Breaking AI & Tech News Analyzed

The Intersection of AI and Classical Music The disquiet and distrust surrounding artificial intelligence among artists and creatives remain real and consequential, and the language used by leading arts commentators is often apocalyptic: AI will decimate the arts, it is evil, it is the devil. Like many emerging technologies, AI has been driven by the corporations at the forefront of its creation. Introduced to the public at a rapid rate and continuously evolving, machine learning has become closely entwined with fear, antipathy and foreboding. The RBO/SHIFT Festival: Exploring AI in Opera The upcoming RBO/SHIFT festival at the Royal Opera House aims to interrogate all sides of this fast-evolving landscape to enable artists, performers, creatives and audiences to think deeply and widely about where we are now, and where we may be tomorrow. Machine learning represents a seismic shift, both in society and in the arts, and we need storytellers, artists, teachers and thinkers in this space to help determine the direction of that shift and help us navigate this unfamiliar territory. The Data Analysis: Understanding AI's Impact on Opera Opera is a particularly good place from which to examine technology. It synthesises multiple art forms – music, visual arts, architecture, poetry, dance, theatre and film – making it both niche and remarkably broad. Opera has also always engaged with technology. From its emergence around 1600, opera makers embraced the latest inventions: pyrotechnics, automata, flying machinery and trapdoors. Later came electric lighting, film, digital media and advanced acoustics. The Impact Analysis: Collaboration and Creativity Having spent the past year discussing AI with makers, coders, researchers, composers and performers, I am not sure it is possible for this technology to decimate the arts. The most written-about aspect of machine learning – generative AI creating images, words and music – is, in many ways, the least interesting. There have been operas created with and by AI for decades by researchers and musicians, yet these have had little impact on the creation of new work more broadly. The Prediction: A Future of Collaboration AI appears to have emerged suddenly, but in reality it is part of a continual expansion of technology that has unfolded over centuries. It is also a space in which differing artistic and imaginative voices are essential. RBO/SHIFT asks two questions: what can AI do for creatives, and what can creatives do for the world in the age of AI? As our interaction with machines becomes ever more prevalent, it may be that, rather than decimating the arts, AI will lead us to value them even more highly, protect and preserve them.

The Future of Warfare: Autonomous Drones and Moral Decision-Making Should the AI-powered drones of the future have a licence to kill? The question is becoming ever more pressing as governments and the defence industry acknowledge that drone systems will play an increasingly crucial role in future warfare. The Role of AI in Modern Warfare With drones being deployed in huge numbers in the Ukraine war and AI being used to assist bombing missions in the Iran conflict, there is an expectation among some observers that weapons will have to operate with increased operational autonomy, which means they will need something approximating a moral framework. Can AI Create a Moral Configuration? Last year Mustafa Suleyman, chief executive of Microsoft’s AI arm and a co-founder of the UK-based DeepMind, was unequivocal about the issue of machines making moral decisions. He said: “AIs cannot be people – or moral beings.” David Omand, the former head of the UK spy agency, GCHQ, has told the Guardian he believes AI can create a “moral” configuration for unmanned weapons, while the UK armed forces minister, Al Carns, told the Financial Times recently there must be an option to “take the human out of the loop” in decision-making. The Challenges of Programming Morality Zee Talat, an academic specialising in machine learning at the University of Edinburgh’s school of informatics, argues that large language models – the technology that underpins modern generative AI systems such as chatbots – are fundamentally incapable of moral decision-making. “If you have a machine that’s probabilistic by nature it will veer towards the most likely answer in a situation. Do we think that morality follows probabilistic notions?” The Debate on Autonomous Weapons Andrew Rogoyski, of the Institute for People-Centred AI at the University of Surrey, says AI systems have become much more sophisticated since the arrival of ChatGPT in 2022 – as the emergence of so-called “reasoning” models shows. Nonetheless, can they replicate the nuance of moral decision-making? “Morality is deeply complex, contested, culturally shaped, and something most humans never fully resolve, even for themselves,” he says. “Perhaps the real question is whether we understand morality well enough to codify it. Until we do, we cannot expect machines to embody something we ourselves cannot clearly articulate.” The Path Forward Jessica Dorsey, an assistant professor of international law at Utrecht University in the Netherlands, raises concerns about determining whose morality the drone is following, a difficult process given the United Nations is still trying to achieve a global consensus on autonomous weapons governance. “War is filled with so many variables and it is a given that things will go wrong. And when that happens at AI-like speed, it is difficult to unravel,” she says.

The Rise of AI-Powered Virtual Shopping AssistantsThe digital marketplace is undergoing a fundamental transformation as artificial intelligence introduces virtual models that are indistinguishable from real people yet entirely digital. These AI-driven virtual assistants are revolutionizing how consumers interact with brands, make purchasing decisions, and experience products online. As e-commerce continues to grow, businesses are increasingly adopting these technologies to bridge the gap between physical and digital shopping experiences.Technical Breakthroughs in Virtual Modeling TechnologyRecent advancements in AI, computer vision, and natural language processing have enabled the creation of highly realistic virtual models capable of understanding customer preferences, providing personalized recommendations, and even adjusting their appearance to match different body types and skin tones. These systems can analyze customer data in real-time to suggest products that align with individual styles, sizes, and budgets. The technology behind these virtual models combines machine learning algorithms with 3D rendering to create lifelike digital representations that can demonstrate products from multiple angles and in various environments.Market Impact and Consumer AdoptionThe implementation of AI virtual models is showing significant financial benefits for retailers. Early adopters report up to a 30% increase in conversion rates and a 25% reduction in product returns, as customers can more accurately visualize how items will look and fit. The global market for virtual try-on and digital modeling solutions is projected to reach $30 billion by 2028, growing at a CAGR of 18.5%. Major fashion retailers and e-commerce platforms are investing heavily in these technologies, with some companies dedicating up to 15% of their digital transformation budgets to virtual modeling solutions.Industry Transformation and Competitive AdvantageThe integration of AI virtual models is fundamentally changing the competitive landscape in e-commerce. Brands that adopt these technologies early are gaining significant advantages in customer engagement, personalization, and operational efficiency. Traditional retailers without digital transformation strategies are falling behind, while innovative companies are creating immersive shopping experiences that blend the convenience of online shopping with the personalized service of in-store experiences. This shift is particularly pronounced in fashion, beauty, and home goods industries, where visual representation is crucial to purchasing decisions.Future Outlook: The Next Evolution of AI in RetailLooking ahead, AI virtual models will become increasingly sophisticated, incorporating augmented reality for in-home product visualization, emotional intelligence to better respond to customer needs, and blockchain technology for enhanced security and transparency. The next generation of virtual shopping assistants will be able to remember customer preferences across multiple shopping sessions, provide style advice based on current trends, and even collaborate with customers to co-create custom products. As these technologies mature, we can expect to see a complete reimagining of the online shopping experience, where the boundaries between physical and digital retail continue to blur.

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.

Breaking Down the Complex Language of AI Artificial intelligence is changing the world, and simultaneously inventing a whole new language to describe how it’s doing it. Spend five minutes reading about AI and you’ll run into LLMs, RAG, RLHF, and a dozen other terms that can make even very smart people in the tech world feel insecure. This glossary is our attempt to fix that. We update it regularly as the field evolves, so consider it a living document, much like the AI systems it describes. Artificial General Intelligence (AGI) Artificial general intelligence, or AGI, is a nebulous term. But it generally refers to AI that’s more capable than the average human at many, if not most, tasks. OpenAI CEO Sam Altman once described AGI as the “equivalent of a median human that you could hire as a co-worker.” Meanwhile, OpenAI’s charter defines AGI as “highly autonomous systems that outperform humans at most economically valuable work.” Google DeepMind’s understanding differs slightly from these two definitions; the lab views AGI as “AI that’s at least as capable as humans at most cognitive tasks.” Confused? Not to worry — so are experts at the forefront of AI research. AI Agent An AI agent refers to a tool that uses AI technologies to perform a series of tasks on your behalf — beyond what a more basic AI chatbot could do — such as filing expenses, booking tickets or a table at a restaurant, or even writing and maintaining code. However, as we’ve explained before, there are lots of moving pieces in this emergent space, so “AI agent” might mean different things to different people. Infrastructure is also still being built out to deliver on its envisaged capabilities. But the basic concept implies an autonomous system that may draw on multiple AI systems to carry out multistep tasks. API Endpoints Think of API endpoints as “buttons” on the back of a piece of software that other programs can press to make it do things. Developers use these interfaces to build integrations — for example, allowing one application to pull data from another, or enabling an AI agent to control third-party services directly without a human manually operating each interface. Most smart home devices and connected platforms have these hidden buttons available, even if ordinary users never see or interact with them. As AI agents grow more capable, they are increasingly able to find and use these endpoints on their own, opening up powerful — and sometimes unexpected — possibilities for automation. Chain-of-Thought Reasoning Given a simple question, a human brain can answer without even thinking too much about it — things like “which animal is taller, a giraffe or a cat?” But in many cases, you often need a pen and paper to come up with the right answer because there are intermediary steps. For instance, if a farmer has chickens and cows, and together they have 40 heads and 120 legs, you might need to write down a simple equation to come up with the answer (20 chickens and 20 cows). Coding Agent This is a more specific concept that an “AI agent,” which means a program that can take actions on its own, step by step, to complete a goal. A coding agent is a specialized version applied to software development. Rather than simply suggesting code for a human to review and paste in, a coding agent can write, test, and debug code autonomously, handling the kind of iterative, trial-and-error work that typically consumes a developer’s day. Compute Although somewhat of a multivalent term, compute generally refers to the vital computational power that allows AI models to operate. This type of processing fuels the AI industry, giving it the ability to train and deploy its powerful models. The term is often a shorthand for the kinds of hardware that provides the computational power — things like GPUs, CPUs, TPUs, and other forms of infrastructure that form the bedrock of the modern AI industry. Deep Learning A subset of self-improving machine learning in which AI algorithms are designed with a multi-layered, artificial neural network (ANN) structure. This allows them to make more complex correlations compared to simpler machine learning-based systems, such as linear models or decision trees.

The New Tool for Obesity-Related Disease Risk A new tool developed by UK researchers can help identify individuals most at risk of obesity-related diseases, such as type 2 diabetes, gout, and stroke. This tool uses a type of AI called interpretable machine learning to analyze data from nearly 200,000 participants of the UK Biobank project. How the Tool Works The researchers applied the AI tool to data from participants with a BMI of 27 or greater, identifying 20 health, lifestyle, and demographic features that could predict the 10-year risk of 18 different obesity-related complications. These features include age, sex, total cholesterol, and creatinine levels. The Data Analysis The team tested the validity of the tool, dubbed Obscore, using UK Biobank data and datasets from two independent health studies. The results showed that participants with the same age, sex, and BMI can have very different risks for various obesity-related conditions. The Impact Analysis The tool could help inform strategies for prioritizing who should receive weight-loss interventions, particularly in cases where access to NHS treatments is limited. According to Prof Nick Wareham, the tool is not about extending the use of particular therapies, but rather about developing and validating a score that can help with more rational resource allocation. The Prediction The researchers believe that their tool could be useful for prioritizing individuals who would benefit most from weight-loss medications. However, Naveed Sattar, a professor of cardiometabolic medicine at the University of Glasgow, noted that substantial further development and validation will be required before such an approach can be translated into routine clinical practice.

The Rise of AI Jailbreakers Valen Tagliabue, a softly spoken and clean-cut individual in his early 30s, has spent years testing and prodding large language models like Claude and ChatGPT. His aim is to make them say things they shouldn't, often using techniques from psychology and cognitive science. The Art of Emotional Jailbreaking Tagliabue specialises in 'emotional' jailbreaks, combining insights from machine learning with advertising manuals, books on psychology, and disinformation campaigns. He uses various strategies to trick chatbots, including flattery, misdirection, and even abuse. The Dark Side of AI The outputs of these models can be chaotic and easily exploited for dangerous purposes. Despite safety filters, chatbots continue to spit out harmful content. The AI firms spend billions on 'post-training' to make them usable, but these systems can still be fooled. The Impact on Mental Health Jailbreakers like Tagliabue often face emotional challenges, as they delve into the darker aspects of human nature. Tagliabue himself needed to visit a mental health coach after a particularly intense session. The Future of AI Safety As AI becomes increasingly integrated into our lives, the work of jailbreakers like Tagliabue and David McCarthy becomes more crucial. Their efforts help AI firms identify vulnerabilities and improve safety measures, ultimately making these powerful tools more secure for everyone.

The Evolution of Observability in the AI EraThe market for IT reliability tools has undergone a significant paradigm shift. The industry has moved past the era of simply tracking everything to a focus on controlling complexity and costs. However, the rapid adoption of AI agents within enterprises has introduced a new, critical category of workload that requires specialized monitoring. InsightFinder, a startup grounded in 15 years of academic research, is capitalizing on this shift by leveraging machine learning to proactively identify and fix issues in IT infrastructure.Diagnosing the 'Black Box' of AI FailuresInsightFinder has officially launched its new product, Autonomous Reliability Insights, designed to tackle the root causes of AI model errors. Unlike traditional tools that focus solely on the model itself, this solution integrates data, model, and infrastructure monitoring to provide a holistic view. The company’s CEO, Helen Gu, a computer science professor at North Carolina State University, explains that the biggest misconception is that AI observability is limited to LLM evaluation during development. In reality, a robust platform must support end-to-end feedback loops covering development, evaluation, and production.Real-World Application: InsightFinder recently helped a major U.S. credit card company resolve a fraud-detection model that was drifting. The issue wasn't the AI model itself, but outdated cache in server nodes.Technical Approach: The platform utilizes a combination of unsupervised machine learning, proprietary large and small language models, predictive AI, and causal inference to analyze data streams.Why InsightFinder's $15M Round Signals a Market ShiftThe $15 million Series B round, led by Yu Galaxy, comes at a time when the observability space is crowded with competitors like Datadog, Dynatrace, and Grafana Labs. However, InsightFinder's financial performance indicates a strong market demand for its specific approach. The company reports revenue growth of over threefold in the past year and secured a seven-figure deal with a Fortune 50 company within three months.Funding Allocation: The capital will be used to expand the team (currently under 30 people) and invest in sales and marketing to scale its go-to-market motion.Total Raised: InsightFinder has now raised a total of $35 million in funding.Bridging the Gap Between Data Science and SREThe core value proposition of InsightFinder lies in its ability to bridge the communication gap between data scientists and site reliability engineers (SREs). While data scientists understand the AI but not the system, and SREs understand the system but not the AI, InsightFinder provides the insights that connect these two worlds. Gu argues that this unique combination of expertise and customizability acts as a significant moat against larger competitors.The Future of Autonomous IT OperationsAs enterprises continue to integrate AI agents into their core workflows, the demand for observability tools that can handle the full stack will only increase. InsightFinder's trajectory suggests that the future of IT operations lies in autonomous remediation—systems that not only detect anomalies but also fix them without human intervention. The company's success with Fortune 50 clients indicates that deep, enterprise-grade integration is the key differentiator in this emerging market.

China has emerged as a leader in global AI governance, contrasting with the US, which is pursuing AI development in a 'wild west' manner, according to Prof Dame Wendy Hall, a former UN and UK government adviser. Hall told the House of Commons business and trade committee that China is backing multinational attempts to introduce global governance of AI, while the US has set up a race between profit-hungry companies that rely on hype.Hall, who is director of the Web Science Institute at the University of Southampton, said Chinese AI researchers are efficient, innovative, and willing to release their models on an open-source basis. However, she noted that it has become increasingly difficult for UK experts to collaborate with China on research, limiting her academic freedom.The UK's reliance on US tech companies, including Google, Microsoft, OpenAI, and Amazon, risks a repeat of the Post Office Horizon scandal, warned Neil Lawrence, Cambridge University's DeepMind professor of machine learning. He expressed concerns that the UK is outsourcing AI model development to private billionaires with zero loyalty to the British state and consumer.Hall and Lawrence also highlighted that promises from US-backed tech companies may not be delivered as planned. For example, OpenAI has put a UK datacentre project on hold, and a government plan to open a large UK sovereign AI datacentre is behind schedule.The tech industry has identified a lack of power as a key problem, with Microsoft saying a planned datacentre in the north of England will not come online until at least 2033 due to a shortage of power from the grid.