Ukraine's ground robot fleet is growing and performing more missions to keep human soldiers further from the fight. Diego Herrera Carcedo/Anadolu via Getty Images Ukraine aims to have robots handle 100% of front-line logistics missions. Its defense minister said Ukraine will contract 25,000 robots in the first half of this year. Their use has surged. They carry gear and troops and attack Russian positions. Ukraine is taking steps to boost its rapidly growing arsenal of war robots, including buying thousands of new ones and pushing to turn over 100% of front-line logistics to robotic systems. Defense Minister Mykhailo Fedorov announced changes to boost the production of ground robots, also known as unmanned ground vehicles, which already help Ukrainian troops carry gear, lay mines, evacuate the wounded, and attack Russian positions. He said that due to increased demand, Ukraine will contract 25,000 new robots in the first half of this year, double last year's total. A key goal, he said, is for 100% of its front-line logistics to be performed by these robots. That would keep the vulnerable soldiers and vehicles making dangerous supply runs out of harm's way. Soldiers have detailed their impact. Oleksandr Yabchanka, the head of robotic systems for Ukraine's Da Vinci Wolves Battalion, previously told Business Insider that one average-sized war robot can carry more than roughly 10 servicemen can. Fedorov called the robots "one of the most dynamic areas of defense tech" and said that more than 280 companies are working in the space in Ukraine. 280 companies are working in the space in Ukraine. Ground robots have been important for Ukraine's fight against Russia's invasion because, like aerial drones, they replace soldiers in dangerous roles, which is critical for a smaller force like Ukraine's that can't afford to waste manpower resources. President Volodymyr Zelenskyy shared last week that Ukraine's ground robot force had carried out more than 22,000 missions on the front lines over the previous three months, saying that means "lives were saved more than 22,000 times when a robot went into the most dangerous areas instead of a warrior." The president's comments also revealed a big jump in the robots' overall usage. Oleksandr Syrskyi, the commander in chief of Ukraine's armed forces, said in December that over the previous six months, Ukraine's ground robots had carried out only 2,000 missions. Then, in half the time, that figure grew to 10 times that. Fedorov also revealed a substantial increase in battlefield usage. He shared that more than 9,000 missions were carried out with ground robots in March, demonstrating how central these assets are becoming to Ukrainian operations. The robots are performing new types of missions. Units and manufacturers say Russian robots have surrendered to them, and Zelenskyy said they were used to take a Russian position by only using unmanned platforms for the first time in this war. "The future is already on the front line — and Ukraine is building it," he said. Ground robots have been used in other conflicts, including by Western militaries, but not to this extent. The West has been paying close attention, increasingly interested in adapting the technology further for NATO militaries. Fedorov said that as part of the new efforts, Ukraine's Agency for Defense Procurement has so far concluded 19 contracts worth around $250 million with manufacturers. He said he met with ground robot manufacturers to accelerate supply, and they identified key challenges, including scaling production and integrating them into combat operations. He listed new policy changes to allow acceleration, including the defense procurement agency speeding up the conclusion of contracts at the ministry of defense's request, changing how contracts work so manufacturers will have contracts for next year this year, and creating a new center for manufacturers. And at the same time, he said, Ukraine is also boosting the production of other components and accessories that make ground robots effective, like weaponry that attaches to them and parts that let them mine and de-mine. Fedorov said the focus is on cheap and effective systems. Partner nations are watching that shift toward mass, low-cost weapons over smaller numbers of advanced systems. Read the original article on Business Insider

Ukraine's General Cherry makes FPV and interceptor drones. AP Photo/Evgeniy Maloletka A Ukrainian drone maker is teaming up with a US company to make its drones on US soil. It fears being "eliminated" if it stays in Ukraine, either by Russian attacks or competition. Ukraine's firms are increasingly building abroad. A Ukrainian drone maker worried about getting "eliminated" at home is teaming up with a US manufacturer to make weapons on American soil. Stanislav Hryshyn, co-founder of General Cherry, told Business Insider that "actually to stay in Ukraine, that means actually to stay local and to be eliminated in a couple of years or even months." Rising Ukrainian defense companies like his are increasingly seeing opportunities beyond their own conflict. They believe staying in Ukraine — where they're in the crosshairs and fighting to survive in a heavily saturated market with limited growth potential — could ultimately kill them, physically and commercially. Staying home risks being outcompeted or destroyed. First, he said, "we're a target for the enemy." Russia routinely targets Ukraine's defense production with missile and drone attacks. But beyond that, the Ukrainian market is "not the biggest one, and it's very local," limited by what Ukraine's limited defense budget can afford. And at General Cherry, "we don't want to stay the local players." Ukrainian products, like interceptor drones, are increasingly in demand. AP Photo/Efrem Lukatsky Another concern is that without an international market presence, foreign companies could copy Ukrainian innovations and scale production, taking market share that could have gone to Ukrainian firms. The company is now partnering with US manufacturer Wilcox Industries to make first-person-view (FPV) and interceptor drones in the US, with development led by General Cherry and production based at Wilcox's facility in New Hampshire. The project is still awaiting formal approval, but the firm says Ukraine has endorsed the move, explaining it's "about scaling what works" and "bringing Ukrainian technology to the global level." Hryshyn said that for security reasons, he could not comment on where the drones produced through the partnership would go, Ukraine or the US. James Teetzel, the founder and CEO of Wilcox Industries Corp, said in a statement on the partnership that "there is no doubt in my mind that combining our Wilcox manufacturing infrastructure and engineering resources with General Cherry's technology, we will bring great value to our US Government." Teetzel's company makes defense equipment, including fire control systems and laser aiming devices. General Cherry wants to be "integrated into the global defense system, to be a global player, to play on an international level," Hryshyn said. And for the company's growth, too, "we think it would be actually a crime if we stay in Ukraine. We're already big in Ukraine." But it's more than that. Hryshyn says "the truth is actually that for US companies, they also have to learn a lot from us. And we hope that it'll be a fruitful partnership for both of us." Officials and industry leaders across the NATO alliance say there is significant value in Ukraine's battlefield-driven innovation and rapid production cycles. Ukrainian gear is in demand Ukraine's defense industry is booming, making up a growing share of the equipment used by its military and drawing increasing interest from allies seeking battle-proven technology. Foreign countries are interested in Ukraine's battlefield tech and production techniques. AP Photo/Evgeniy Maloletka But staying in Ukraine comes with constraints. Companies are prioritizing supplying the war effort, but operating only domestically limits how much they can produce, hindering not just the companies but also wartime production capacity. Industry leaders say that manufacturers have the capacity to make far more than they currently do, but they can't because Ukraine's defense budget can't afford to buy more gear. They argue exports would generate revenue to scale production, including for Ukraine's own forces. Staying in Ukraine also means spreading out manufacturing and slowing production efforts to account for Russian attacks. Companies sometimes split their production across almost a dozen sites, which makes operations safer but also a lot less efficient. Ukraine's leaders have been taking steps to expand which gear manufacturers can send abroad and which partnerships the industry can establish with other countries, including agreements that allow allies to obtain some Ukrainian weaponry. Hryshyn said that his company chose Wilcox Industries because it is an established manufacturer with proven products. He said that while some startups were very interesting, they had not yet delivered tangible results. Additionally, he said, "We think that the US market is actually the biggest one in the world." General Cherry is one of a growing number of Ukrainian companies partnering with NATO companies to produce weapons abroad — reducing exposure to attacks while giving Western partners access to Ukrainian technology and experience. Denmark, a leader in the "Build with Ukraine" initiative, has said NATO companies have much to learn from Ukrainian firms and is seeking access to their tech. Hryshyn said that he sees that demand firsthand. In discussions with Western partners, "our experience and expertise are pretty valuable for them." He added that the US is "learning some things from Ukraine," including how to work with startups and operate in a more open, flexible defense ecosystem. Read the original article on Business Insider

Aisha Malik / TechCrunch: Deezer says AI-generated tracks now account for 44% of daily uploads, totaling ~75K tracks per day and 2M+ per month, but account for just 1-3% of consumption  —  Deezer announced on Monday that AI-generated tracks now represent 44% of all new music uploaded to its platform.

The move comes after the vendor forged significant deals with OpenAI and AWS earlier this year.

Yazhou Sun / Bloomberg: Sources: UK-based CuspAI, which aims to use AI for discovering new materials, is in discussions to raise at least $200M, taking its valuation to over $1B  —  CuspAI, a British startup that aims to use artificial intelligence for discovering new materials, is in discussions to raise …

Kea parrot missing his upper beak "has rewritten what disability means for behaviorally complex species."

Conceptual overview and practical guidance The post Context Payload Optimization for ICL-Based Tabular Foundation Models appeared first on Towards Data Science.

Deezer says consumption of AI-generated music on the platform is still very low, between 1-3% of the total streams, and that 85% of these streams are detected as fraudulent and are demonetized.

Damn! It looks like I've got bixonimania! monshtein/Shutterstock.com Until a few years ago, no one had heard of bixonimania. Then, in 2024, a group of scientists posted findings online announcing the condition, which they claimed affected the eyes after computer use. However, the scientists had made it up – not just the work, but the authors’ names, affiliations, locations and funding, which was the University of Fellowship of the Ring and the Galactic Triad. Large language models like ChatGPT and Gemini treated it as real anyway, and in doing so, helped turn a fictional disease into a legitimate-sounding health concern. Bixonimania is not an isolated case. Being deceived – whether you are a person or an AI model – is concerningly common, in science and beyond. Whether we’re talking about AI hallucinations, state-backed disinformation or just everyday lies, humans have a remarkable knack for naivety, owing to our biases and increasing need to outsource learning to others. These are problems we – individually and collectively – urgently need to better understand and overcome. Our shared fascination with deception may help explain the popularity of The Traitors, a TV programme built around the tension between trust and suspicion, where contestants must decide who among them is deceiving the group. The show captures something intrinsic to being human: the persistent threat of being unsure about whether we’re placing trust effectively. Yet in the modern era of mass digital communication and AI, we’re now almost constantly faced with a similar threat, often without realising it. At a recent event at the Cambridge Festival, we aimed to highlight this risk through a Traitors-themed science event. Four panellists presented work, all of which could have been a lie. The audience was asked to vote on which of the presenters was deceiving them and why. We deliberately made the presenters and their work outlandish. From their varying backgrounds and with varying accents, the panellists presented their work in global health, climate, media and astrophysics. Some dressed formally, while one – a Nigerian researcher presenting her work on immigration in a healthcare context – wore clothes linked to her ethnic identity. We were interested in exploring which of these signals – accent, gender, ethnicity, and dress and presentation style – influenced the audience’s decisions. Both content and presentation styles influenced them, but the signals they relied on led them to the wrong conclusions, rating the traitors as more credible than honest researchers. The ones who received the most votes were the two “faithful” researchers (to use the language of The Traitors) – Ada, from the non-profit Development Media Initiative, and Sarah, an astrophysicist working in galactic archaeology. Ada’s team had saved lives by sharing health information with communities in the global south through running ten radio broadcasts daily. The audience thought the results were implausibly impressive. “Ada’s data is too good to be true,” one person reported in our questionnaire. She was also presenting work she hadn’t personally contributed to. Even though this is common in large collaborations, this distance led to perceptions of a lack of confidence, undermining her credibility. Sarah, an astrophysicist, had presented her subfield of galactic archaeology – the study of the Milky Way’s formation history through the chemical signatures of ancient stars. Yet with only four minutes to speak, she was unable to convey significant depth. The audience read that as a lack of understanding. The outlandishness of her field’s name also harmed perceptions of her legitimacy. “Galaxy [sic] archaeology is too cool a name to exist,” one audience member wrote. By contrast, the two traitors, Jack and Joyce, received the fewest votes. Jack was an actor who created the persona of a climate researcher specialising in rain. Joyce presented her own work but falsified the results. Interestingly, Joyce’s personal connection to her work – she is a Nigerian woman conducting research into Nigerian communities – helped to convince the audience of her authenticity. “Joyce’s presentation sounded very considered and genuine – the process of her research and recounts of her personal experiences sounded like she had lots of interest in the area,” one person wrote. The traitors, as the audience saw them. University of Cambridge, CC BY-SA The event was meant to be fun and engaging. Yet we also wanted to illustrate the many ways people can misrepresent themselves, whether in science or beyond. Our traitors showed that lies don’t just have to be about who you are (Jack is an actor, not a researcher) but about what you say (Joyce is a researcher but falsified her results). Misinformation has always existed. What’s new is the speed at which it spreads, the tools that generate it, and how convincingly it mimics the real thing. Why maths isn’t enough Our collective capacity to recognise false information is also at risk. This is because, as a society, we continue to promote the importance of hard science subjects at the expense of the critical thinking skills derived from studies of the arts, humanities and social sciences. This can be seen, for example, in the 2023 UK governmental push to require all school students to take maths until age 18. No such push exists to promote and develop the critical thinking skills of young people. It’s easy to see how increasingly convincing falsehoods like bixonimania’s existence can be accepted as truth, especially when touted by AI models. Tools are helpful. AI is a tool, the internet is a tool, the media is a tool. But it’s up to us to ensure that we are using them and not being manipulated by them. In The Traitors, we have little to go on to determine what is true. Yet in the real world, we have the ability to check the truth of claims. With effective caution and critical thinking, it is entirely possible to determine what is trustworthy, but it requires thinking for ourselves. Trust is ours to give, and we need to learn to give it wisely. Jonathan R. Goodman receives funding from the National Institute of Health Research, the Wellcome Sanger Institute, and the Wellcome Trust. Mariam Rashid receives funding from the Isaac Newton Trust and the Kavli Foundation.

Vercel blamed its breach on an earlier hack at Context AI, which allowed hackers to hijack a Vercel employee's account to steal customer data.

Strangers can infer limited info about you without knowing or messaging you, which could theoretically aid certain kinds of malicious activity.

Discussion | Link

Bloomberg: China's iQiyi plans to overhaul its streaming service into an AI content hub, with an app redesign and Nadou Pro AI tool handling “every aspect of film-making”  —  IQiyi Inc. expects AI to create the bulk of its films and shows in five years, a monumental industry shift that spurred …

Cyberdefenses could be exposed faster than fixes could be deployed.

Discussion | Link

A new report from French IT consulting firm Capgemini found that businesses are rapidly moving physical AI from experimentation to implementation.

In today’s hyper-competitive landscape, "speed to insight" is no longer the finish...

Encryption at rest is a cloud baseline, but for enterprises operating in highly regulated...

What do the large language models behind human-like conversational AI really know and what does it mean to live alongside them? A new book Understanding Conversational AI by Thierry Poibeau offers a critical and interdisciplinary exploration of large language models (LLMs), examining how they reshape our understanding of language, cognition and society. Drawing on philosophy of language, linguistics, cognitive science and AI ethics, it investigates how LLMs generate meaning, simulate reasoning and perform tasks that once seemed uniquely human from translation to moral judgement and literary creation. The book explores the limitations of these models, their embedded biases and their role in processes of automation, misinformation and platform enclosure while reflecting on how they prompt us to revisit fundamental questions: What is understanding? What is creativity? How do we ascribe agency or trust in a world of synthetic language? Recent findings are suggesting that non-expert poetry readers prefer AI works to poems written by humans. Thierry Poibeau reminds us how “aesthetic judgment in poetry often involves elements that are not easily codifiable, such as originality, emotional resonance, metaphorical depth, and cultural embeddedness”. Poibeau sheds light on why AI-generated poems can be appealing, by highlighting how literary value is “shaped by evolving norms within particular reading communities, making judgments of poetic merit historically contingent and socially negotiated”. AI Poetry in motion Excerpts from Understanding Conversational AI “The arrival of AI-generated poetry raises fundamental questions about whether current evaluation criteria, often rooted in human experience, intentional expression, and historical context are adequate for assessing texts produced through large-scale statistical recombination. As large language models generate increasingly fluent and plausible verse, their outputs may challenge existing notions of aesthetic authenticity, precisely because they blur the boundary between craft, imitation, and genuine creative insight. These systems are capable of reproducing recognizable poetic forms, stylistic conventions, and affective tones, raising the question of how such texts should be evaluated when traditional assumptions about authorship and intention no longer clearly apply. However, despite this formal competence, recent studies reveal recurrent stylistic tendencies characteristic of large language model–generated poetry. GPT-4, for instance, shows a strong preference for quatrain structures, a frequent reliance on iambic meter and end rhyme, and a tendency to employ recurring lexical choices such as heart, whisper, or dream. These outputs often reflect a flattening of emotional and metaphorical complexity, favoring literal formulations and conventional poetic tropes over ambiguity, innovation, or semantic depth. In comparison to human poetry, LLM-generated verse appears more homogeneous and less nuanced, and is less capable of producing the kinds of conceptual tension or unexpected imagery often found in more original human compositions. In addition to these stylistic patterns, recent reader evaluation studies suggest that some AI-generated poems are judged comparably to human-authored texts when assessed blind. However, once authorship is disclosed, evaluations tend to decline, indicating a persistent skepticism toward machine-generated poetry. This reception dynamic highlights the ongoing tension between the formal fluency of AI-generated verse and readers’ perceptions of authenticity and creative agency. Interestingly, readers often find AI-generated poems easier to interpret. They can more readily grasp the images, themes, and emotions, which are typically presented in a more accessible and transparent manner than in the often denser and more ambiguous work of human poets. As a result, readers may develop a preference for these texts and mistakenly interpret their own ease of understanding as evidence of human authorship. When the machine origin of the poem is revealed, this interpretation shifts, and the same textual features that previously facilitated comprehension may be reinterpreted as signs of superficiality or lack of depth. Such findings should be interpreted with caution. They do not demonstrate that AI-generated poetry is superior in literary quality, but rather that it tends to conform to familiar forms and accessible conventions. In contrast, human-authored poetry frequently relies on layers of complexity, allusion, and ambiguity that demand substantial background knowledge in literature, history, and poetic traditions. Interpreting such works is a cognitively demanding task, and it is therefore problematic to ask lay readers to evaluate poetry in a meaningful way without accounting for differences in expertise and interpretive competence. Beyond questions of evaluation, the reception of AI-generated poetry raises broader issues of attribution, disclosure, and creative ownership. Literary institutions, publishers, and competition organizers increasingly require authors to disclose the use of AI tools, reflecting uncertainty about how machine-generated texts should be situated within existing frameworks of creativity and value. The discomfort that emerges when readers discover that a poem was generated by a machine is not merely a reaction to deception but an indication that aesthetic judgment remains deeply entangled with assumptions about intention, experience, and authorship.” Thierry Poibeau is Director of Research at the Centre national de la recherche scientifique (CNRS), Paris and head of the LaTTiCE laboratory in Paris, France. He is a member of PRAIRIE-PSAI (Paris AI Research Institute – Paris School of Artificial Intelligence) and is also an affiliated lecturer at the Department of Theoretical and Applied Linguistics (DTAL) of the University of Cambridge. He works on natural language processing (NLP), particularly focusing on information extraction, question answering, semantic zoning, knowledge acquisition from text, and named entity tagging. A weekly e-mail in English featuring expertise from scholars and researchers. It provides an introduction to the diversity of research coming out of the continent and considers some of the key issues facing European countries. Get the newsletter! Thierry Poibeau is a member of PRAIRIE-PSAI (Paris AI Research Institute -- Paris School of Artificial Intelligence) and has received funding in this capacity.

Why it tickles your brain to use an LLM, and what that means for the AI industry The post The LLM Gamble appeared first on Towards Data Science.