Breaking AI & Tech News Analyzed

The Vision for AI Coding Agents Cognition CEO Scott Wu made headlines again this week when his two-year-old AI coding agent startup raised $1 billion at a $26 billion valuation. Cognition is the maker of Devin, one of the first and, arguably, most successful AI coding agents. Devin, the CEO says, “naturally owns tasks end to end.” The Future of Software Development In fact, in the blog post announcing that raise, Cognition laid out a vision where “we are shifting to a world of self-driving software development.” So, could Devin replace, say, a mid-level L4 programmer? Yes, and no, Wu told TechCrunch. “We’ve never thought about it as replacing humans. I know it’s like a scenario, folks have said these things. It has never been our view.” Preserving the Joy of Programming Wu emphasizes that the goal is not to make human programmers obsolete. “We are all programmers ourselves,” he explained. “I started coding when I was nine.” He views agents as another layer of abstraction between envisioning a software product and producing it, similar to how visual development environments abstracted software creation away from machine instructions. The Role of Devin in Cognition Cognition says that Devin’s role in its own company is to ship nearly all the software. The company says that 89% of code committed by its engineers was committed by Devin, and the rest by local agents. Wu explains that his agent’s role is largely to do the kinds of long-tail maintenance tasks that many programmers don’t like to do anyway: bringing old software up to date; moving applications off one platform and onto another. The Future of AI Agents Wu predicts that agents will enter other fields where they will learn tasks, from customer service to medicine, but hopes the goal will be to augment human workers in those areas, too. “Code and software has been the first to move, but we’ll see this happen in all these other industries,” he predicts. “One thing that’s been clear to us since the beginning is, it should always be up to the human what to do … you really see this in software engineering, but I think it’s true in all these other professions too.”

UK Study Reveals Air Pollution's Impact on Children's Lung DevelopmentResearch shows that air pollution is slowing the lung growth of children in the UK. Scientists tracked the lung function of more than 5,000 people who were born in and around Bristol in the 1990s. Their health was assessed from birth onwards and their lungs were tested as they grew up, at eight and 15 years old and then as adults, aged 24, when their lung function should have reached its maximum.Longitudinal Study Tracks Lung Function From Birth to AdulthoodProf Ann Hansell, of the University of Leicester, who led the study, said: "Much of the evidence on health effects of air pollution relates to adults or pregnancy, but we think it's highly plausible it has impacts on growth and development of children. Those whose lungs didn't grow to maximum potential in childhood may be more vulnerable to the respiratory diseases of later life because they have a lower reserve."Dr Katie Eminson, also of the University of Leicester and a first author of the study, explained: "Lung function was measured using spirometry by trained technicians. Participants were asked to take a deep breath in, then blow out as hard and as fast as possible into a mouthpiece. A machine measured both the amount of air they can breathe out and the speed of that breath, providing an indication of how well their lungs are working."Pollution Exposure Linked to Reduced Lung CapacityThe researchers calculated the children's air pollution exposure in each trimester of pregnancy and then for each year of early childhood. This included particle pollution as well as nitrogen dioxide, a gas that comes mainly from diesel cars and fossil gas boilers.Hansell noted: "We spent literally years creating the particulate air pollution exposure estimates in pregnancy and early life, including sourcing road traffic data from Bristol city council that are not available in the national database."The researchers allowed for other factors that can affect children's health, including premature birth, breastfeeding, parental smoking and home conditions including damp.They found that breathing more air pollution during pregnancy, infancy and early childhood can slow lung development all the way up to early adulthood. The greatest impact was during adolescence, which is the time when lung growth accelerates.Health Implications Extend Beyond Respiratory SystemThose with reduced lung function face multiple health risks. "They are also more vulnerable to poorer health generally," Hansell explained. "For instance, low lung function in adults is associated with the same level of risk of heart disease as having high cholesterol. Research has also shown that people whose lung health has been affected by air pollution may be at greater risk of heart disease."An earlier study found that air pollution was reducing the growth of children's lungs in east London. There, the average nine-year-old's lungs were between 90 and 100 millilitres smaller than they should be—approximately the volume of two hen's eggs.Studies on children in Sweden showed that lung growth increased when air quality improved. Reductions in air pollution might have also allowed the Bristol children's lung growth to return closer to normal rates.Call for Action on Air Quality ProtectionEminson concluded: "While the effects in individual children are small and unlikely to have immediate clinical consequences, they shouldn't be dismissed. Because lung function tends to track from childhood into adulthood, small differences early in life may have implications for long-term health. This reinforces the importance of reducing exposures and protecting children's environments."

The Rise of Machine-Generated Traffic Cloud infrastructure has long been designed around humans who search, click, scroll, and stream in a steady and predictable fashion. However, AI agents behave differently. They can unleash a swell of activity, spinning up multiple sub-agents that query hundreds of databases, search documents, and call APIs in seconds and then disappear as quickly as they arrived. AWS's Next-Gen OpenSearch Serverless Under that premise, Amazon is redesigning a core piece of its cloud infrastructure. On Thursday, AWS launched its next generation of OpenSearch Serverless, a fully managed search and vector database — essentially a system for storing and retrieving information at scale — that's designed specifically for agentic workloads. AWS says the new system can instantly scale up when agents trigger tasks and scale back down to zero when idle. The Data Analysis Cloudflare says bots accounted for 31% of overall HTTP traffic over the last six months. AI crawlers, search engines, and assistants made up roughly a quarter of all bot requests during that period. 'Non-human traffic will exceed human traffic sometime in the first half of 2027,' said Lai Yi Ohlsen, senior product manager at Cloudflare. The Impact Analysis The launch reflects a growing realization across the tech industry: Infrastructure originally designed for a human-driven internet doesn't work as well in a world increasingly populated by agents. As AI agents still represent a relatively small portion of internet activity, machine-generated traffic is already significant, and poised to grow. The Prediction As a result, cloud providers and infrastructure companies have been reckoning with how to adapt systems built for humans to a world of agents that are constantly and autonomously retrieving information, invoking tools, and generating machine-to-machine traffic. The more companies deploy AI agents, the more pressure there will be to redesign infrastructure around machine-generated workloads, which in turn could make agents cheaper and easier to deploy at larger scales.

The Rise of Recursive Self-Improvement in AIThe word "recursion" is the latest buzzword in AI circles. Two separate startups have taken on the name, and many more have started referencing recursive self-improvement (RSI) in their roadmaps. Like AGI before it, RSI has become a three-letter byword for a cataclysmic AI takeoff – even if there's still a little disagreement about what it exactly means.In basic terms, RSI refers to an AI system that can continuously upgrade itself. Once AI systems can manage the upgrade cycle better than humans, the process can become a closed loop, limited only by the compute power they can access, and humans are no longer necessary or even helpful.Scary or not, that's a vision that a lot of AI labs are eager to chase.Key Players Pursuing Recursive SystemsEarlier this month, well-known AI researcher Richard Socher launched the aptly named Recursive Superintelligence with RSI as an explicit goal. "Our main focus is to build truly recursive, self-improving superintelligence at scale," Socher told TechCrunch at launch, "which means that the entire process of ideation, implementation, and validation of research ideas would be automatic."A number of other prominent researchers are already chasing that same goal, hoping for a breakthrough that will make recursive self-improvement possible.One of the most prominent is Andrej Karpathy, a legendary figure from Tesla and OpenAI, who is using agent swarms to train LLMs on simple tasks for a project he calls Auto-Research. Karpathy has been unusually open about the project, tweeting about milestones regularly and making the building blocks available through a public GitHub repo. So far, the work has mostly been confined to making minor improvements on a GPT-2 scale model — as Karpathy noted in March, "It's not novel, ground-breaking 'research' (yet)" — but it's been enough to convince lots of other researchers to follow the RSI dream. And with Karpathy now working on pre-training at Anthropic, he will have plenty of opportunity to apply the idea at a larger scale.Adaption — founded by Cohere and Google alum Sara Hooker — recently launched a similar tool called AutoScientist in an effort to automate frontier training. Like Karpathy's auto-researchers, the system trains agents to make incremental improvements — but for Adaption, the goal is to make it easier to train a full-scale frontier model. If those same researchers start to push the frontier forward, the system could quickly spiral into something very much like RSI.Disarray founder Doris Xin drew more specific RSI interest when her self-trained machine learning agent took home 28 medals in a recent Kaggle competition, beating out many human-trained agents. As she sees it, the major challenge is reliability."I would argue, given infinite compute and infinite time horizon, we are already there," Xin told me. "I want to make an argument that this is not a creative endeavor, really. It's just a lot of meat-and-potatoes engineering."The Current State of Self-Improving AIThere's also plenty of evidence that the AI industry isn't very close to recursive systems in any meaningful way — and is still grappling with talking to a wary public about its progress. So Google CEO Sundar Pichai basically admitted in a recent podcast interview."It's a continuum, and we are all definitely making progress," Pichai said. "But in the way people describe RSI, that would represent a next level of acceleration and would have a lot of implications, but we aren't quite there yet."But the continuum includes an awful lot of self-improving AI systems.In January, one of Anthropic's lead programmers for Claude Code estimated that "close to 100%" of his team's code was written by the tool — a frank admission that Claude Code was literally writing itself.Just because engineers are using an AI tool doesn't mean the tool can replace them — but Anthropic seems to be getting close to replacing engineers too. In a recent survey tied to the Mythos preview, five out of 18 Anthropic engineers believed that, with harness improvements, this version of Mythos could soon substitute for an L4 engineer — a midlevel programmer who can take on involved projects without supervision.Still, there were some of the same weaknesses you might expect."Some of Claude's major reported weaknesses compared to an L4 include: self-managing week-long ambiguous tasks, understanding org priorities, taste, verification, instruction-following, and epistemics," the report reads.In other words, its weaknesses are everything involved with self-direction, which is the cornerstone for RSI. But sure, for everything else, Claude is ready to step right in.Expert Perspectives on RSI TimelinesJust like the AGI term before it, the AI industry also can't tell us how far away it is from showcasing a meaningful recursive system. When Georgetown's Center for Security and Emerging Technology assembled a group of experts to study RSI last year, the group found a major split in assessments — some expecting an imminent "superintelligence" style explosion while others expected slower progress and an eventual plateau. But all agreed that recursion made the future especially difficult to predict.Helen Toner, director of CSET and a former board member at OpenAI, told TechCrunch that simply using AI tools to do AI research isn't enough to qualify as RSI. "They're just using AI for as much as they can," Toner told TechCrunch. "And I think that is different from the classic definition of RSI, which is really that there are no humans needed."Toner pointed to a recent post by METR's Ajeya Cotra, which distinguishes different milestones on the path to the AI research takeover. One step, which Cotra calls "adequacy," would come when the system can still perform research after all humans are removed — even if the resulting research isn't as valuable or efficient. "Parity" comes when an AI-only system is as good at research as a human-only system. "Supremacy," the final stage, comes when an AI-only system outperforms a collaborative system between humans and AI.Ultimately, Cotra concludes that AI is very close to the adequacy threshold of being able to produce some work on its own — similar to the incremental changes made by Karpathy's Auto-Research system. "I wouldn't be totally shocked if you told me this milestone had already passed, and I expect it to happen in the next couple years," Cotra wrote.She was less clear on when parity will come, but once it does, she thinks it would "massively accelerate the pace of AI progress, leading to AI research supremacy within another year."The Challenges Ahead for Recursive AIWith so much of AI built on scaling laws, there's a strong tendency to think RSI will follow the same curve. Toner thinks that many of those pursuing AI research and development via RSI "think of it as a pretty smooth ladder, where you can just keep scaling up."But even if AI researchers are able to make incremental improvements like Karpathy's auto-researchers, there will be larger challenges in handing off the whole process of research. Toner put it in terms of the history of computing, which has seen human beings handing off more and more of the process while still directing things from the top."We went from machine languages to assembly language and compiled languages; you're getting further and further from the guts of the computer," Toner said. "But the human is still, in some intuitive sense, running the show."Moving beyond that paradigm will take significant challenges, both in engineering and alignment. But even with the massive investments happening, there's no infinite compute available — and the basic trade-off between human labor and machine intelligence will be hard to overcome.The Future of Recursive Self-ImprovementAs for a total recursive AI system of apocalyptic visions? The only thing researchers essentially agree on is that, like AGI, it's not here yet.

The LeadA novel featuring a protagonist whose job is taken by AI has won the Climate fiction prize. Hum by Helen Phillips, the American writer's third novel, explores themes of technological displacement and the commodification of nature in a dystopian future.The Disneyfication of Nature in LiteratureHum is about a woman, May, who loses her job to a "hum" of the title – a humanoid robot. Struggling to find work, she becomes a guinea pig for an experimental injection that alters her face so it can't be recognised by surveillance. When she gets paid for it, she splashes out on family passes to the Botanical Garden, the last remaining green space in her city. There, things take a turn for the worse.The Climate Fiction Prize RecognitionThe prize, worth £10,000, was first awarded last year to Abi Daré for And So I Roar, the follow-up to her bestselling debut The Girl with the Louding Voice. This year's judging panel was chaired by Guardian theatre critic and former literary editor of the Independent, Arifa Akbar. Alongside Kit de Waal and Friederike Otto on the panel were author Jessie Greengrass and book influencer Simon Savidge.Judges' Perspectives on the NovelJudge and writer Kit de Waal described Phillips's book as being about the "Disneyfication of nature … turning nature into a rare place that we have to pay to see". Fellow judge and climate scientist Friederike Otto added that it "tackles the central reason that nothing is done about the climate crisis – privilege", while writer Daisy Hildyard described it as "mesmerising and scary".The Author's InspirationPhillips was inspired to write the book after walking home from work one day and having the thought that she needed to buy dishcloths, before opening her computer at home and finding that dishcloths were being advertised to her. "That eerie feeling stuck with me, and I started to think about what worst-case scenarios might arise from surveillance by an algorithm."Impact on Climate AwarenessHum "helps us connect with what really matters and stops us from sleepwalking into an inevitable dystopia", said Lucy Stone, CEO of Climate Spring, which funds the prize. "In the novel, the machines themselves start to question the insane volume of advertising and the consumer treadmill, and then show the family that there are multiple different futures lying ahead of them."Future RecognitionPhillips will appear at Hay festival to discuss the book on Saturday 30 May. Alongside Phillips on this year's shortlist were Madeleine Thien for The Book of Records, Robbie Arnott for Dusk, Keshava Guha for The Tiger's Share, Susanna Kwan for Awake in the Floating City and Maria Reva for Endling.

The Sound of Summer and the Cost of SilenceThe 'lawnmower hum' is a defining auditory backdrop of the season, yet it has evolved from a mere annoyance into a significant source of legal liability. While some view the rip, chug, and whine of garden machinery as the 'music of May,' it frequently pits neighbours against one another, turning weekend relaxation into a battleground of noise complaints.The Legal Threshold: From Annoyance to FelonyThe conflict between maintaining a pristine lawn and respecting community peace has escalated to serious criminal levels. A recent case in Wiltshire highlights the extreme end of this spectrum, where a woman faces up to a year in jail for deliberately leaving her lawnmower running during a neighbour's dinner party. This incident was not spontaneous but the culmination of a 15-year feud and a violation of a restraining order.Weekday Hours: Generally 8am to 8pm are considered acceptable.Saturday Hours: Some councils recommend restrictions between 9am to 7pm.Sunday Hours: Stricter limits often apply, typically 10am to 7pm.The Financial Impact of Noise ViolationsLocal authorities possess the power to impose strict penalties on those who disregard these quiet hours. The financial cost of a single noise violation can be substantial, with fines reaching up to £5,000. This financial burden applies not only to the operator of the machinery but also to those using other noisy tools such as hedge trimmers, edge strimmers, and pressure washers that disturb the peace.The Social Shift in Suburban EtiquetteThe core of the issue lies in a clash of lifestyles: the hard-working individual who must mow on weekends versus the neighbour seeking relaxation. The irritation is compounded by the antisocial use of tools during sensitive times, creating a breakdown in mutual consideration. As urban density increases and gardens become smaller, the margin for error in noise management is shrinking.Future Outlook: Enforcing the PeaceAs legal frameworks tighten and community standards evolve, the expectation for 'quiet enjoyment' of a property is likely to become more rigid. Future disputes may see a rise in litigation rather than informal compromise, as the financial and legal risks of ignoring noise restrictions continue to mount.

The Challenge: Designing Escape RoutesThe puzzle presents a scenario with two snakes of equal width but different lengths trapped in a cage. The objective is to design two distinct escape passages, A and B, that allow one snake to pass while blocking the other.Passage A: Must allow the short snake to escape but block the long snake.Passage B: Must allow the long snake to escape but block the short snake.The Logic of the SolutionThe solution relies on exploiting the physical dimensions of the snakes. For Passage A, the design features a loop that is longer than the short snake but shorter than the long one. When the long snake enters the loop and doubles back, its body blocks the exit point, trapping it. The short snake, being shorter, can navigate the loop without obstruction.Passage B utilizes a floor hole. Assuming the snakes have non-zero rigidity, the short snake cannot stretch far enough to move over the hole without falling in, whereas the long snake can bridge the gap and pass safely.Why Spatial Reasoning MattersThis puzzle underscores the critical role of spatial intelligence in problem-solving. It demonstrates how understanding the relationship between length, width, and path constraints can create solutions that are counter-intuitive yet logically sound.The Future of Logic Puzzles in AIAs AI models continue to advance in spatial reasoning, puzzles like this will likely serve as benchmarks for testing the flexibility of machine intelligence. The future of puzzle design may shift towards scenarios that require not just calculation, but a nuanced understanding of physical constraints.

The LeadApple's latest MacBook Air is its most powerful yet, comes with double the starting storage and is better than ever for getting work done and as the benchmark for a consumer laptop. The Event DetailsThe M5 MacBook Air starts at £1,099 (€1,199/$1,099/A$1,799) for the 13in version, which is £100 or equivalent more than last year's excellent M4 version, but comes with at least 512GB of storage. It sits above the £599 MacBook Neo and below the £1,699 M5 MacBook Pro, making the Air Apple's mid-range machine. The Data Analysis Screen: 13.6in LCD (2560x1600; 224 ppi) True Tone Processor: Apple M5 with eight or 10-core GPU RAM: 16, 24 or 32GB Storage: 512GB, 1, 2 or 4TB SSD Operating system: macOS 26 Tahoe Camera: 12MP centre stage Connectivity: wifi 7, Bluetooth 6, 2x Thunderbolt/USB 4, headphones Dimensions: 215 x 304.1 x 11.3mm Weight: 1.23kg The Impact AnalysisThe new M5 chip marks a watershed moment for Apple's laptop line. It is about 10-20% faster than the M4 in the previous edition, which is nothing to be sniffed at. But with the progress over the last few years, the M5 makes this MacBook Air between 75% and 108% faster than the M1 MacBook Air depending on the task. The PredictionThe MacBook Air M5 is a top-notch consumer laptop that offers pro-level performance, long battery life, and sustainability features. With its improved performance, storage, and features, it is likely to remain a top choice for consumers in the market for a reliable and powerful laptop.

Opening the Japanese Horizon: Playground’s Deep‑Dive ResearchPlayground Games set out to make Japan the centerpiece of Forza Horizon 6, a task that required years of on‑the‑ground research, cultural consultancy and cutting‑edge imaging technology.On‑Location Photography and 360‑Degree Scanning for AuthenticityThe design team dispatched crews across the country to capture thousands of photos, hours of video and detailed sky captures. They partnered with cultural consultant and former Porsche ambassador Kyoko Yamashita, who spent three years advising on signage colours, local nuances and the broader racing scene. Additional expertise came from Kyoto‑based bodyshop Rocket Bunny and car‑culture photographer Larry Chen, who also fronted the YouTube series “Art of Driving”. Crucially, the team employed 360‑degree cameras, allowing them to record environments in both 2D and 3D – a process the developers liken to building their own Google Maps.Release Timeline and Market PositioningForza Horizon 6 launched on 19 May 2026. Its Japan map is the largest in the franchise to date, blending iconic roads such as Hakone Nanamagari and Mount Haruna with detailed urban districts that mirror Shibuya, Akihabara and suburban Tokyo. The inclusion of classic drift icons – the Nissan Skyline, Toyota Supra and Mazda RX‑7 – reinforces the game’s appeal to both racing enthusiasts and fans of Japanese car culture.Redefining Player Expectations of Japan in Open‑World RacingBy moving beyond the stylised, fragmented portrayals common in titles like Persona 4, Shenmue and the Yakuza series, Playground aims to deliver a version of Japan that feels lived‑in and recognisable. Early feedback from Japanese streamers praised the accuracy of locations such as the Tokyo railway station, suggesting the research paid off. The focus on “car‑culture‑adjacent” details – vending machines, roadside temples, grassroots time‑attack circuits – adds layers of authenticity that could shift how gamers visualise the country.Future Directions for Forza Horizon’s Global SettingsHaving proven that exhaustive fieldwork and 360‑degree imaging can produce a convincing Japanese environment, the studio is likely to apply the same methodology to future locales. Expect deeper collaborations with local consultants, more immersive scanning techniques and an ongoing push to align virtual worlds with real‑world cultural expectations.