Stefaan Verhulst
Article by Madeleine I. G. Daepp, Kiran Tomlinson, Scott Counts & Siddharth Suri: “Knowledge work has been key to economic flourishing in most advanced and many emerging economies in the last half century. Defined by the synthesis and creation of ideas rather than the production of physical goods, knowledge work involves the processing of non-routine problems that require judgment-based and creative intellectual capabilities. Such work is a large and important component of contemporary economies, accounting for an estimated third to half of all jobs in high-income countries and a fifth of all jobs globally. Achieving sustained economic growth increasingly depends on the ability to leverage and create knowledge, with countries actively seeking to transition to knowledge economies to improve their economic outcomes. Knowledge work is also the foremost application for which workers are using generative artificial intelligence (AI). A critical question for the future of twenty-first-century economies, then, is whether generative AI could democratize knowledge work by expanding the set of people who can engage in and benefit from it.
Generative AI’s effect on knowledge work hinges on emerging challenges along those two dimensions: (1) who benefits from AI’s use and (2) who actually uses AI. In this Perspective, we synthesize recent empirical work to map out these challenges and describe both technical and policy interventions to mitigate harms and ensure that benefits are widely shared. Technological and institutional fixes will need to be developed in tandem. Policies will need to be calibrated—towards either sharing productivity gains or building skills—according to what current models enable, and tooling will need to be made broadly usable if AI literacy and adoption pushes are to be effective in closing persistent participation gaps…(More)”.
Article by John Fell, Sándor Gardó, Domenic Kellner, Benjamin Klaus, Jan Hannes Lang, Lukas Nagy, Pucho Vendrell, Marek Rusnák, Jonas Wendelborn and Stefan Wredenborg: “Financial stability communication is challenging because its task is not to forecast financial crises, let alone predict their precise timing. Rather, it is to identify vulnerabilities and explain how the financial system is likely to fare should it be confronted with adverse shocks. Great care is needed in this endeavour, because the sentiment of financial stability communication can influence market perceptions and risk assessments, as well as broader economic and financial outcomes. Given the presence of this potential feedback loop, the task of financial stability communication at the ECB has long been guided by a broad concept of financial stability: the smooth allocation of financial resources, effective management of risk by financial institutions and the capacity of the financial system to absorb shocks. Using the messages conveyed in the ECB’s Financial Stability Review over two decades, this special feature compares dictionary-based, FinBERT and prompt-based AI approaches to extracting financial stability sentiment. It finds broad co-movement across methods, while the GPT-based filter isolates sentences that contain explicit risk assessments, capturing subtle shifts in tone and context that were previously difficult to quantify. Used carefully, such tools can support risk monitoring and drafting consistency over time, but they remain complementary to expert judgement, vulnerability analysis and stress testing, rather than substitutes for it. A deep-dive box in the special feature also shows how AI can be used to systematically extract information from financial news to create an indicator for the severity and probability of triggers (SPOT) for financial stability risks…(More)”.
Report by the Asia Society: “…surfaced nine factors that define the conditions that national strategies must get right to enable responsible and rapid AI adoption: trusted datasets, AI infrastructure, AI skills and awareness, global AI value chain leverage, ethical AI development, misinformation governance, AI governance frameworks and institutions, environmental sustainability, and cybersecurity. Each factor represents a domain in which the absence of measurable progress creates conditions that can stall adoption, erode public confidence, or concentrate power in ways that undermine the broader ecosystem. The following sections elaborate on each metric. The discussion is grounded in both policy analysis and stakeholder consultations, so that policymakers can identify where the most significant trust gaps lie.
The metrics proposed here are not yet available as standardized, comparable data across Asian economies. However, the policy analysis and stakeholder consultations reveal consistent patterns in legislation, governance commitments, institutional design, and stated priorities, providing a foundation from which measurement frameworks can be built. The value of naming these metrics lies precisely in making that construction a shared and deliberate project, rather than leaving it to occur ad hoc…(More)”.
Article by Pokere Paewai: “A new decentralised data storage network will put Māori data in Māori hands with the goal of ensuring Māori sovereignty doesn’t “stop at the server door”.
Designed by Te Kāhui Raraunga, Te Pā Tūwatawata will be available to marae, hapū, iwi or other organisations who wish to store their data within the protection of the Pā.
Principal advisor Erena Mikaere said it was a commercial storage service designed specifically to meet the needs of iwi Māori, hapū and marae.
The project was built on open source technology and led by Māori scientists, Māori engineers and grounded in tikanga Māori, she said.
“Central to everything from its architecture, to its initial conceptions, to the values that drive it, and then also to our customer service delivery, it’s really about doing things in a very Māori way, based on a Te Ao Māori worldview. And so to that end, we didn’t just want to offer like an automated store with us and push this button and register your name and company and here’s the invoice type of style. It starts with a conversation, it starts with a kōrero, like all good things. And so that means that we can provide them with a really tailored service.”
Te Pā Tūwatawata provides end-to-end encryption of data, both in transit and at rest, which Mikaere said would mean only the group who submitted the data to the platform would have the “keys” required to decrypt it.
“What it does is it provides a safe place for some of our data that we might consider, or that whānau and hapū, iwi might consider are some of our most sensitive sets of mātauranga. It provides a way in which we can protect that and ensure extra restriction, say over another data set, which perhaps isn’t as sensitive.”..(More)”.
Article by Rina Chandran: “Since OpenClaw burst onto the scene as Clawdbot last November, individuals and businesses have embraced artificial intelligence agents to write code, send emails, run a shop, and more. AI agents are forecast to become ubiquitous in the coming years, raising concerns about agentic inequality, and its economic consequences for companies, countries, and people.
AI agents are built on top of large language models, and can reason and take actions to complete tasks on behalf of users. They have been touted as a way to do repetitive and mundane tasks to free up workers’ time for higher-value activities. Many agents still fail at the most basic tasks, and some perform unauthorized actions, yet big tech firms including Google, Amazon, Anthropic, and Perplexity are launching agents that can do increasingly complex tasks autonomously.
As AI agents become more integrated into the economy, companies and entities that deploy them will benefit disproportionately compared to those that cannot, Nick Srnicek, a senior lecturer in digital economy at King’s College London, told Rest of World.
“We will see new inequalities of access, scale, quality and trust: divides between those who have agents and those who don’t; those who have good agents and those who have bad agents; those who have many agents and those who have few agents; and those who can trust their agents and those who cannot,” he said…(More)”.
Paper by Paris21: “…argues that the statistical community has reached a fork in the road. Incremental adjustment alone may no longer be sufficient. Countries and the international community face a strategic choice between two broad paths, each with distinct implications for legitimacy, financing, risk and equity. Rather than prescribing a single solution, the paper provides a framework to facilitate informed debate among all stakeholders of official statistics as they chart a course towards more sustainable and inclusive data systems. It is also the foundation for a forthcoming series of policy briefs that will explore specific aspects of this theme in greater depth and help sharpen policy attention to key areas.
This is a decisive moment for official statistical systems.
Deep cuts in development financing for statistics, legitimacy issues, artificial intelligence and other rapid technological changes, and rising expectations for more inclusive and participatory data are colliding with long-standing weaknesses in trust, capacity and data use. For many national statistical offices (NSOs), particularly in low- and middle-income countries, this convergence amounts to a systemic data crisis that threatens their relevance, credibility and sustainability. At the same time, these pressures create a rare opportunity to rethink how data systems are designed, governed and embedded in society…(More)”.
Article by Adele Peters: “Last October, days before Hurricane Melissa slammed into Jamaica, it wasn’t obvious how quickly the storm would intensify or the path it would take. But inside Google, an experimental AI model was spinning through dozens of scenarios, including the possibility that it might be the strongest hurricane on record to hit the island.
Five days before the storm made landfall—while traditional weather models were undecided on whether it would weaken and turn in another direction—the AI model, called WeatherNext, predicted with 80% confidence that Melissa would rapidly intensify from a Category 1 storm to a Category 5 and land in Jamaica. Google sent its predictions to the U.S.’s National Hurricane Center, which used the models to help make a record-breaking high-intensity forecast.
That early forecast “was critical,” says Evan Thompson, principal director of the Meteorological Service of Jamaica. “We want to get the information as soon as possible and then continuously drill that message to the public.” A Category 5 hurricane had never made landfall on the island. The weather office warned residents that anything they had experienced before “would pale in comparison,” Thompson says, and urged people to prepare however they could…
The Google DeepMind model was more accurate than any other model the National Hurricane Center used during the storm. Now, as the new hurricane season begins on June 1, the NHC will work with Google again. Last year, the model ran a set of 50 possible futures every six hours; this year, it will look at 1,000 futures every six hours, making it even more likely that it can predict unusual storms. “This significant increase should provide more stable and consistent guidance,” says Philippe Papin, NHC senior hurricane specialist.
Google is one of several companies working to use AI to reshape forecasting as weather becomes more extreme. That includes other large tech companies like Microsoft, Nvidia, and Huawei, and startups like Atmo, Tomorrow.io, and WindBorne—some of which are also collecting better data through cheap satellites or redesigned weather balloons…(More)”.
Book by Joseph Hilger, Lulit Tesfaye, and Zachary Wahl: “The semantic layer framework is the practical combination of core technologies and best-in-practice design principles that deliver a framework connecting all of an organization’s knowledge assets in context and with the greatest accuracy. It combines elements from data, information, and knowledge management in order to enable the reliable (re)use of an organization’s assets.
This book guides the reader through every element of the semantic layer, from its core definition and business value, associated components and design requirements, methodological keys to success, realworld success stories and applications, and finally, through its relationship with AI. The book is divided into five parts. “Chapter I: Understanding the Semantic Layer” defines the semantic layer, explains its value, and details business outcomes and potential areas of return on investment. “Chapter II: Semantic Layer Component Design and Implementation” breaks the semantic layer down into its core components (knowledge assets, business glossary, metadata, taxonomy, and knowledge graph) defining each, explaining their role in the semantic layer, discussing associated technologies, and delving into the modeling of these components. “Chapter III: Methodologies for Designing and Integrating the Semantic Layer” delivers a step-by-step guide to the design and implementation of a semantic layer, covering a range of best practices to ensure the success of an initiative. “Chapter IV: Real-World Success Stories” expands on various real-world examples, presenting case studies from a range of organizations that have successfully implemented a semantic layer and detailing the use cases, approaches, and outcomes. Finally, “Chapter V: Powering Artificial Intelligence and Beyond” covers the interrelationship between the semantic layer framework and AI, elaborating on the ways in which these two fields can mutually benefit one another and concluding with a look into the future on how mature organizations are evolving their semantic ecosystems to drive responsible AI adoption and further competitive capabilities…(More)”.
Article by The National Archives: “Re-using public sector information is a government priority, one the UK Industrial Strategy estimates could unlock billions in economic growth. Yet publishing data openly, and enabling its re-use, is a process full of friction for public institutions. As the authority on licensing, The National Archives and their digital and policy teams partnered with IF to understand why the licensing framework is challenging to use today, prototype a redesign, and recommend how The National Archives can take a leadership role in closing the gaps so that people can use it with confidence…(More)”.
Article by Sarah O’Connor and John Burn-Murdoch: “In the UK, the beleaguered Labour government has high hopes that AI can help to deliver better, faster and cheaper public services. Just this week, the FT’s public policy editor Chris Smyth reported on a draft proposal to scale back recruitment plans for the NHS, which is partly premised on the expectation that much wider use of AI can boost productivity and even “completely substitute” for some (unspecified) roles.
But how easy is it to insert AI into public services? And how will we know if it’s working?…
Here in the UK, adoption of AI tools is gathering pace across many parts of the public sector. As far back as early 2024, the National Audit Office found that almost three-quarters of government departments and national bodies were either deploying or piloting AI tools. Some of these deployments are very narrow. The Department for Transport, for example, is trialling an AI tool to prevent fraud when people apply for subsidies for electric vehicle chargers, by checking “proof of installation” photos against similar or stock photos. Others are more widespread and seem to have developed quite organically, such as the rapid adoption of AI transcription tools by social workers to record meetings and save time typing up notes.
But how do we know if any of this is making a difference when it comes to productivity?..(More)”.