese entities that are targeted by the administration. It's not yet clear which Chinese firms and labs would be impacted. The broad expansion of what is known as the foreign direct product rule is just one part of Washington's planned restrictions. The administration is also expected to try to control the sale of cutting-edge U.S.-made tools to China's domestic chip makers.

Read more of this story at Slashdot.

An anonymous reader quotes a report from The Conversation: To detect audio deepfakes, we and our research colleagues at the University of Florida have developed a technique that measures the acoustic and fluid dynamic differences between voice samples created organically by human speakers and those generated synthetically by computers. The first step in differentiating speech produced by humans from speech generated by deepfakes is understanding how to acoustically model the vocal tract. Luckily scientists have techniques to estimate what someone -- or some being such as a dinosaur -- would sound like based on anatomical measurements of its vocal tract. We did the reverse. By inverting many of these same techniques, we were able to extract an approximation of a speaker's vocal tract during a segment of speech. This allowed us to effectively peer into the anatomy of the speaker who created the audio sample. From here, we hypothesized that deepfake audio samples would fail to be constrained by the same anatomical limitations humans have. In other words, the analysis of deepfaked audio samples simulated vocal tract shapes that do not exist in people. Our testing results not only confirmed our hypothesis but revealed something interesting. When extracting vocal tract estimations from deepfake audio, we found that the estimations were often comically incorrect. For instance, it was common for deepfake audio to result in vocal tracts with the same relative diameter and consistency as a drinking straw, in contrast to human vocal tracts, which are much wider and more variable in shape. This realization demonstrates that deepfake audio, even when convincing to human listeners, is far from indistinguishable from human-generated speech. By estimating the anatomy responsible for creating the observed speech, it's possible to identify the whether the audio was generated by a person or a computer.

Read more of this story at Slashdot.

Tesla CEO Elon Musk revealed a prototype of a humanoid robot that he said utilizes the company's AI software, as well as the sensors that power its advanced driver assist features. The Verge reports: The robot was showcased at Tesla's AI Day, and reps said it features the same technology used to enable the Full Self-Driving beta in Tesla's cars. According to Musk, it can do more than what has been shown, but "the first time it walked without a tether was tonight on stage." Musk said they're targeting a price of "probably less than $20,000." The back doors of the stage open to reveal a deconstructed Optimus that walked forward and did a "raise the roof" dance move. Musk would admit after the motion that they wanted to keep it safe and not make too many moves on stage and have it "fall flat on its face." "It'll be a fundamental transformation for civilization as we know it." said Musk. Afterward, the company showed a few video clips of the robot doing other tasks like picking up boxes. Then Tesla's team brought out another prototype that has its body fully assembled but not fully functional. [...] Future applications could include cooking, gardening, or even "catgirl" sex partners, Musk has said, while also claiming that production could start as soon as next year. Musk says the robot is "the most important product development we're doing this year," predicting that it will have the potential to be "more significant than the vehicle business over time." Musk first announced the "Tesla Bot" at last year's AI Day.

Read more of this story at Slashdot.

An anonymous reader quotes a report from TechCrunch: Dubbed the Facial Recognition Act, the bill would compel law enforcement to obtain a judge-authorized warrant before using facial recognition. By adding the warrant requirement, law enforcement would first have to show a court it has probable cause that a person has committed a serious crime, rather than allowing largely unrestricted use of facial recognition under the existing legal regime. The bill also puts other limits on what law enforcement can use facial recognition for, such as immigration enforcement or peaceful protests, or using a facial recognition match as the sole basis for establishing probable cause for someone's arrest. If passed, the bill would also require law enforcement to annually test and audit their facial recognition systems, and provide detailed reports of how facial recognition systems are used in prosecutions. It would also require police departments and agencies to purge databases of photos of children who were subsequently released without charge, whose charges were dismissed or were acquitted. [...] The bill has so far received glowing support from privacy advocates, rights groups and law enforcement-adjacent groups and organizations alike. Woodrow Hartzog, a law professor at Boston University, praised the bill for strengthening baseline rules and protections across the U.S. "without preempting more stringent limitations elsewhere."

Read more of this story at Slashdot.

AI text-to-image generators have been making headlines in recent months, but researchers are already moving on to the next frontier: AI text-to-video generators. From a report: A team of machine learning engineers from Facebook's parent company Meta has unveiled a new system called Make-A-Video. As the name suggests, this AI model allows users to type in a rough description of a scene, and it will generate a short video matching their text. The videos are clearly artificial, with blurred subjects and distorted animation, but still represent a significant development in the field of AI content generation. "Generative AI research is pushing creative expression forward by giving people tools to quickly and easily create new content," said Meta in a blog post announcing the work. "With just a few words or lines of text, Make-A-Video can bring imagination to life and create one-of-a-kind videos full of vivid colors and landscapes." In a Facebook post, Meta CEO Mark Zuckerberg described the work as "amazing progress," adding: "It's much harder to generate video than photos because beyond correctly generating each pixel, the system also has to predict how they'll change over time."

Read more of this story at Slashdot.

Bruce Willis has become the first Hollywood star to sell his rights to allow a "digital twin" of himself to be created for use on screen. From a report: Using deepfake technology, the actor appeared in a phone advert without ever being on set, after his face was digitally transplanted onto another performer. Willis allowed US firm Deepcake, which makes "digital twins," to use his face. In a statement on its website, Willis said: "I liked the precision with which my character turned out. It's a mini-movie in my usual action-comedy genre. For me, it is a great opportunity to go back in time. With the advent of modern technology, even when I was on another continent, I was able to communicate, work and participate in the filming. It's a very new and interesting experience, and I thank our entire team."

Read more of this story at Slashdot.

Elon Musk is widely expected to show off a new humanoid robot Friday at a Tesla artificial intelligence event. From a report: Mr. Musk first laid out the vision for the robot, called Optimus, little more than a year ago at Tesla's first-ever AI day. At the time, a dancer in a costume appeared onstage. This time, Mr. Musk has said he wants a prototype to be at the gathering that is scheduled to unfold from 5 p.m. local time in Palo Alto, Calif. Mr. Musk has painted a vision of Optimus as helping Tesla make cars more efficiently. He has also suggested the robot could serve broader functions and potentially alleviate labor shortages. "My guess is Optimus will be more valuable than the car long term," Mr. Musk said Aug. 4 at Tesla's annual shareholder meeting. "It will, I think, turn the whole notion of what's an economy on its head, at the point at which you have no shortage of labor," he added. When he first unveiled the Optimus concept, Mr. Musk said such a robot could have such an impact on the labor market it could make it necessary to provide a universal basic income, or a stipend to people without strings attached.

Read more of this story at Slashdot.

"First they came for factory jobs. Then they showed up in service industries. Now, machines are making inroads into the kind of white-collar office work once thought to be the exclusive preserve of humans," write Alexandre Tanzi and Reade Pickert via Bloomberg. An anonymous reader shares an excerpt from the report: It's not just corporate giants, capable of spending millions of dollars to develop their own technologies, that are getting in on the act. One feature of the new automation wave is that companies like Kizen have popped up to make it affordable even for smaller firms. Based in Austin, Texas, Kizen markets an automated assistant called Zoe, which can perform tasks for sales teams like carrying out initial research and qualifying leads. Launched a year ago, it's already sold more than 400,000 licenses. "Our smallest customer pays us $10 a month and our largest customer pays us $9.5 million a year,'' says John Winner, Kizen's chief executive officer. There are plenty of other ambitious companies cashing in on the trend, and posting steep increases in revenue -- like UiPath Inc., a favorite of star investment manager Cathie Wood, as well as Appian Corp. and EngageSmart Inc. Alongside the growth of AI and what economists call "robotic process automation" -- essentially, when software performs certain tasks previously done by humans -- old-school automation is still going strong too. The number of robots sold in North America hit a new record in the first quarter of 2022, according to the Association for Advancing Automation. The World Economic Forum predicts that by 2025, machines will be working as many hours as humans. What all of this innovation means for the world's workers is one of the key open questions in economics. The upbeat view says it's tasks that get automated, not entire jobs -- and if the mundane ones can be handled by computers or robots, that should free up employees for more challenging and satisfying work. The downside risk: occupations from sales reps to administrative support, could begin to disappear -- without leaving obvious alternatives for the people who earned a living from them. That adds another employment threat for white-collar workers who may already be vulnerable right now to an economic downturn, largely because so many got hired in the boom of the past couple of years. KC Harvey Environmental, a consultancy based in Bozeman, Montana that works with businesses and governments on environmental issues, is one of Kizen's clients. It uses the software to automate document control -- for example, archiving and delivering new contracts to the right places and people. "A new project probably took our accounting group and project management team a day," says Rio Franzman, KC Harvey's chief operating officer. "This now probably streamlines it down to about an hour." The firm employs about 100 people and "we didn't lose any'' as a result of automation, he says. "What it did allow is for the reallocation of time and resources to more meaningful tasks." KC Harvey is now working with Kizen to bring AI into its marketing, too, with a partly automated newsletter among other projects. Some of the biggest firms at the forefront of automation also say they've been able to do it without cutting jobs. Engineering giant Siemens AG says it's automated all kinds of production and back-office tasks at its innovative plant in Amberg, Germany, where it makes industrial computers, while keeping staffing steady at around 1,350 employees over several decades. The firm has developed a technology known as "digital twinning," which builds virtual versions of everything from specific products to administrative processes. Managers can then run simulations and stress-tests to see how things can be made better. "We're not going to automate people out of the process," says Barbara Humpton, CEO of Siemens USA. "By optimizing automation systems, and by using digital tools and AI, workers have increased productivity at Amberg by more than 1,000%." [...] Whatever the outcome, it's unlikely to allay the deep unease that the idea of automation triggers among workers who feel their jobs are vulnerable. With the rise of AI, that group increasingly includes white-collar employees.

Read more of this story at Slashdot.

An anonymous reader quotes a report from Motherboard: Everyone wants to be able to just zap a bug and have it go away. But now, thanks to a recent development from Ildar Rakhmatulin, a research associate at Heriot-Watt University interested in machine learning and engineering, this dream is now a reality. In the study -- which was conducted last year but published in Oriental Insects last week -- Rakhmatulin and his co-authors used a laser insect control device automated with machine vision to perform a series of experiments on domiciliary cockroaches. They were able to not only detect cockroaches at high accuracy but also neutralize and deter individual insects at a distance up to 1.2 meters. This is a follow-up of sorts to earlier projects, in which he used a Raspberry Pi and lasers to zap mosquitoes. However, for this project, Rakhmatulin used a different kind of computer which allowed for more precision in detecting the bug. "I started using a Jetson Nano that allowed me to use deep learning technologies with higher accuracy to detect an object," Rakhmatulin explained. The Jetson Nano is a small computer that can run machine learning algorithms. The computer processes a digital signal from two cameras to determine the cockroach's position. It transmits that information to a galvanometer (a machine that measures electric current), which changes the direction of the laser to shoot the target. According to the paper, Rakhmatulin tried this configuration at different power levels for the laser. At a lower power level, he found that he could influence the behavior of roaches by simply triggering their flight response with a laser; this way, they could potentially be trained to not shelter in a particular dark area. At a higher power level, the cockroaches were effectively "neutralized," in the paper's language -- in other words, killed. "I use very cheap hardware and cheap technology and it's open source," Rakhmatulin said. "All sources are uploaded in my GitHub and see how to do it and use it. If it can damage cockroaches, it can also damage other pests in agriculture." It's not quite ready for household use though. "It's not recommended because it's a little dangerous," Rakhmatulin said. "Lasers can damage not only cockroaches but your eyes." You can view a video of the device in action here.

Read more of this story at Slashdot.

An anonymous reader shares a report:Since the research lab OpenAI debuted the latest version of DALL-E in April, the AI has dazzled the public, attracting digital artists, graphic designers, early adopters, and anyone in search of online distraction. The ability to create original, sometimes accurate, and occasionally inspired images from any spur-of-the-moment phrase, like a conversational Photoshop, has startled even jaded internet users with how quickly AI has progressed. Five months later, 1.5 million users are generating 2 million images a day. On Wednesday, OpenAI said it will remove its waitlist for DALL-E, giving anyone immediate access. The introduction of DALL-E has triggered an explosion of text-to-image generators. Google and Meta quickly revealed that they had each been developing similar systems, but said their models weren't ready for the public. Rival start-ups soon went public, including Stable Diffusion and Midjourney, which created the image that sparked controversy in August when it won an art competition at the Colorado State Fair.

Read more of this story at Slashdot.

An anonymous reader quotes a report from Smithsonian Magazine: It's an increasingly common sight on vacation, particularly in tourist destinations: An influencer sets up in front of a popular local landmark, sometimes even using props (coffee, beer, pets) or changing outfits, as a photographer or self-timed camera snaps away. Others are milling around, sometimes watching. But often, unbeknownst to everyone involved, another device is also recording the scene: a surveillance camera. Belgian artist Dries Depoorter is exploring this dynamic in his controversial new online exhibit, The Followers, which he unveiled last week. The art project places static Instagram images side-by-side with video from surveillance cameras, which recorded footage of the photoshoot in question. To make The Followers, Depoorter started with EarthCam, a network of publicly accessible webcams around the world, to record a month's worth of footage in tourist attractions like New York City's Times Square and Dublin's Temple Bar Pub. Then he enlisted an artificial intelligence (A.I.) bot, which scraped public Instagram photos taken in those locations, and facial-recognition software, which paired the Instagram images with the real-time surveillance footage. Depoorter calls himself a "surveillance artist," and this isn't his first project using open-source webcam footage or A.I. Last year, for a project called The Flemish Scrollers, he paired livestream video of Belgian government proceedings with an A.I. bot he built to determine how often lawmakers were scrolling on their phones during official meetings. "On its face, The Followers is an attempt, like many other studies, art projects and documentaries in recent years, to expose the staged, often unattainable ideals shown in many Instagram and influencer photos posted online," writes Smithsonian's Molly Enking. "But The Followers also tells a darker story: one of increasingly worrisome privacy concerns amid an ever-growing network of surveillance technology in public spaces. And the project, as well as the techniques used to create it, has sparked both ethical and legal controversy." Depoorter told Vice's Samantha Cole that he got the idea when he "watched an open camera and someone was taking pictures for like 30 minutes." He wondered if he'd be able to find that person on Instagram.

Read more of this story at Slashdot.

T-Mobile has partnered with the startup Pano AI to spot wildfires before they get out of control. CNET reports: The startup Pano AI uses a series of cameras that survey the wilderness and AI algorithms that watch for telltale smoke -- an indicator of small blazes that could grow into raging wildfires. That footage is sent to the startup's headquarters for human confirmation, and if a fire is burning, evidence is sent to clients who could be affected. While Pano AI had been sending evidence photos over 4G LTE networks at slow rates of around 20 to 30 6-megapixel images per minute, its new partnership with T-Mobile has it using the carrier's 5G network to send video at 30 frames per second, which is around 90 times more data. Ultimately, getting evidence to Pano AI's clients, which include utility companies, much quicker on 5G means a faster response from firefighters and potentially squashing big fires before they get dangerous. Pano AI works with a number of utilities, governments, fire authorities, forestry companies and private landlords who in turn work with local emergency responders. Its newest client and the first with a system using T-Mobile's 5G network is Portland General Electric (PGE), a utility supplying gas and electricity to 16 million customers around Portland, Oregon. Pano AI has 20 cameras set up in the forests surrounding the city that give 10-mile panoramic views, which include powerlines. This lets PGE know if fires are headed toward its infrastructure. T-Mobile recruited Pano AI to be part of its Innovation Lab alongside other companies harnessing 5G to improve their services, such as Mixhalo, which is using the carrier's 5G network to pipe in concert audio directly to audience members' phones. But Pano AI's partnership goes deeper, as it's mounting its cameras on T-Mobile's cell towers, saving months of time and paperwork needed to request and install its equipment on other signal towers or similar vantage points.

Read more of this story at Slashdot.

An anonymous reader quotes a report from TechCrunch: Deepfakes, AI generated porn, and a thousand more innocent uses -- there's been a lot of news about neural network-generated images. It makes sense that people started getting curious; were my photos used to train the robots? Are photos of me in the image-generating training sets? A brand new site tries to give you an answer. Spawning AI creates image generation tools for artists, and the company just launched Have I Been Trained? which you can use to search a set of 5.8 billion images that have been used to train popular AI art models. When you search the site, you can search through the images that are the closest match, based on the LAION-5B training data, which is widely used for training AI search terms. It's a fun tool to play with, and may help give a glimpse into the data that the AI is using as the basis for its own.

Read more of this story at Slashdot.

Getty Images has banned the upload and sale of illustrations generated using AI art tools like DALL-E, Midjourney, and Stable Diffusion. From a report: It's the latest and largest user-generated content platform to introduce such a ban, following similar decisions by sites including Newgrounds, PurplePort, and FurAffinity. Getty Images CEO Craig Peters told The Verge that the ban was prompted by concerns about the legality of AI-generated content and a desire to protect the site's customers. "There are real concerns with respect to the copyright of outputs from these models and unaddressed rights issues with respect to the imagery, the image metadata and those individuals contained within the imagery," said Peters. Given these concerns, he said, selling AI artwork or illustrations could potentially put Getty Images users at legal risk. "We are being proactive to the benefit of our customers," he added. One of Getty Images' biggest competitors, Shutterstock, also seems to be limiting some searches for AI content but hasn't yet introduced specific policies banning the material.

Read more of this story at Slashdot.

Shutterstock appears to be removing images generated by AI systems like DALL-E and Midjourney. Motherboard reports: On Shutterstock, searches for images tagged "Midjourney" yielded several photos with the AI tool's unmistakable aesthetic, with many having high popularity scores and marked as "frequently used." But late Monday, the results for "Midjourney" seem to have been reduced, leaving mainly stock photos of the tool's logo. Other images use tags like "AI generated" -- one image, for example, is an illustration of a futuristic building with an image description reading "Ai generated illustration of futuristic Art Deco city, vintage image, retro poster." The image is part of a collection the artist titled "Midjourney," which has since been removed from the site. Other images marked "AI generated," like this burning medieval castle, seem to remain up on the site. As Ars Technica notes, neither Shutterstock nor Getty Images explicitly prohibits AI-generated images in their terms of service, and Shutterstock users typically make around 15 to 40 percent of what the company makes when it sells an image. Some creators have not taken kindly to this trend, pointing out that these systems use massive datasets of images scraped from the web. [...] In other words, the generated works are the result of an algorithmic process which mines original art from the internet without credit or compensation to the original artists. Others have worried about the impacts on independent artists who work for commissions, since the ability for anyone to create custom generated artwork potentially means lost revenue.

Read more of this story at Slashdot.

Shady scientists trying to publish bad research may want to think twice as academic publishers are increasingly using AI software to automatically spot signs of data tampering. The Register: Duplications of images, where the same picture of a cluster of cells, for example, is copied, flipped, rotated, shifted, or cropped is, unfortunately, quite common. In cases where the errors aren't accidental, the doctored images are created to look as if the researchers have more data and conducted more experiments then they really did. Image duplication was the top reason papers were retracted for the American Association for Cancer Research (AACR) over 2016 to 2020, according to Daniel Evanko, the company's Director of Journal Operations and Systems. Having to retract a paper damages the authors and the publishers' reputation. It shows that the quality of work from the researchers was poor, and the editor's peer review process missed mistakes. To prevent embarrassment for both parties, academic publishers like AACR have turned to AI software to detect image duplication before a paper is published in a journal. The AACR started trialling Proofig, an image-checking programme developed by a startup going by the same name as their product based in Israel. Evanko presented results from the pilot study to show how Proofig impacted AACR's operations at the International Congress on Peer Review and Scientific Publication conference held in Chicago this week. AACR publishes ten research journals and reviews over 13,000 submissions every year. From January 2021 to May 2022, officials used Proofig to screen 1,367 manuscripts that had been provisionally accepted for publication and contacted authors in 208 cases after reviewing image duplicates flagged by the software. In most cases, the duplication is a sloppy error that can be fixed easily. Scientists may have accidentally got their results mixed up and the issue is often resolved by resubmitting new data. On rare occasions, however, the dodgy images highlighted by the software are a sign of foul play.

Read more of this story at Slashdot.