AI – Page 27

AI trains on kids’ photos even when parents use strict privacy settings

AI, ai deepfakes, AI training, Artificial Intelligence, australia, child online safety, children's online privacy, generative ai, Human rights, image generators, large language models, Policy, youtube / Kelly Newman / July 2, 2024

“Outrageous” —

Even unlisted YouTube videos are used to train AI, watchdog warns.

Ashley Belanger – Jul 2, 2024 7: 37 pm UTC

Human Rights Watch (HRW) continues to reveal how photos of real children casually posted online years ago are being used to train AI models powering image generators—even when platforms prohibit scraping and families use strict privacy settings.

Last month, HRW researcher Hye Jung Han found 170 photos of Brazilian kids that were linked in LAION-5B, a popular AI dataset built from Common Crawl snapshots of the public web. Now, she has released a second report, flagging 190 photos of children from all of Australia’s states and territories, including indigenous children who may be particularly vulnerable to harms.

These photos are linked in the dataset “without the knowledge or consent of the children or their families.” They span the entirety of childhood, making it possible for AI image generators to generate realistic deepfakes of real Australian children, Han’s report said. Perhaps even more concerning, the URLs in the dataset sometimes reveal identifying information about children, including their names and locations where photos were shot, making it easy to track down children whose images might not otherwise be discoverable online.

That puts children in danger of privacy and safety risks, Han said, and some parents thinking they’ve protected their kids’ privacy online may not realize that these risks exist.

From a single link to one photo that showed “two boys, ages 3 and 4, grinning from ear to ear as they hold paintbrushes in front of a colorful mural,” Han could trace “both children’s full names and ages, and the name of the preschool they attend in Perth, in Western Australia.” And perhaps most disturbingly, “information about these children does not appear to exist anywhere else on the Internet”—suggesting that families were particularly cautious in shielding these boys’ identities online.

Stricter privacy settings were used in another image that Han found linked in the dataset. The photo showed “a close-up of two boys making funny faces, captured from a video posted on YouTube of teenagers celebrating” during the week after their final exams, Han reported. Whoever posted that YouTube video adjusted privacy settings so that it would be “unlisted” and would not appear in searches.

Only someone with a link to the video was supposed to have access, but that didn’t stop Common Crawl from archiving the image, nor did YouTube policies prohibiting AI scraping or harvesting of identifying information.

Reached for comment, YouTube’s spokesperson, Jack Malon, told Ars that YouTube has “been clear that the unauthorized scraping of YouTube content is a violation of our Terms of Service, and we continue to take action against this type of abuse.” But Han worries that even if YouTube did join efforts to remove images of children from the dataset, the damage has been done, since AI tools have already trained on them. That’s why—even more than parents need tech companies to up their game blocking AI training—kids need regulators to intervene and stop training before it happens, Han’s report said.

Han’s report comes a month before Australia is expected to release a reformed draft of the country’s Privacy Act. Those reforms include a draft of Australia’s first child data protection law, known as the Children’s Online Privacy Code, but Han told Ars that even people involved in long-running discussions about reforms aren’t “actually sure how much the government is going to announce in August.”

“Children in Australia are waiting with bated breath to see if the government will adopt protections for them,” Han said, emphasizing in her report that “children should not have to live in fear that their photos might be stolen and weaponized against them.”

AI uniquely harms Australian kids

To hunt down the photos of Australian kids, Han “reviewed fewer than 0.0001 percent of the 5.85 billion images and captions contained in the data set.” Because her sample was so small, Han expects that her findings represent a significant undercount of how many children could be impacted by the AI scraping.

“It’s astonishing that out of a random sample size of about 5,000 photos, I immediately fell into 190 photos of Australian children,” Han told Ars. “You would expect that there would be more photos of cats than there are personal photos of children,” since LAION-5B is a “reflection of the entire Internet.”

LAION is working with HRW to remove links to all the images flagged, but cleaning up the dataset does not seem to be a fast process. Han told Ars that based on her most recent exchange with the German nonprofit, LAION had not yet removed links to photos of Brazilian kids that she reported a month ago.

LAION declined Ars’ request for comment.

In June, LAION’s spokesperson, Nathan Tyler, told Ars that, “as a nonprofit, volunteer organization,” LAION is committed to doing its part to help with the “larger and very concerning issue” of misuse of children’s data online. But removing links from the LAION-5B dataset does not remove the images online, Tyler noted, where they can still be referenced and used in other AI datasets, particularly those relying on Common Crawl. And Han pointed out that removing the links from the dataset doesn’t change AI models that have already trained on them.

“Current AI models cannot forget data they were trained on, even if the data was later removed from the training data set,” Han’s report said.

Kids whose images are used to train AI models are exposed to a variety of harms, Han reported, including a risk that image generators could more convincingly create harmful or explicit deepfakes. In Australia last month, “about 50 girls from Melbourne reported that photos from their social media profiles were taken and manipulated using AI to create sexually explicit deepfakes of them, which were then circulated online,” Han reported.

For First Nations children—”including those identified in captions as being from the Anangu, Arrernte, Pitjantjatjara, Pintupi, Tiwi, and Warlpiri peoples”—the inclusion of links to photos threatens unique harms. Because culturally, First Nations peoples “restrict the reproduction of photos of deceased people during periods of mourning,” Han said the AI training could perpetuate harms by making it harder to control when images are reproduced.

Once an AI model trains on the images, there are other obvious privacy risks, including a concern that AI models are “notorious for leaking private information,” Han said. Guardrails added to image generators do not always prevent these leaks, with some tools “repeatedly broken,” Han reported.

LAION recommends that, if troubled by the privacy risks, parents remove images of kids online as the most effective way to prevent abuse. But Han told Ars that’s “not just unrealistic, but frankly, outrageous.”

“The answer is not to call for children and parents to remove wonderful photos of kids online,” Han said. “The call should be [for] some sort of legal protections for these photos, so that kids don’t have to always wonder if their selfie is going to be abused.”

AI trains on kids’ photos even when parents use strict privacy settings Read More »

Google’s greenhouse gas emissions jump 48% in five years

AI, data centers, Google, greenhouse gas emissions, syndication, Tech / Kelly Newman / July 2, 2024

computationally intensive means energy intensive —

Google’s 2030 “Net zero” target looks increasingly doubtful as AI use soars.

Camilla Hodgson and Stephen Morris, Financial Times – Jul 2, 2024 5: 25 pm UTC

Cooling pipes at a Google data center in Douglas County, Georgia.

Google’s greenhouse gas emissions have surged 48 percent in the past five years due to the expansion of its data centers that underpin artificial intelligence systems, leaving its commitment to get to “net zero” by 2030 in doubt.

The Silicon Valley company’s pollution amounted to 14.3 million tonnes of carbon equivalent in 2023, a 48 percent increase from its 2019 baseline and a 13 percent rise since last year, Google said in its annual environmental report on Tuesday.

Google said the jump highlighted “the challenge of reducing emissions” at the same time as it invests in the build-out of large language models and their associated applications and infrastructure, admitting that “the future environmental impact of AI” was “complex and difficult to predict.”

Chief Sustainability Officer Kate Brandt said the company remained committed to the 2030 target but stressed the “extremely ambitious” nature of the goal.

“We do still expect our emissions to continue to rise before dropping towards our goal,” said Brandt.

She added that Google was “working very hard” on reducing its emissions, including by signing deals for clean energy. There was also a “tremendous opportunity for climate solutions that are enabled by AI,” said Brandt.

As Big Tech giants including Google, Amazon, and Microsoft have outlined plans to invest tens of billions of dollars into AI, climate experts have raised concerns about the environmental impacts of the power-intensive tools and systems.

In May, Microsoft admitted that its emissions had risen by almost a third since 2020, in large part due to the construction of data centers. However, Microsoft co-founder Bill Gates last week also argued that AI would help propel climate solutions.

Meanwhile, energy generation and transmission constraints are already posing a challenge for the companies seeking to build out the new technology. Analysts at Bernstein said in June that AI would “double the rate of US electricity demand growth and total consumption could outstrip current supply in the next two years.”

In Tuesday’s report, Google said its 2023 energy-related emissions—which come primarily from data center electricity consumption—rose 37 percent year on year and overall represented a quarter of its total greenhouse gas emissions.

Google’s supply chain emissions—its largest chunk, representing 75 percent of its total emissions—also rose 8 percent. Google said they would “continue to rise in the near term” as a result in part of the build-out of the infrastructure needed to run AI systems.

Google has pledged to achieve net zero across its direct and indirect greenhouse gas emissions by 2030 and to run on carbon-free energy during every hour of every day within each grid it operates by the same date.

However, the company warned in Tuesday’s report that the “termination” of some clean energy projects during 2023 had pushed down the amount of renewables it had access to.

Meanwhile, the company’s data center electricity consumption had “outpaced” Google’s ability to bring more clean power projects online in the US and Asia-Pacific regions.

Google’s data center electricity consumption increased 17 percent in 2023, and amounted to approximately 7-10 percent of global data center electricity consumption, the company estimated. Its data centers also consumed 17 percent more water in 2023 than during the previous year, Google said.

Google’s greenhouse gas emissions jump 48% in five years Read More »

Lightening the load: AI helps exoskeleton work with different strides

AI, Computer science, exoskeleton, https://doi.org/10.1038/s41586-024-07382-4, robotics, Science, Software / Shannon Garcia / July 1, 2024

One model to rule them all —

A model trained in a virtual environment does remarkably well in the real world.

Jacek Krywko – Jul 1, 2024 5: 31 pm UTC

Image of two people using powered exoskeletons to move heavy items around, as seen in the movie Aliens. — Enlarge / Right now, the software doesn’t do arms, so don’t go taking on any aliens with it.

20th Century Fox

Exoskeletons today look like something straight out of sci-fi. But the reality is they are nowhere near as robust as their fictional counterparts. They’re quite wobbly, and it takes long hours of handcrafting software policies, which regulate how they work—a process that has to be repeated for each individual user.

To bring the technology a bit closer to Avatar’s Skel Suits or Warhammer 40k power armor, a team at North Carolina University’s Lab of Biomechatronics and Intelligent Robotics used AI to build the first one-size-fits-all exoskeleton that supports walking, running, and stair-climbing. Critically, its software adapts itself to new users with no need for any user-specific adjustments. “You just wear it and it works,” says Hao Su, an associate professor and co-author of the study.

Tailor-made robots

An exoskeleton is a robot you wear to aid your movements—it makes walking, running, and other activities less taxing, the same way an e-bike adds extra watts on top of those you generate yourself, making pedaling easier. “The problem is, exoskeletons have a hard time understanding human intentions, whether you want to run or walk or climb stairs. It’s solved with locomotion recognition: systems that recognize human locomotion intentions,” says Su.

Building those locomotion recognition systems currently relies on elaborate policies that define what actuators in an exoskeleton need to do in each possible scenario. “Let’s take walking. The current state of the art is we put the exoskeleton on you and you walk on a treadmill for an hour. Based on that, we try to adjust its operation to your individual set of movements,” Su explains.

Building handcrafted control policies and doing long human trials for each user makes exoskeletons super expensive, with prices reaching $200,000 or more. So, Su’s team used AI to automatically generate control policies and eliminate human training. “I think within two or three years, exoskeletons priced between $2,000 and $5,000 will be absolutely doable,” Su claims.

His team hopes these savings will come from developing the exoskeleton control policy using a digital model, rather than living, breathing humans.

Digitizing robo-aided humans

Su’s team started by building digital models of a human musculoskeletal system and an exoskeleton robot. Then they used multiple neural networks that operated each component. One was running the digitized model of a human skeleton, moved by simplified muscles. The second neural network was running the exoskeleton model. Finally, the third neural net was responsible for imitating motion—basically predicting how a human model would move wearing the exoskeleton and how the two would interact with each other. “We trained all three neural networks simultaneously to minimize muscle activity,” says Su.

One problem the team faced is that exoskeleton studies typically use a performance metric based on metabolic rate reduction. “Humans, though, are incredibly complex, and it is very hard to build a model with enough fidelity to accurately simulate metabolism,” Su explains. Luckily, according to the team, reducing muscle activations is rather tightly correlated with metabolic rate reduction, so it kept the digital model’s complexity within reasonable limits. The training of the entire human-exoskeleton system with all three neural networks took roughly eight hours on a single RTX 3090 GPU. And the results were record-breaking.

Bridging the sim-to-real gap

After developing the controllers for the digital exoskeleton model, which were developed by the neural networks in simulation, Su’s team simply copy-pasted the control policy to a real controller running a real exoskeleton. Then, they tested how an exoskeleton trained this way would work with 20 different participants. The averaged metabolic rate reduction in walking was over 24 percent, over 13 percent in running, and 15.4 percent in stair climbing—all record numbers, meaning their exoskeleton beat every other exoskeleton ever made in each category.

This was achieved without needing any tweaks to fit it to individual gaits. But the neural networks’ magic didn’t end there.

“The problem with traditional, handcrafted policies was that it was just telling it ‘if walking is detected do one thing; if walking faster is detected do another thing.’ These were [a mix of] finite state machines and switch controllers. We introduced end-to-end continuous control,” says Su. What this continuous control meant was that the exoskeleton could follow the human body as it made smooth transitions between different activities—from walking to running, from running to climbing stairs, etc. There was no abrupt mode switching.

“In terms of software, I think everyone will be using this neural network-based approach soon,” Su claims. To improve the exoskeletons in the future, his team wants to make them quieter, lighter, and more comfortable.

But the plan is also to make them work for people who need them the most. “The limitation now is that we tested these exoskeletons with able-bodied participants, not people with gait impairments. So, what we want to do is something they did in another exoskeleton study at Stanford University. We would take a one-minute video of you walking, and based on that, we would build a model to individualize our general model. This should work well for people with impairments like knee arthritis,” Su claims.

Nature, 2024. DOI: 10.1038/s41586-024-07382-4

Lightening the load: AI helps exoskeleton work with different strides Read More »

The telltale words that could identify generative AI text

AI / Kelly Newman / July 1, 2024

Delving deep —

New paper counts “excess words” that started appearing more often in the post-LLM era.

Kyle Orland – Jul 1, 2024 11: 30 am UTC

Enlarge / If your right hand starts typing “delve,” you may, in fact, be an LLM.

Getty Images

Thus far, even AI companies have had trouble coming up with tools that can reliably detect when a piece of writing was generated using a large language model. Now, a group of researchers has established a novel method for estimating LLM usage across a large set of scientific writing by measuring which “excess words” started showing up much more frequently during the LLM era (i.e., 2023 and 2024). The results “suggest that at least 10% of 2024 abstracts were processed with LLMs,” according to the researchers.

In a pre-print paper posted earlier this month, four researchers from Germany’s University of Tubingen and Northwestern University said they were inspired by studies that measured the impact of the COVID-19 pandemic by looking at excess deaths compared to the recent past. By taking a similar look at “excess word usage” after LLM writing tools became widely available in late 2022, the researchers found that “the appearance of LLMs led to an abrupt increase in the frequency of certain style words” that was “unprecedented in both quality and quantity.”

Delving in

To measure these vocabulary changes, the researchers analyzed 14 million paper abstracts published on PubMed between 2010 and 2024, tracking the relative frequency of each word as it appeared across each year. They then compared the expected frequency of those words (based on the pre-2023 trendline) to the actual frequency of those words in abstracts from 2023 and 2024, when LLMs were in widespread use.

The results found a number of words that were extremely uncommon in these scientific abstracts before 2023 that suddenly surged in popularity after LLMs were introduced. The word “delves,” for instance, shows up in 25 times as many 2024 papers as the pre-LLM trend would expect; words like “showcasing” and “underscores” increased in usage by nine times as well. Other previously common words became notably more common in post-LLM abstracts: the frequency of “potential” increased 4.1 percentage points; “findings” by 2.7 percentage points; and “crucial” by 2.6 percentage points, for instance.

Enlarge / Some examples of words that saw their use increase (or decrease) substantially after LLMs were introduced (bottom three words shown for comparison).

These kinds of changes in word use could happen independently of LLM usage, of course—the natural evolution of language means words sometimes go in and out of style. However, the researchers found that, in the pre-LLM era, such massive and sudden year-over-year increases were only seen for words related to major world health events: “ebola” in 2015; “zika” in 2017; and words like “coronavirus,” “lockdown” and “pandemic” in the 2020 to 2022 period.

In the post-LLM period, though, the researchers found hundreds of words with sudden, pronounced increases in scientific usage that had no common link to world events. In fact, while the excess words during the COVID pandemic were overwhelmingly nouns, the researchers found that the words with a post-LLM frequency bump were overwhelmingly “style words” like verbs, adjectives, and adverbs (a small sampling: “across, additionally, comprehensive, crucial, enhancing, exhibited, insights, notably, particularly, within”).

This isn’t a completely new finding—the increased prevalence of “delve” in scientific papers has been widely noted in the recent past, for instance. But previous studies generally relied on comparisons with “ground truth” human writing samples or lists of pre-defined LLM markers obtained from outside the study. Here, the pre-2023 set of abstracts acts as its own effective control group to show how vocabulary choice has changed overall in the post-LLM era.

An intricate interplay

By highlighting hundreds of so-called “marker words” that became significantly more common in the post-LLM era, the telltale signs of LLM use can sometimes be easy to pick out. Take this example abstract line called out by the researchers, with the marker words highlighted: “A comprehensive grasp of the intricate interplay between […] and […] is pivotal for effective therapeutic strategies.”

After doing some statistical measures of marker word appearance across individual papers, the researchers estimate that at least 10 percent of the post-2022 papers in the PubMed corpus were written with at least some LLM assistance. The number could be even higher, the researchers say, because their set could be missing LLM-assisted abstracts that don’t include any of the marker words they identified.

Enlarge / Before 2023, it took a major world event like the coronavirus pandemic to see large jumps in word usage like this.

Those measured percentages can vary a lot across different subsets of papers, too. The researchers found that papers authored in countries like China, South Korea, and Taiwan showed LLM marker words 15 percent of the time, suggesting “LLMs might… help non-natives with editing English texts, which could justify their extensive use.” On the other hand, the researchers offer that native English speakers “may [just] be better at noticing and actively removing unnatural style words from LLM outputs,” thus hiding their LLM usage from this kind of analysis.

Detecting LLM use is important, the researchers note, because “LLMs are infamous for making up references, providing inaccurate summaries, and making false claims that sound authoritative and convincing.” But as knowledge of LLMs’ telltale marker words starts to spread, human editors may get better at taking those words out of generated text before it’s shared with the world.

Who knows, maybe future large language models will do this kind of frequency analysis themselves, lowering the weight of marker words to better mask their outputs as human-like. Before long, we may need to call in some Blade Runners to pick out the generative AI text hiding in our midst.

The telltale words that could identify generative AI text Read More »

ChatGPT outperforms undergrads in intro-level courses, falls short later

AI, college, exams, grades / Kelly Newman / June 29, 2024

Overhead view of a classroom full of students at desks, focused on writing on papers.

“Since the rise of large language models like ChatGPT there have been lots of anecdotal reports about students submitting AI-generated work as their exam assignments and getting good grades. So, we stress-tested our university’s examination system against AI cheating in a controlled experiment,” says Peter Scarfe, a researcher at the School of Psychology and Clinical Language Sciences at the University of Reading.

His team created over 30 fake psychology student accounts and used them to submit ChatGPT-4-produced answers to examination questions. The anecdotal reports were true—the AI use went largely undetected, and, on average, ChatGPT scored better than human students.

Rules of engagement

Scarfe’s team submitted AI-generated work in five undergraduate modules, covering classes needed during all three years of study for a bachelor’s degree in psychology. The assignments were either 200-word answers to short questions or more elaborate essays, roughly 1,500 words long. “The markers of the exams didn’t know about the experiment. In a way, participants in the study didn’t know they were participating in the study, but we’ve got necessary permissions to go ahead with that”, Scarfe claims.

Shorter submissions were prepared simply by copy-pasting the examination questions into ChatGPT-4 along with a prompt to keep the answer under 160 words. The essays were solicited the same way, but the required word count was increased to 2,000. Setting the limits this way, Scarfe’s team could get ChatGPT-4 to produce content close enough to the required length. “The idea was to submit those answers without any editing at all, apart from the essays, where we applied minimal formatting,” says Scarfe.

Overall, Scarfe and his colleagues slipped 63 AI-generated submissions into the examination system. Even with no editing or efforts to hide the AI usage, 94 percent of those went undetected, and nearly 84 percent got better grades (roughly half a grade better) than a randomly selected group of students who took the same exam.

“We did a series of debriefing meetings with people marking those exams and they were quite surprised,” says Scarfe. Part of the reason they were surprised was that most of those AI submissions that were detected did not end up flagged because they were too repetitive or robotic—they got flagged because they were too good.

Which raises a question: What do we do about it?

AI-hunting software

“During this study we did a lot of research into techniques of detecting AI-generated content,” Scarfe says. One such tool is Open AI’s GPTZero; others include AI writing detection systems like the one made by Turnitin, a company specializing in delivering tools for detecting plagiarism.

“The issue with such tools is that they usually perform well in a lab, but their performance drops significantly in the real world,” Scarfe explained. Open AI claims the GPTZero can flag AI-generated text as “likely” AI 26 percent of the time, with a rather worrisome 9 percent false positive rate. Turnitin’s system, on the other hand, was advertised as detecting 97 percent of ChatGPT and GPT-3 authored writing in a lab with only one false positive in a hundred attempts. But, according to Scarfe’s team, the released beta version of this system performed significantly worse.

ChatGPT outperforms undergrads in intro-level courses, falls short later Read More »

Brussels explores antitrust probe into Microsoft’s partnership with OpenAI

AI, antitrust, brussels, European Union, microsoft, microsoft openai, openai, Policy, syndication / Kelly Newman / June 28, 2024

still asking questions —

EU executive arm drops merger review into US tech companies’ alliance.

Javier Espinoza and Tim Bradshaw, Financial Times – Jun 28, 2024 5: 03 pm UTC

Enlarge / EU competition chief Margrethe Vestager said the bloc was looking into practices that could in effect lead to a company controlling a greater share of the AI market.

Brussels is preparing for an antitrust investigation into Microsoft’s $13 billion investment into OpenAI, after the European Union decided not to proceed with a merger review into the most powerful alliance in the artificial intelligence industry.

The European Commission, the EU’s executive arm, began to explore a review under merger control rules in January, but on Friday announced that it would not proceed due to a lack of evidence that Microsoft controls OpenAI.

However, the commission said it was now exploring the possibility of a traditional antitrust investigation into whether the tie-up between the world’s most valuable listed company and the best-funded AI start-up was harming competition in the fast-growing market.

The commission has also made inquiries about Google’s deal with Samsung to install a modified version of its Gemini AI system in the South Korean manufacturer’s smartphones, it revealed on Friday.

Margrethe Vestager, the bloc’s competition chief, said in a speech on Friday: “The key question was whether Microsoft had acquired control on a lasting basis over OpenAI. After a thorough review we concluded that such was not the case. So we are closing this chapter, but the story is not over.”

She said the EU had sent a new set of questions to understand whether “certain exclusivity clauses” in the agreement between Microsoft and OpenAI “could have a negative effect on competitors.” The move is seen as a key step toward a formal antitrust probe.

The bloc had already sent questions to Microsoft and other tech companies in March to determine whether market concentration in AI could potentially block new companies from entering the market, Vestager said.

Microsoft said: “We appreciate the European Commission’s thorough review and its conclusion that Microsoft’s investment and partnership with OpenAI does not give Microsoft control over the company.”

Brussels began examining Microsoft’s relationship with the ChatGPT maker after OpenAI’s board abruptly dismissed its chief executive Sam Altman in November 2023, only to be rehired a few days later. He briefly joined Microsoft as the head of a new AI research unit, highlighting the close relationship between the two companies.

Regulators in the US and UK are also scrutinizing the alliance. Microsoft is the biggest backer of OpenAI, although its investment of up to $13 billion, which was expanded in January 2023, does not involve acquiring conventional equity due to the startup’s unusual corporate structure. Microsoft has a minority interest in OpenAI’s commercial subsidiary, which is owned by a not-for-profit organization.

Antitrust investigations tend to last years, compared with a much shorter period for merger reviews, and they focus on conduct that could be undermining rivals. Companies that are eventually found to be breaking the law, for example by bundling products or blocking competitors from access to key technology, risk hefty fines and legal obligations to change their behavior.

Vestager said the EU was looking into practices that could in effect lead to a company controlling a greater share of the AI market. She pointed to a practice called “acqui-hires,” where a company buys another one mainly to get its talent. For example, Microsoft recently struck a deal to hire most of the top team from AI start-up Inflection, in which it had previously invested. Inflection remains an independent company, however, complicating any traditional merger investigation.

The EU’s competition chief said regulators were also looking into the way big tech companies may be preventing smaller AI models from reaching users.

“This is why we are also sending requests for information to better understand the effects of Google’s arrangement with Samsung to pre-install its small model ‘Gemini nano’ on certain Samsung devices,” said Vestager.

Jonathan Kanter, the top US antitrust enforcer, told the Financial Times earlier this month that he was also examining “monopoly choke points and the competitive landscape” in AI. The UK’s Competition and Markets Authority said in December that it had “decided to investigate” the Microsoft-OpenAI deal when it invited comments from customers and rivals.

Brussels explores antitrust probe into Microsoft’s partnership with OpenAI Read More »

Researchers craft smiling robot face from living human skin cells

AI, Biz & IT, foreskin, Harvard, human skin, human skin cells, PromoCell, robot, Science, skin cells, slimy, Tech, University of Tokyo / Kelly Newman / June 28, 2024

Enlarge / A movable robotic face covered with living human skin cells.

In a new study, researchers from the University of Tokyo, Harvard University, and the International Research Center for Neurointelligence have unveiled a technique for creating lifelike robotic skin using living human cells. As a proof of concept, the team engineered a small robotic face capable of smiling, covered entirely with a layer of pink living tissue.

The researchers note that using living skin tissue as a robot covering has benefits, as it’s flexible enough to convey emotions and can potentially repair itself. “As the role of robots continues to evolve, the materials used to cover social robots need to exhibit lifelike functions, such as self-healing,” wrote the researchers in the study.

Shoji Takeuchi, Michio Kawai, Minghao Nie, and Haruka Oda authored the study, titled “Perforation-type anchors inspired by skin ligament for robotic face covered with living skin,” which is due for July publication in Cell Reports Physical Science. We learned of the study from a report published earlier this week by New Scientist.

The study describes a novel method for attaching cultured skin to robotic surfaces using “perforation-type anchors” inspired by natural skin ligaments. These tiny v-shaped cavities in the robot’s structure allow living tissue to infiltrate and create a secure bond, mimicking how human skin attaches to underlying tissues.

To demonstrate the skin’s capabilities, the team engineered a palm-sized robotic face able to form a convincing smile. Actuators connected to the base allowed the face to move, with the living skin flexing. The researchers also covered a static 3D-printed head shape with the engineered skin.

Enlarge / “Demonstration of the perforation-type anchors to cover the facial device with skin equivalent.”

Takeuchi et al. created their robotic face by first 3D-printing a resin base embedded with the perforation-type anchors. They then applied a mixture of human skin cells in a collagen scaffold, allowing the living tissue to grow into the anchors.

Researchers craft smiling robot face from living human skin cells Read More »

OpenAI’s new “CriticGPT” model is trained to criticize GPT-4 outputs

AI, Biz & IT, chatgpt, chatgtp, CriticGPT, GPT-4, GPT-4-turbo, GPT-4o, machine learning, openai / Mike M. / June 27, 2024

automated critic —

Research model catches bugs in AI-generated code, improving human oversight of AI.

Benj Edwards – Jun 27, 2024 7: 40 pm UTC

Enlarge / An illustration created by OpenAI.

On Thursday, OpenAI researchers unveiled CriticGPT, a new AI model designed to identify mistakes in code generated by ChatGPT. It aims to enhance the process of making AI systems behave in ways humans want (called “alignment”) through Reinforcement Learning from Human Feedback (RLHF), which helps human reviewers make large language model (LLM) outputs more accurate.

As outlined in a new research paper called “LLM Critics Help Catch LLM Bugs,” OpenAI created CriticGPT to act as an AI assistant to human trainers who review programming code generated by the ChatGPT AI assistant. CriticGPT—based on the GPT-4 family of LLMS—analyzes the code and points out potential errors, making it easier for humans to spot mistakes that might otherwise go unnoticed. The researchers trained CriticGPT on a dataset of code samples with intentionally inserted bugs, teaching it to recognize and flag various coding errors.

The researchers found that CriticGPT’s critiques were preferred by annotators over human critiques in 63 percent of cases involving naturally occurring LLM errors and that human-machine teams using CriticGPT wrote more comprehensive critiques than humans alone while reducing confabulation (hallucination) rates compared to AI-only critiques.

Developing an automated critic

The development of CriticGPT involved training the model on a large number of inputs containing deliberately inserted mistakes. Human trainers were asked to modify code written by ChatGPT, introducing errors and then providing example feedback as if they had discovered these bugs. This process allowed the model to learn how to identify and critique various types of coding errors.

In experiments, CriticGPT demonstrated its ability to catch both inserted bugs and naturally occurring errors in ChatGPT’s output. The new model’s critiques were preferred by trainers over those generated by ChatGPT itself in 63 percent of cases involving natural bugs (the aforementioned statistic). This preference was partly due to CriticGPT producing fewer unhelpful “nitpicks” and generating fewer false positives, or hallucinated problems.

The researchers also created a new technique they call Force Sampling Beam Search (FSBS). This method helps CriticGPT write more detailed reviews of code. It lets the researchers adjust how thorough CriticGPT is in looking for problems, while also controlling how often it might make up issues that don’t really exist. They can tweak this balance depending on what they need for different AI training tasks.

Interestingly, the researchers found that CriticGPT’s capabilities extend beyond just code review. In their experiments, they applied the model to a subset of ChatGPT training data that had previously been rated as flawless by human annotators. Surprisingly, CriticGPT identified errors in 24 percent of these cases—errors that were subsequently confirmed by human reviewers. OpenAI thinks this demonstrates the model’s potential to generalize to non-code tasks and highlights its ability to catch subtle mistakes that even careful human evaluation might miss.

Despite its promising results, like all AI models, CriticGPT has limitations. The model was trained on relatively short ChatGPT answers, which may not fully prepare it for evaluating longer, more complex tasks that future AI systems might tackle. Additionally, while CriticGPT reduces confabulations, it doesn’t eliminate them entirely, and human trainers can still make labeling mistakes based on these false outputs.

The research team acknowledges that CriticGPT is most effective at identifying errors that can be pinpointed in one specific location within the code. However, real-world mistakes in AI outputs can often be spread across multiple parts of an answer, presenting a challenge for future iterations of the model.

OpenAI plans to integrate CriticGPT-like models into its RLHF labeling pipeline, providing its trainers with AI assistance. For OpenAI, it’s a step toward developing better tools for evaluating outputs from LLM systems that may be difficult for humans to rate without additional support. However, the researchers caution that even with tools like CriticGPT, extremely complex tasks or responses may still prove challenging for human evaluators—even those assisted by AI.

OpenAI’s new “CriticGPT” model is trained to criticize GPT-4 outputs Read More »

AI-generated Al Michaels to provide daily recaps during 2024 Summer Olympics

2024 Summer Olympics, AI, Al Michaels, Biz & IT, chatgpt, chatgtp, Deepfakes, machine learning, nbc, nfl, olympic games, Olympics, speech synthesis, sports, sports commentators, Summer Olympics, voice synthesis / Kelly Newman / June 27, 2024

forever young —

AI voice clone will narrate daily Olympics video recaps; critics call it a “code-generated ghoul.”

Benj Edwards – Jun 27, 2024 3: 30 pm UTC

Al Michaels looks on prior to the game between the Minnesota Vikings and Philadelphia Eagles at Lincoln Financial Field on September 14, 2023 in Philadelphia, Pennsylvania. — Enlarge / Al Michaels looks on prior to the game between the Minnesota Vikings and Philadelphia Eagles at Lincoln Financial Field on September 14, 2023, in Philadelphia, Pennsylvania.

On Wednesday, NBC announced plans to use an AI-generated clone of famous sports commentator Al Michaels‘ voice to narrate daily streaming video recaps of the 2024 Summer Olympics in Paris, which start on July 26. The AI-powered narration will feature in “Your Daily Olympic Recap on Peacock,” NBC’s streaming service. But this new, high-profile use of voice cloning worries critics, who say the technology may muscle out upcoming sports commentators by keeping old personas around forever.

NBC says it has created a “high-quality AI re-creation” of Michaels’ voice, trained on Michaels’ past NBC appearances to capture his distinctive delivery style.

The veteran broadcaster, revered in the sports commentator world for his iconic “Do you believe in miracles? Yes!” call during the 1980 Winter Olympics, has been covering sports on TV since 1971, including a high-profile run of play-by-play coverage of NFL football games for both ABC and NBC since the 1980s. NBC dropped him from NFL coverage in 2023, however, possibly due to his age.

Michaels, who is 79 years old, shared his initial skepticism about the project in an interview with Vanity Fair, as NBC News notes. After hearing the AI version of his voice, which can greet viewers by name, he described the experience as “astonishing” and “a little bit frightening.” He said the AI recreation was “almost 2% off perfect” in mimicking his style.

The Vanity Fair article provides some insight into how NBC’s new AI system works. It first uses a large language model (similar technology to what powers ChatGPT) to analyze subtitles and metadata from NBC’s Olympics video coverage, summarizing events and writing custom output to imitate Michaels’ style. This text is then fed into an unspecified voice AI model trained on Michaels’ previous NBC appearances, reportedly replicating his unique pronunciations and intonations.

NBC estimates that the system could generate nearly 7 million personalized variants of the recaps across the US during the games, pulled from the network’s 5,000 hours of live coverage. Using the system, each Peacock user will receive about 10 minutes of personalized highlights.

A diminished role for humans in the future?

Al Michaels reports on the Sweden vs. USA men's ice hockey game at the 1980 Olympic Winter Games on February 12, 1980. — Enlarge / Al Michaels reports on the Sweden vs. USA men’s ice hockey game at the 1980 Olympic Winter Games on February 12, 1980.

It’s no secret that while AI is wildly hyped right now, it’s also controversial among some. Upon hearing the NBC announcement, critics of AI technology reacted strongly. “@NBCSports, this is gross,” tweeted actress and filmmaker Justine Bateman, who frequently uses X to criticize technologies that might replace human writers or performers in the future.

A thread of similar responses from X users reacting to the sample video provided above included criticisms such as, “Sounds pretty off when it’s just the same tone for every single word.” Another user wrote, “It just sounds so unnatural. No one talks like that.”

The technology will not replace NBC’s regular human sports commentators during this year’s Olympics coverage, and like other forms of AI, it leans heavily on existing human work by analyzing and regurgitating human-created content in the form of captions pulled from NBC footage.

Looking down the line, due to AI media cloning technologies like voice, video, and image synthesis, today’s celebrities may be able to attain a form of media immortality that allows new iterations of their likenesses to persist through the generations, potentially earning licensing fees for whoever holds the rights.

We’ve already seen it with James Earl Jones playing Darth Vader’s voice, and the trend will likely continue with other celebrity voices, provided the money is right. Eventually, it may extend to famous musicians through music synthesis and famous actors in video-synthesis applications as well.

The possibility of being muscled out by AI replicas factored heavily into a Hollywood actors’ strike last year, with SAG-AFTRA union President Fran Drescher saying, “If we don’t stand tall right now, we are all going to be in trouble. We are all going to be in jeopardy of being replaced by machines.”

For companies that like to monetize media properties for as long as possible, AI may provide a way to maintain a media legacy through automation. But future human performers may have to compete against all of the greatest performers of the past, rendered through AI, to break out and forge a new career—provided there will be room for human performers at all.

“Al Michaels became Al Michaels because he was brought in to replace people who died, or retired, or moved on,” tweeted a writer named Geonn Cannon on X. “If he can’t do the job anymore, it’s time to let the next Al Michaels have a shot at it instead of just planting a code-generated ghoul in an empty chair.“

AI-generated Al Michaels to provide daily recaps during 2024 Summer Olympics Read More »

Toys “R” Us riles critics with “first-ever” AI-generated commercial using Sora

AI, Air Head, Biz & IT, Joe Russo, machine learning, Mike Drucker, Native Foreign, Patrick Cederberg, shy kids, Toys, Toys"R"Us, video synthesis / DJ Henderson / June 26, 2024

Enlarge / A screen capture from the partially AI-generated Toys “R” Us brand film created using Sora.

Toys R Us

On Monday, Toys “R” Us announced that it had partnered with an ad agency called Native Foreign to create what it calls “the first-ever brand film using OpenAI’s new text-to-video tool, Sora.” OpenAI debuted Sora in February, but the video synthesis tool has not yet become available to the public. The brand film tells the story of Toys “R” Us founder Charles Lazarus using AI-generated video clips.

“We are thrilled to partner with Native Foreign to push the boundaries of Sora, a groundbreaking new technology from OpenAI that’s gaining global attention,” wrote Toys “R” Us on its website. “Sora can create up to one-minute-long videos featuring realistic scenes and multiple characters, all generated from text instruction. Imagine the excitement of creating a young Charles Lazarus, the founder of Toys “R” Us, and envisioning his dreams for our iconic brand and beloved mascot Geoffrey the Giraffe in the early 1930s.”

The company says that The Origin of Toys “R” Us commercial was co-produced by Toys “R” Us Studios President Kim Miller Olko as executive producer and Native Foreign’s Nik Kleverov as director. “Charles Lazarus was a visionary ahead of his time, and we wanted to honor his legacy with a spot using the most cutting-edge technology available,” Miller Olko said in a statement.

In the video, we see a child version of Lazarus, presumably generated using Sora, falling asleep and having a dream that he is flying through a land of toys. Along the way, he meets Geoffery, the store’s mascot, who hands the child a small red car.

Many of the scenes retain obvious hallmarks of AI-generated imagery, such as unnatural movement, strange visual artifacts, and the irregular shape of eyeglasses. In February, a few Super Bowl commercials intentionally made fun of similar AI-generated video defects, which became famous online after fake AI-generated beer commercial and “Pepperoni Hug Spot” clips made using Runway’s Gen-2 model went viral in 2023.

A screen capture from the partially AI-generated Toys “R” Us brand film created using Sora.

Toys “R” Us
A screen capture from the partially AI-generated Toys “R” Us brand film created using Sora.

Toys “R” Us
A screen capture from the partially AI-generated Toys “R” Us brand film created using Sora.

Toys “R” Us
A screen capture from the partially AI-generated Toys “R” Us brand film created using Sora.

Toys R Us
A screen capture from the partially AI-generated Toys “R” Us brand film created using Sora.

Toys R Us
A screen capture from the partially AI-generated Toys “R” Us brand film created using Sora.

Toys “R” Us
A screen capture from the partially AI-generated Toys “R” Us brand film created using Sora.

Toys “R” Us
A screen capture from the partially AI-generated Toys “R” Us brand film created using Sora.

Toys “R” Us
A screen capture from the partially AI-generated Toys “R” Us brand film created using Sora.

Toys “R” Us
A screen capture from the partially AI-generated Toys “R” Us brand film created using Sora.

Toys “R” Us
A screen capture from the partially AI-generated Toys “R” Us brand film created using Sora.

Toys “R” Us

AI-generated artwork receives frequent criticism online due to the use of human-created artwork to train AI models that create the works, the perception that AI synthesis tools will replace (or are currently replacing) human creative jobs, and the potential environmental impact of AI models, which are seen as energy-wasteful by some critics. Also, some people just think the output quality looks bad.

On the social network X, comedy writer Mike Drucker wrapped up several of these criticisms into one post, writing, “Love this commercial is like, ‘Toys R Us started with the dream of a little boy who wanted to share his imagination with the world. And to show how, we fired our artists and dried Lake Superior using a server farm to generate what that would look like in Stephen King’s nightmares.'”

Other critical comments were more frank. Filmmaker Joe Russo posted: “TOYS ‘R US released an AI commercial and it fucking sucks.”

Toys “R” Us riles critics with “first-ever” AI-generated commercial using Sora Read More »

YouTube tries convincing record labels to license music for AI song generator

AI, AI music, AI music generation, music, syndication, youtube / DJ Henderson / June 26, 2024

Jukebox zeroes —

Video site needs labels’ content to legally train AI song generators.

Anna Nicolaou and Madhumita Murgia, Financial Times – Jun 26, 2024 1: 48 pm UTC

Chris Ratcliffe/Bloomberg via Getty

YouTube is in talks with record labels to license their songs for artificial intelligence tools that clone popular artists’ music, hoping to win over a skeptical industry with upfront payments.

The Google-owned video site needs labels’ content to legally train AI song generators, as it prepares to launch new tools this year, according to three people familiar with the matter.

The company has recently offered lump sums of cash to the major labels—Sony, Warner, and Universal—to try to convince more artists to allow their music to be used in training AI software, according to several people briefed on the talks.

However, many artists remain fiercely opposed to AI music generation, fearing it could undermine the value of their work. Any move by a label to force their stars into such a scheme would be hugely controversial.

“The industry is wrestling with this. Technically the companies have the copyrights, but we have to think through how to play it,” said an executive at a large music company. “We don’t want to be seen as a Luddite.”

YouTube last year began testing a generative AI tool that lets people create short music clips by entering a text prompt. The product, initially named “Dream Track,” was designed to imitate the sound and lyrics of well-known singers.

But only 10 artists agreed to participate in the test phase, including Charli XCX, Troye Sivan and John Legend, and Dream Track was made available to just a small group of creators.

YouTube wants to sign up “dozens” of artists to roll out a new AI song generator this year, said two of the people.

YouTube said: “We’re not looking to expand Dream Track but are in conversations with labels about other experiments.”

Licenses or lawsuits

YouTube is seeking new deals at a time when AI companies such as OpenAI are striking licensing agreements with media groups to train large language models, the systems that power AI products such as the ChatGPT chatbot. Some of those deals are worth tens of millions of dollars to media companies, insiders say.

The deals being negotiated in music would be different. They would not be blanket licenses but rather would apply to a select group of artists, according to people briefed on the discussions.

It would be up to the labels to encourage their artists to participate in the new projects. That means the final amounts YouTube might be willing to pay the labels are at this stage undetermined.

The deals would look more like the one-off payments from social media companies such as Meta or Snap to entertainment groups for access to their music, rather than the royalty-based arrangements labels have with Spotify or Apple, these people said.

YouTube’s new AI tool, which is unlikely to carry the Dream Track brand, could form part of YouTube’s Shorts platform, which competes with TikTok. Talks continue and deal terms could still change, the people said.

YouTube’s latest move comes as the leading record companies on Monday sued two AI start-ups, Suno and Udio, which they allege are illegally using copyrighted recordings to train their AI models. A music industry group is seeking “up to $150,000 per work infringed,” according to the filings.

After facing the threat of extinction following the rise of Napster in the 2000s, music companies are trying to get ahead of disruptive technology this time around. The labels are keen to get involved with licensed products that use AI to create songs using their music copyrights—and get paid for it.

Sony Music, which did not participate in the first phase of YouTube’s AI experiment, is in negotiations with the tech group to make available some of its music to the new tools, said a person familiar with the matter. Warner and Universal, whose artists participated in the test phase, are also in talks with YouTube about expanding the product, these people said.

In April, more than 200 musicians including Billie Eilish and the estate of Frank Sinatra signed an open letter.

“Unchecked, AI will set in motion a race to the bottom that will degrade the value of our work and prevent us from being fairly compensated for it,” the letter said.

YouTube added: “We are always testing new ideas and learning from our experiments; it’s an important part of our innovation process. We will continue on this path with AI and music as we build for the future.”

YouTube tries convincing record labels to license music for AI song generator Read More »

Taking a closer look at AI’s supposed energy apocalypse

AI, Features / DJ Henderson / June 25, 2024

Enlarge / Someone just asked what it would look like if their girlfriend was a Smurf. Better add another rack of servers!

Getty Images

Late last week, both Bloomberg and The Washington Post published stories focused on the ostensibly disastrous impact artificial intelligence is having on the power grid and on efforts to collectively reduce our use of fossil fuels. The high-profile pieces lean heavily on recent projections from Goldman Sachs and the International Energy Agency (IEA) to cast AI’s “insatiable” demand for energy as an almost apocalyptic threat to our power infrastructure. The Post piece even cites anonymous “some [people]” in reporting that “some worry whether there will be enough electricity to meet [the power demands] from any source.”

Digging into the best available numbers and projections available, though, it’s hard to see AI’s current and near-future environmental impact in such a dire light. While generative AI models and tools can and will use a significant amount of energy, we shouldn’t conflate AI energy usage with the larger and largely pre-existing energy usage of “data centers” as a whole. And just like any technology, whether that AI energy use is worthwhile depends largely on your wider opinion of the value of generative AI in the first place.

Not all data centers

While the headline focus of both Bloomberg and The Washington Post’s recent pieces is on artificial intelligence, the actual numbers and projections cited in both pieces overwhelmingly focus on the energy used by Internet “data centers” as a whole. Long before generative AI became the current Silicon Valley buzzword, those data centers were already growing immensely in size and energy usage, powering everything from Amazon Web Services servers to online gaming services, Zoom video calls, and cloud storage and retrieval for billions of documents and photos, to name just a few of the more common uses.

The Post story acknowledges that these “nondescript warehouses packed with racks of servers that power the modern Internet have been around for decades.” But in the very next sentence, the Post asserts that, today, data center energy use “is soaring because of AI.” Bloomberg asks one source directly “why data centers were suddenly sucking up so much power” and gets back a blunt answer: “It’s AI… It’s 10 to 15 times the amount of electricity.”

Enlarge / The massive growth in data center power usage mostly predates the current mania for generative AI (red 2022 line added by Ars).

Unfortunately for Bloomberg, that quote is followed almost immediately by a chart that heavily undercuts the AI alarmism. That chart shows worldwide data center energy usage growing at a remarkably steady pace from about 100 TWh in 2012 to around 350 TWh in 2024. The vast majority of that energy usage growth came before 2022, when the launch of tools like Dall-E and ChatGPT largely set off the industry’s current mania for generative AI. If you squint at Bloomberg’s graph, you can almost see the growth in energy usage slowing down a bit since that momentous year for generative AI.

Determining precisely how much of that data center energy use is taken up specifically by generative AI is a difficult task, but Dutch researcher Alex de Vries found a clever way to get an estimate. In his study “The growing energy footprint of artificial intelligence,” de Vries starts with estimates that Nvidia’s specialized chips are responsible for about 95 percent of the market for generative AI calculations. He then uses Nvidia’s projected production of 1.5 million AI servers in 2027—and the projected power usage for those servers—to estimate that the AI sector as a whole could use up anywhere from 85 to 134 TWh of power in just a few years.

To be sure, that is an immense amount of power, representing about 0.5 percent of projected electricity demand for the entire world (and an even greater ratio in the local energy mix for some common data center locations). But measured against other common worldwide uses of electricity, it’s not representative of a mind-boggling energy hog. A 2018 study estimated that PC gaming as a whole accounted for 75 TWh of electricity use per year, to pick just one common human activity that’s on the same general energy scale (and that’s without console or mobile gamers included).

Enlarge / Worldwide projections for AI energy use in 2027 are on the same scale as the energy used by PC gamers.

More to the point, de Vries’ AI energy estimates are only a small fraction of the 620 to 1,050 TWh that data centers as a whole are projected to use by 2026, according to the IEA’s recent report. The vast majority of all that data center power will still be going to more mundane Internet infrastructure that we all take for granted (and which is not nearly as sexy of a headline bogeyman as “AI”).

Taking a closer look at AI’s supposed energy apocalypse Read More »