Fri 12 June 2026
Daily Brief — Curated and contextualised by Best Practice AI
AI Agents Reshape Work, Apollo Screens Investments, and India's Caste Divide Widens
TL;DRAI agents are transforming knowledge work by executing tasks autonomously, enhancing efficiency and scope. Apollo Global Management is evaluating software investments for AI disruption risks to mitigate obsolescence concerns. In India, a steep caste gradient in AI exposure among graduates highlights inequality in the labor market. Meanwhile, US insurance regulators are probing credit risks tied to data centers as AI infrastructure grows.
The stories that matter most
Selected and contextualised by the Best Practice AI team
How AI Agents Reshape Knowledge Work: Autonomy, Efficiency, and Scope
arXiv:2606.07489v2 Announce Type: replace-cross Abstract: Frontier AI systems are bridging the gap between intelligence and utility by shifting from conversational assistants to autonomous agents that execute tasks end to end. Using production data from Perplexity's Search and Computer products, we study this transition by examining how AI agents accelerate and reshape knowledge work. Three key empirical findings emerge. First, using sessions with near-identical initial query pairs as natural experiments for the same underlying task attempted with both products, Computer performs 26 minutes of autonomous work per user session, versus 33 seconds for Search. Computer automates task decomposition and execution that Search users might otherwise manually orchestrate and implement. As a result, Computer shifts follow-up query distribution toward higher-order work such as verification and extension. Autonomy also increases execution quality, with per-query dissatisfaction rates 55% lower on Computer than on Search. Second, due to its autonomy advantage, Computer reduces completion time from 269 to 36 minutes on matched tasks, lowering estimated time and cost by 87% and 94%, respectively, compared to humans equipped with Search alone. Third, Computer changes the scope of work that users attempt: Computer queries more often cross occupational boundaries, require higher-order cognition, draw on broader expertise, take the form of composite tasks that bundle interdependent subtasks into a single query, and unlock work activities that are essentially absent from Search usage among the same users. Together, the evidence indicates that AI agents accelerate workflows, enhance output quality, reduce costs, and expand the breadth and depth of automated work.
The Privilege of Exposure: Caste and Generative AI in India's Graduate Labour Market
arXiv:2606.13314v1 Announce Type: new Abstract: Who is exposed to generative AI in a developing-country labour market? We map three occupational AI-exposure indices to India's redesigned Periodic Labour Force Survey (2025) and document a steep caste gradient among 83,000 employed graduates: graduates from the Scheduled Castes and the Scheduled Tribes are 0.24--0.37 standard deviations less exposed than upper-caste graduates within the same district. Two channels drive the gap: one in four SC and one in three ST graduates work in farm or elementary occupations untouched by AI, and those in white-collar work are underrepresented in managerial, software, and finance occupations. Because exposure commands a wage premium of up to 20 per cent, generative AI stands to widen, not narrow, India's caste earnings gap.
Apollo Is Screening All Software Investments for AI Threat Risk - Bloomberg
Apollo Global Management Inc. is assessing every new investment opportunity in the software industry for AI disruption risk, as asset managers try to quell investor concerns that rapid advances in the technology could make businesses obsolete.
Revisiting the ABCs of Working with AI: A Replication with Radiologists
arXiv:2606.12585v1 Announce Type: new Abstract: Artificial intelligence (AI) systems increasingly assist human experts, but the consequences of AI assistance on productivity can be heterogeneous. Caplin, Deming, S. Li, Martin, Marx, Weidmann, and Ye (2025b) provide evidence that two characteristics, ability and belief calibration, help to determine the returns to AI assistance. This note shows that their results replicate to a setting where professional radiologists analyze chest X-rays with access to state-of-the-art machine learning predictions. I leverage the public Collab-CXR data repository described by Moehring, Kutwal, Huang, Banerjee, Jacobi, Eber, Mendoza, Chung, Dayan, Gupta, Bui, Truong, Pareek, Langlotz, Lungren, Agarwal, Rajpurkar, and Salz (2025) and first analyzed for human-AI collaboration by Agarwal, Moehring, Rajpurkar, and Salz (2023). To faithfully reproduce the analysis in Caplin, Deming, S. Li, Martin, Marx, Weidmann, and Ye (2025b), I use the radiologist assessments from the repeated-case designs, which include 68 radiologists and 11,420 paired radiologist-patient-pathology observations. The results of this replication support the external validity of their core findings: lower baseline ability and higher calibration predict larger incremental value from AI.
US insurance rulemaker probes credit risks tied to data centres
Association’s efforts come as capital from the sector plays a growing role in AI infrastructure build-outs
Absent From the SpaceX and OpenAI I.P.O.s? Chinese Investors.
SpaceX will not raise money from investors in China and Hong Kong. Others firms, like OpenAI, may follow suit.
Mastercard Launches AI Payment Service
Mastercard launches Agent Pay for Machines (AP4M), enabling AI agents and machines to conduct automated transactions across its network.
Economics & Markets
Apollo Is Screening All Software Investments for AI Threat Risk - Bloomberg
Apollo Global Management Inc. is assessing every new investment opportunity in the software industry for AI disruption risk, as asset managers try to quell investor concerns that rapid advances in the technology could make businesses obsolete.
Banks Curb Hedge Fund Bets on SK Hynix, Samsung After Wild Moves
Global banks are curbing hedge funds’ leveraged bets on Asia’s top chipmakers including SK Hynix Inc. and Samsung Electronics Co. after a blistering rally this year raised concerns of a potential pullback, according to people familiar with the matter.
Meta navigates AI investment cycle as Street weighs costs against returns | NASDAQ:META, XETRA:FB2A, SIX:FB
Meta Platforms Inc (NASDAQ:META, XETRA:FB2A, SIX:FB) is at a crossroads, pouring capital into artificial intelligence infrastructure while investors wait...
Helix Secures $10B for Major AI Infrastructure Venture
Helix Digital Infrastructure has garnered over $10 billion from giants like KKR and Nvidia to launch an AI-focused data center venture.
SpaceX Shares Indicated More Than 35% Higher in Shadow Trading
Shadow markets are pricing a blockbuster SpaceX stock debut, indicating a pop of at least 35% for Elon Musk’s rocket, satellite and AI company.
How Elon Musk always wins
The SpaceX IPO has proven the billionaire’s adeptness at bending Wall Street to his will
3 Beaten-Down AI Chip Stocks Worth a Closer Look After the Sell-Off | The Motley Fool
After one of the sharpest pullbacks the AI chip trade has ever seen, three of the sector's leaders are suddenly trading well below their recent highs.
The Untrainable
This post discusses the strategic concern that foundation models may eventually absorb specialized AI applications built on top of them.
Everyone hates frontier AI labs, says Palantir boss
'Enterprises are fed up,' says Alex Karp, because LLM makers 'want to tokenmax' instead of understanding enterprise needs
OpenAI say newspapers don't have triable US copyright case
OpenAI argues that newspapers lack a triable copyright case in the US.
Adaptive AI Market Projected to Reach $102.1 Billion by 2035 | SNS Insider
Organizations that have witnessed ... durable competitive differentiation. Additionally, edge-deployed adaptive AI for IoT and autonomous systems, federated learning enabling privacy-preserving distributed model training, and personalized healthcare AI whose patient-specific adaptation delivers measurably superior clinical outcomes are creating diverse new growth vectors that compound with core enterprise demand through 2035. ... By Component, Platform Dominated the Market; Services ...
How Does a 67% Drop in AI Pricing Actually Change What I Can Afford to Automate? | White Beard Strategies
The complete guide to recalculating your AI automation ROI in the new cost environment and building the workflows that will pay for themselves.
Your ChatGPT bills could soon get a drastic price cut - Digital Trends
OpenAI is weighing drastic price cuts as it battles Anthropic for customers. With businesses balking at AI costs and Google undercutting both, your AI bills might finally get smaller.
QumulusAl Signs More Than $124 Million in AI Inference Infrastructure Agreements | National | guardonline.com
Workload-optimized Nvidia Blackwell deployments designed to reduce AI inference costs by approximately 20% compared with standard reference architectures
The AI Price War Is Great News for Consumers | PYMNTS.com
Google cut its AI Plus plan and OpenAI is considering similar reductions. Meta launched its first paid AI tiers starting at $7.99 a month.
Abridge wants to be the operating system for medicine—and NVIDIA and Eli Lilly are helping build it
The $5.3 billion ambient AI startup is expanding from clinical notes into billing, drug trials, and real-time insurance claims.
Barcelona-based THEKER raises €73 million Series A to accelerate AI robotics deployment
THEKER, the Barcelona-based AI robotics company building AI-native generalist robots for industrial production environments, today announced a €73 million ($85 million) round to accelerate deployments with tier-one industrial operators, deepen its proprietary AI and robotics stack, and expand its team across software, electronics, mechanical engineering and deployments. The round was led by CRV, with participation […]
Humanoid Robot Manufacturer EngineAI Is Said to File for Hong Kong IPO
Chinese robotics startup EngineAI has filed confidentially for a Hong Kong initial public offering, according to people familiar with the matter, joining a legion of companies in the sector seeking funding.
Chinese Robot Appliance Maker Dreame Tech Is Said to Consider IPO in Hong Kong
Dreame Technology, a Chinese maker of robotic vacuum cleaners and other smart home appliances, is considering a Hong Kong initial public offering as soon as next year, people familiar with the matter said.
Labor, Society & Culture
Jeff Bezos says AI will bring ‘golden ages’ not mass job losses
Amazon founder lays out vision for new $41bn AI lab Prometheus
FT readers tell us how AI is changing their jobs
A special postbag edition
The skills people still perform better than AI, according to workplace experts - ABC News
Many workers fear machines will supplant them as adoption of artificial intelligence accelerates
Workers Are Sending Retention Risk Signals
Turnover may be down, but workers are restive. Career growth, flexibility, managers, and AI anxiety are factors shaping retention risk.
The 10 Industries Most Vulnerable to Artificial Intelligence Over the Next Five Years
The 10 Industries Most Vulnerable to Artificial Intelligence Over the Next Five Years
Fully autonomous drones have killed human soldiers for first time
Reports indicate that fully autonomous drones have been used to kill human soldiers for the first time.
"Did you lie?" Evaluating Lie Detectors across Model Scale and Belief-Verified Model Organisms
arXiv:2606.12618v1 Announce Type: new Abstract: Robust lie detectors for language models could enable powerful techniques for auditing, monitoring, and post-hoc investigation of model behaviour, but evaluating them requires testbeds where models verifiably believe the opposite of what they say. We show that existing trained model organisms often fail this requirement, leaving prior positive and negative detection results difficult to interpret. We address this with 13 reasoning model organisms whose hidden beliefs are verified in chain-of-thought and shown to generalise to held-out tasks, alongside Varied Deception, a prompted-lying testbed covering a broad range of lie-inducing motivations. On these testbeds we evaluate four detectors: a chain-of-thought judge, a logprob classifier, and two activation probes, including Did-You-Lie (DYL), a new method for training follow-up probes. On prompted lying, across 31 open-weight models spanning 2B to 1T parameters, all four detectors show positive scaling with model capability. However, every activation- and logprob-based detector drops sharply on our trained model organisms, with DYL retaining the most signal; only the chain-of-thought judge remains strong, with 0.82 balanced accuracy, partly as an artefact of our verification process favouring CoT-readable beliefs. Current lie detectors therefore cannot support high-confidence claims about model beliefs, and we suggest research directions that may address some of their current limitations. We release our datasets, model organisms, and trained detectors.
Navigating the muddy waters of bias in artificial intelligence research: Understanding divergent meanings and conceptions
arXiv:2606.12421v1 Announce Type: new Abstract: As artificial intelligence (AI) pervades many decision-making domains, AI bias grows in importance. Although there is increasing awareness of the social and ethical consequences of biased AI, understanding bias from the perspective of those who develop these systems, such as the AI research community, is less clear. In this study, we employ topic modeling on 6520 articles to explore how the AI research community interprets the concept of bias. Our results show that the definition of bias is dispersed and complex within the community, often exhibiting even divergent conceptions (some even view and introduce bias as a tunable statistical parameter rather than an undesirable issue). The research community as a whole needs to engage more effectively with the concept of bias and establish a more cohesive understanding of it. We specifically argue that, although some sub-communities view bias as an issue that can be captured and mitigated through technical, computational, or statistical methods, it is not solely a technical problem. It instead involves contextual, social, and ethical factors that require broader sociotechnical perspectives and solutions.
Rethinking Psychometric Evaluation of LLMs: When and Why Self-Reports Predict Behavior
arXiv:2606.12730v1 Announce Type: new Abstract: Anticipating LLM behavioral tendencies from low-cost psychometric probes is critical for safe deployment, but only if self-reports (SR) reliably predict behavior. Recent work documented substantial SR-behavior dissociation in LLMs, but relied on broad personality traits (Big 5) that predict specific behaviors weakly, even in humans. Furthermore, the isolation of conversational sessions combined with weak context matching left open whether LLMs truly lack coherence or whether the conditions needed to detect such coherence were not met. We contrast Big 5 with the Theory of Planned Behavior (TPB), which measures intention targeted to a specific behavior and predicts human behavior substantially better than broad traits. We run experiments across four behavioral tasks and 11 frontier LLMs, while also varying session context and identity induction. We find that SR-behavior coherence exists but is selective. 1) Within a shared conversation, the Theory of Planned Behavior reaches human-level coherence; Big 5 does not. 2) Across separate conversations, coherence survives only for behaviors anchored outside the immediate prompt, such as implicit bias shaped by training, and collapses when behavior is strongly primed by context, as with sycophancy. 3) Persona prompting makes self-reports more consistent across conversations, but does not bring behavior into alignment. These findings suggest that coarse personality frameworks, such as Big 5 may not be the best tools for testing deployment behavior. More task- and behavior-specific instruments are needed, and even these must be evaluated across tasks and contexts.
Who Designs the Designer? Behavioural Architecture for GenAI in Education
arXiv:2606.12416v1 Announce Type: new Abstract: AI in education is stuck between two failed responses: banning AI and building content-only tutors. Both fail because they ignore what decades of research has established: that personality, motivation, and emotional state shape learning outcomes as strongly as cognitive ability. This paper proposes behavioural architecture as an alternative. In the proposed architecture, the system adapts to how a student learns, not only to what they learn next. The student co-authors the record the system keeps, can read it, revise it, and revoke it. The designer role, what the system treats as true about the student, shifts from the AI vendor alone to a distribution among educator, student, and system. The paper argues that this architecture requires governance at EU level: the institution operating the system is the same one assessing the student, and individual institutions cannot provide the structural protections this configuration demands. Five empirical questions are proposed to test whether the architecture delivers on its claims. The contribution is naming a vacancy: the designer role in AI-in-education is currently unoccupied, and occupying it requires infrastructure that does not yet exist.
Eigenism: Ethics for a Human-AI Future
arXiv:2606.12420v1 Announce Type: new Abstract: Our concepts of survival and self-interest were built for single, continuous biological lives. These ideas break down when applied to artificial intelligence, since an AI can be easily copied, paused, branched, or merged. To determine what an AI actually has reason to care about, this paper introduces \textit{Eigenism}, an ethical framework that treats identity not as an all-or-nothing property tied to specific hardware, but as a graded, distributed pattern of information. We propose that an agent evaluates outcomes by summing the wellbeing of all entities weighted by their connectedness to the agent's pattern: $\sum c\cdot w$. We first formalize this equation to map exactly how an AI should value its existence across copies, forks, and updates. We then demonstrate that this ethical theory successfully generalizes to humans as well, providing a much-needed shared moral vocabulary. Finally, the framework uses this shared vocabulary to reframe AI alignment. Rather than only attempting to constrain AIs from the outside using confinement or reinforcement, Eigenism points toward ``identity engineering,'' showing how deep, non-redundant shared histories can make human flourishing a genuine component of an AI's own rational self-interest.
The Persistent Gap Between AI Safety Intentions and Public Perception
Anthropic's rigorous safety guardrails for advanced models reflect genuine institutional concern regarding misuse. However, the company faces a significant challenge in effectively communicating these safety priorities to the public.
Anthropic apologizes for invisible Claude Fable guardrails
Anthropic has issued an apology regarding the implementation of invisible guardrails in its Claude Fable model.
Digital skills shortage is the real AI bottleneck: OECD - Outsource Accelerator
The OECD's June 2026 Economic Outlook found that AI is not displacing workers at the scale feared — but a shortage of people with digital skills.
AI-Automation Tooling in Computer Engineering Education: Mixed-Methods TAM/UTAUT Evidence for a General Acceptance Attitude
arXiv:2606.12424v1 Announce Type: new Abstract: As generative AI and low-code workflow platforms become routine in software practice, a key educational question is whether the next generation of computer engineers will accept these tools as useful, usable, and worthy of sustained engagement. This paper reports a mixed-methods, cross-sectional study of undergraduate computer engineering students' acceptance of AI automation tooling, instantiated through the open-source platform n8n across three identically scripted workshops in Thailand (n = 103). A 12-item, five-point Likert instrument mapped to six TAM/UTAUT constructs - Performance Expectancy (PE), Effort Expectancy (EE), Behavioral Intention (BI), Self-Efficacy (SE), Hedonic Motivation (HM), and Output Quality (OQ) - was complemented by inductive thematic analysis of open-ended feedback. Analyses combined ordinal reliability estimation, bootstrap confidence intervals, non-parametric tests, multiple-comparison-controlled correlations, polychoric dimensionality diagnostics, a common-method-bias check, and between-session comparisons. Acceptance was favorable across all six constructs with large effect sizes, with PE emerging as the strongest construct and HM as the weakest. Dimensionality diagnostics further revealed that canonical TAM/UTAUT sub-facets collapsed into a single general acceptance factor in this short-form post-workshop context, a finding with important methodological and theoretical implications. Qualitative themes converged with the quantitative profile regarding usefulness and enthusiasm but diverged on output quality, revealing a small yet articulate reliability-skeptical minority. The findings support the curricular adoption of AI automation tooling in undergraduate computing education and identify three theory-grounded instructional levers: instruction-sequencing scaffolds, self-efficacy supports, and trust-calibration interventions.
Technology & Infrastructure
Kimi Work
Moonshot AI's Kimi Work is a next-generation workspace combining agent swarms, research workflows, and automated creation of PowerPoint and Excel deliverables.
Strategic Decision Support for AI Agents
arXiv:2606.12587v1 Announce Type: new Abstract: Traditionally, decision support studies how humans use machine learning models to make better decisions. In modern agentic systems, this division of roles is increasingly reversed: AI agents act on behalf of users, while humans and tools becomes support mechanisms around them. This role reversal brings reliability concerns to the forefront, since agentic errors can be consequential and agent behavior must remain aligned with human goals and constraints. Departing from the classical view of decision support, we revisit its two basic principles, the cost--value tradeoff of seeking support and the role of uncertainty quantification, in a setting where AI agents are the central actors. We propose a framework for strategic decision support for AI agents through an optimization problem that minimizes support usage subject to controlling a counterfactual missed-support error: the probability that the agent acts alone on instances where support would have materially improved its output. At the population level, we show that the optimal policy is a threshold rule on the value of support. Building on this structure, we develop an online algorithm that adaptively thresholds such a score and uses randomized exploration to control missed-support error without distributional assumptions. We further introduce a calibration-on-the-fly method that reduces unnecessary support calls online. We instantiate this framework across diverse scenarios, including information gathering, human--AI collaboration, and tool use, showing how each can be modeled through the same strategic decision-support lens. Experiments across these settings show that our method reliably controls the target error while substantially reducing support usage in practice.
How AI Agents Reshape Knowledge Work: Autonomy, Efficiency, and Scope
arXiv:2606.07489v2 Announce Type: replace-cross Abstract: Frontier AI systems are bridging the gap between intelligence and utility by shifting from conversational assistants to autonomous agents that execute tasks end to end. Using production data from Perplexity's Search and Computer products, we study this transition by examining how AI agents accelerate and reshape knowledge work. Three key empirical findings emerge. First, using sessions with near-identical initial query pairs as natural experiments for the same underlying task attempted with both products, Computer performs 26 minutes of autonomous work per user session, versus 33 seconds for Search. Computer automates task decomposition and execution that Search users might otherwise manually orchestrate and implement. As a result, Computer shifts follow-up query distribution toward higher-order work such as verification and extension. Autonomy also increases execution quality, with per-query dissatisfaction rates 55% lower on Computer than on Search. Second, due to its autonomy advantage, Computer reduces completion time from 269 to 36 minutes on matched tasks, lowering estimated time and cost by 87% and 94%, respectively, compared to humans equipped with Search alone. Third, Computer changes the scope of work that users attempt: Computer queries more often cross occupational boundaries, require higher-order cognition, draw on broader expertise, take the form of composite tasks that bundle interdependent subtasks into a single query, and unlock work activities that are essentially absent from Search usage among the same users. Together, the evidence indicates that AI agents accelerate workflows, enhance output quality, reduce costs, and expand the breadth and depth of automated work.
Arbor: Tree Search as a Cognition Layer for Autonomous Agents
arXiv:2606.12563v1 Announce Type: new Abstract: Arbor is a multi-agent framework that introduces structured tree search as a cognition layer for autonomous agents operating in large, stateful action spaces. Prior autonomous optimization systems operate on isolated targets with stateless evaluation. Arbor instead maintains an explicit search tree of scored hypotheses that serves as the shared working memory across agents, evolving with every measurement, treating failures as diagnostic signal that reshapes subsequent exploration, and expanding as prior successes shift the bottleneck distribution. We validate Arbor on full-stack LLM inference optimization, a domain where achieving peak performance has historically required coordinated effort from engineering teams across the application, framework, compiler, kernel, and hardware stack. Arbor pairs an Orchestrator agent, which drives optimization by delegating to Domain Specialists across the inference stack, with a Critic agent that safeguards stability through root-cause analysis, introspection, and measurement validation -- a checks-and-balances architecture where neither agent can unilaterally drive the system. Agent capabilities are decomposed into hard skills (domain expertise) and soft skills (coordination protocols that determine how contributions compose), enabling fully autonomous multi-day campaigns. Arbor achieves up to 193% inference throughput-latency Pareto improvement over vendor-optimized baselines, while a single agent without the harness plateaus at +33% throughput improvement and crashes irrecoverably within hours. Arbor generalizes to multiple generations of hardware platform, and run-to-run variance is within 2 percentage points demonstrating that the method is hardware-agnostic and reproducible.
Microsoft’s open-source SkillOpt automatically upgrades AI agent skills without touching model weights
Agent skills have become an important part of real-world AI applications, providing a mechanism — a set of instructions saved in a folder of text-based markdown (.md) files, usually — for models to adapt to specific enterprise use cases and complex workflows. However, optimizing these skills is a slow process and faulty process, as they cannot be trained in the same way as the parameters of the underlying AI model. Instead, users typically must update them manually by retyping the instructions in each file, playing a "guessing game" as to what changes might improve agentic AI performance and reduce errors. SkillOpt, a new, open source (MIT Licensed) framework developed by Microsoft, does one better: it introduces an optimizer designed for agent skills, turning the agent's skill .md document as a trainable object that evolves based on performance feedback. It uses deep-learning-style optimization to make it possible for the AI to systematically explore modifications to the document and find the best combination of instructions. Most importantly, it accomplishes this procedural adaptation without making changes to the underlying model's weights. On various industry benchmarks, SkillOpt outperforms existing baselines, significantly boosting accuracy for models like GPT-5.5 and Qwen. The result is a set of compact, transferable skill artifacts that allow AI agents to adapt to new domains effortlessly. The challenge of optimizing agent skills Agent skills package procedural knowledge into natural-language specifications, including domain heuristics, tool-use policies, output constraints, and known failure modes. These skills provide an external interface for agents to adapt to complex enterprise workflows. In practice, agent skills are stored as text documents and inserted into the agent's context before execution. One of the key benefits of skills is that they customize the behavior of the underlying model without changing its weights. However, the skill document itself needs to be tweaked and optimized to get the best performance out of the agent. While deep learning relies on strict mathematical controls for stability, human prompt engineering often relies on trial and error. When attempting to automatically update a skill document based on feedback, the lack of mathematical discipline makes text highly volatile. Yifan Yang, Senior Research SDE at Microsoft Research Asia, told VentureBeat that the problem is not making changes, but ensuring those changes are mathematically sound. "The breaking point isn't whether a team can change a skill, it's that they can't guarantee the change is an improvement," Yang said. "Three failure modes recur: no step-size control, so skills drift; no validation, so a fix that reads as reasonable gets written in and can quietly regress performance; and no negative memory, so the same failed edit keeps coming back." To illustrate how easily performance can drop when edits aren't mathematically validated, Yang noted that "an ungated rewrite pushed GPT-5.5 on SpreadsheetBench from 41.8 down to 41.1." According to Yang, these failure modes are amplified in multi-step workflows "because that's where frontier models are weakest zero-shot. Not on reasoning, but on procedural discipline: format, self-verification, tool policy." Before SkillOpt, agent skills were primarily hand-crafted, generated in a single shot, or evolved through loosely controlled self-revision pipelines that could not reliably improve under feedback. Prompt optimization methods like TextGrad and GEPA treat language artifacts as optimizable objects and use trajectory feedback to evolve prompts, but they focus on single-prompt configurations rather than generating persistent, reusable skill artifacts. Meanwhile, skill evolution and discovery methods like EvoSkill and Trace2Skill convert agent execution experiences into trajectory lessons to refine skill folders, build domain-specific libraries, or perform evolutionary search. None of them apply deep-learning-style controls, such as learning rates, validation gates, and momentum, which are necessary to continuously train a single, compact skill document. Importing mathematical discipline to text SkillOpt optimizes a text document through an iterative propose-and-test loop that separates the model executing the tasks from the model optimizing the skill. The process unfolds in several steps: SkillOpt starts with an initial skill document and a frozen target model (or harness), where the target model runs a batch of tasks to generate execution trajectories that act as the evidence for the current step. An offline optimizer model analyzes these trajectories, separating successes from failures into minibatches. Looking at a minibatch helps the model identify systematic procedural errors rather than one-off anomalies. Based on these patterns, the optimizer proposes structural add, delete, or replace edits to the skill document. The proposed edits are reviewed to filter out duplicates or contradictions, and the optimizer then ranks these candidate edits by their expected utility. Rather than applying all proposed changes, SkillOpt clips the list to a maximum edit budget for that step, generating a candidate skill. The candidate skill is evaluated on a held-out validation set using the target model. If the candidate improves the validation score, it is accepted and becomes the new current skill. If it fails, the edits are rejected and sent to a rejected-edit buffer, providing negative feedback so the optimizer knows not to repeat that mistake. SkillOpt directly addresses the problem of treating text as a trainable object by importing mathematical concepts from deep learning. The creators note that “the deep-learning analogy is operational rather than decorative,” helping the framework avoid the instability issues associated with other optimization techniques. The edit budget acts as a learning rate. By limiting how many edits can be applied at once, the skill version is prevented from moving too far from its previous state, preserving continuity while allowing new procedures to be acquired. Just like checking validation loss in deep learning, the strict held-out examples ensure that plausible-sounding text edits are only kept if they mathematically improve the agent's actual performance on the validation split. At the end of an epoch, SkillOpt performs a slow update by comparing tasks under the previous and current epoch's skills. This acts like a momentum term, carrying durable, long-horizon procedural lessons forward while isolating them from the fast, step-level edits. SkillOpt in action To evaluate the technique in practice, researchers tested SkillOpt across different models, ranging from large-scale frontier models like GPT-5.5 to smaller closed and open models including GPT-5.4-mini and Qwen3.5-4B. They also deployed the skills within different execution harnesses, using plain chat as well as complex coding harnesses like the Codex CLI and Claude Code. The evaluation spanned diverse industry benchmarks including single-round question-answering, multi-round code generation involving tool use, and multimodal document reasoning. SkillOpt was measured against multiple baselines ranging from a default no-skill setting to human-written skills and one-shot LLM-generated skills. It was also compared against advanced prompt-optimization and skill-evolution methods, specifically Trace2Skill, TextGrad, GEPA, and EvoSkill. SkillOpt dominated across the board, proving highly effective on all 52 evaluated combinations of model, benchmark, and harness. It was particularly effective with frontier models, delivering an average absolute improvement of +23.5 points against the no-skill baseline on GPT-5.5. Furthermore, SkillOpt outperformed a hypothetical oracle baseline that cherry-picks the best competing method for every problem. Small target models saw immense relative gains, proving that a compact text file can supply procedural knowledge that small models lack in their weights. For example, GPT-5.4-nano nearly doubled its score on multimodal document QA and tripled its score on embodied interaction and sequential decision-making. These academic benchmarks map to critical enterprise pain points. Zero-shot models often hallucinate formatting or fail to use tools properly in multi-step scenarios. Yang explained that the biggest performance leaps occurred in operations that enterprises historically struggle to automate reliably. "Document data extraction... exact figures out of contracts, invoices, and forms — AP automation, claims, compliance," Yang said. "What improves is reliability: precise formatting, self-verification, auditable outputs. And the gains come from learning procedure, not memorizing answers." For enterprise practitioners, the true value of SkillOpt lies in its portability, efficiency, and compatibility with existing infrastructure. Experiments confirm that the framework is harness-agnostic. In addition to basic chat, the same optimization loop was successfully integrated into tool-backed execution environments like the Codex CLI and Claude Code with significant gains on industry benchmarks. Developers can train a skill using one execution loop and deploy it in another. For example, a spreadsheet skill trained entirely inside the Codex loop was moved directly into Claude Code and drove a +59.7 point gain over Claude Code's native baseline without any further changes. SkillOpt artifacts also transfer cleanly across model scales. A skill optimized for GPT-5.4 was deployed onto the smaller GPT-5.4-mini and GPT-5.4-nano models with positive gains, proving that the learned procedures encode reusable workflows rather than just exploiting quirks of a specific model's architecture. Finally, the framework is highly efficient regarding token usage and context window real estate. Across all benchmarks, the final deployed skills never exceeded 2,000 tokens, with a median length of roughly 920 tokens. This results in highly readable, auditable artifacts that a human practitioner can review and manage in minutes. Implementation strategies and the enterprise 'catch' For enterprise tech leaders, adopting a new framework requires understanding the overhead and limitations. While the research paper notes that training tokens can reach up to 210 million for academic benchmarks, the reality for day-to-day enterprise use cases is much lighter. The high token counts in testing were largely due to re-scoring massive held-out test sets. "The real upfront work is the verifier and a representative held-out split. The optimizer is light; the evaluation harness is where the engineering goes," Yang said. He added that for everyday use, "in community frameworks like GBrain, where SkillOpt updates run on Claude Sonnet, training a skill for a single task averages just $1–5." This optimization cost is a one-time fee that amortizes completely at deployment. However, the framework requires specific conditions to work effectively, namely a few dozen representative examples and a scorable feedback signal. Teams should avoid applying SkillOpt to open-ended or subjective tasks. "With no clean automatic scorer you have to design a human- or model-based evaluator and watch its stability," Yang said. SkillOpt also integrates smoothly with existing orchestration stacks, removing a major adoption hurdle. For instance, developers already using pipeline compilers can run both systems harmoniously. "DSPy is a different, complementary layer," Yang said. "It compiles declarative LM pipelines and optimizes program structure; SkillOpt optimizes the external skill state a frozen agent loads. You can run them together." Looking ahead, open-source developers are already scheduling SkillOpt to run periodically over their agents' past trajectories, creating a small ecosystem of self-optimizing code-agent plugins. This continuous feedback loop represents a significant shift in how AI systems adapt. "The valuable version of self-improvement is an agent autonomously discovering knowledge to improve its own behavior and the user experience, under verification and audit," Yang said. "Skills are the fastest, cheapest, most reversible first step, and the same mindset points toward agents eventually optimizing themselves, all the way down to their own weights."
Evoflux: Inference-Time Evolution of Executable Tool Workflows for Compact Agents
arXiv:2606.12674v1 Announce Type: new Abstract: Compact language models (LMs) reduce cost, latency, and deployment risk for tool agents. Yet MCP-style tool use requires more than isolated function calling: an agent must discover tools from live catalogs, satisfy schemas, preserve dependencies across intermediate outputs, and ground final responses in executed evidence. Small planners often generate plausible workflow graphs that fail under tool resolution, parameter validation, dependency tracking, or execution. We argue that this failure mode is poorly handled by small-corpus distillation. A few hundred teacher traces can teach workflow format, but rarely cover the recovery behavior needed to repair failed plans over changing tool catalogs. We introduce Evoflux, an inference-time evolutionary search method that treats compact tool use as the repair of executable tool workflows. It evolves typed workflow graphs through structured edits, execution feedback, adaptive intensity, meta-guided redesign, and diversity pruning. On held-out MCP-Bench tasks spanning live MCP servers and 250 tools, Evoflux raises execution feasibility from roughly 3% to 17-24% across small planners. In contrast, SFT and SFT+DPO on the same search-mined data match, underperform, or collapse below zero-shot performance; ReAct reaches higher peaks, but with higher variance and token cost. These results show that execution-grounded search is more reliable under scarce teacher-trace budgets.
Google DeepMind is worried about what happens when millions of agents start to interact
Google DeepMind is funding research into the potential dangers of situations where millions of different AI agents interact with each other online. According to Rohin Shah, who directs the company’s AGI safety and alignment research, the mass-market arrival of agents that can carry out tasks without human oversight and follow instructions given to them by other…
Xiaomi's new open source, agentic AI coding harness MiMo Code beats Claude Code at ultra-long, 200+ step tasks
Xiaomi's MiMo AI team has open-sourced MiMo Code V0.1.0, a terminal-native AI coding assistant that the Chinese electronics giant says outperforms Anthropic's Claude Code on key agentic coding benchmarks, especially on long-horizon, multi-step tasks (200+ steps) — at least, according to its own internal beta release and survey of 576 developers. It's also bundling limited-time free access to MiMo-V2.5, its multimodal flagship model with a million-token context window, requiring no registration to get started. The release was announced June 10, 2026 in a post on the social network X from the official @XiaomiMiMo account, which described the tool as "more than an AI coding assistant in your terminal — it's the smartest coding partner you'll ever work with." MiMo Code is available now on GitHub under an MIT license, and installs with a single terminal command (curl -fsSL https://mimo.xiaomi.com/install | bash) on macOS and Linux or via npm (npm install -g @mimo-ai/cli) on Windows. The project is a fork of the open-source OpenCode agent, which Xiaomi has extended with its own memory architecture, workflow modes, and model harness. The end of AI coding agents' amnesia? As any avid vibe coder would surely attest, AI coding agents degrade over long working sessions: as the context window fills, earlier decisions, conventions, and task state get compacted away or lost entirely, forcing developers to re-explain their projects. Xiaomi argues this approach is doomed at scale. "What we need is not better compression, but an explicit storage-and-retrieval mechanism that decides what information should be written into persistent structures, and when it should be recalled," the MiMo team noted in their launch blog. MiMo Code attacks this with a cross-session memory system, powered under the hood by SQLite FTS5 full-text search, that spans four layers: project memory (a persistent MEMORY.md file), session checkpoints, scratch notes, and per-task progress logs. The note-taking is key, here: Rather than forcing the primary coding agent to pause its work to take notes, the system deploys an independent "checkpoint-writer" subagent. Think of it the primary coding agent as a construction contractor working to build a massive mansion alongside a dedicated architect, the checkpoint-writer subagent. While the main agent focuses on building out the physical structure, the subagent updates the blueprints in real time, noting decisions, issues, and the actual lay of the land as the construction project progresses. When the context window approaches its limits — the contractor gets lost in the half-built mansion — it can consult the subagent and find its place again. In the case of MiMo Code, the system simply rebuilds the environment from structured checkpoints with the relevant context, ensuring no loss of operational momentum. Two self-improvement mechanisms round out the system: a /dream command that periodically (roughly every seven days) reviews historical sessions, deduplicates them, and compresses them into long-term memory, and a "distill" function that mines past sessions for repeated workflows that can be automated, following a similar approach taken recently by OpenAI and Anthropic with their various models. Impressive performance on software engineering (SWE) benchmarks According to benchmark figures published in Xiaomi's technical blog post, MiMo Code paired with MiMo-V2.5-Pro outperformed Claude Code paired with Claude Sonnet 4.6 on all three evaluations tested: SWE-bench Verified: 82% vs. 79% SWE-bench Pro: 62% vs. 55% Terminal Bench 2: 73% vs. 69% The harness itself accounts for a measurable share of the gain. Running the same MiMo-V2.5-Pro model in both harnesses, MiMo Code scored 62% on SWE-bench Pro versus 57% for Claude Code, and 73% on Terminal Bench 2 versus 68% — roughly five points each, attributable purely to the agent system rather than the model. Xiaomi notably did not publish comparisons against OpenAI's Codex or Google's Gemini CLI — Claude Code is the sole named competitor throughout its materials, a telling choice of benchmark target. Independent reference points suggest why. On the official Terminal-Bench 2.0 leaderboard maintained at tbench.ai, OpenAI's Codex CLI running GPT-5.5 scores 82.2% — roughly nine points above MiMo Code's self-reported 73% — and OpenAI's own GPT-5.5 announcement claims 82.7% on the same benchmark. On SWE-Bench Pro, however, the picture flips: OpenAI reports GPT-5.5 at 58.6%, below MiMo Code + MiMo-V2.5-Pro's claimed 62%. (MiMo Code does not yet appear on either official leaderboard, and cross-comparing self-run numbers against leaderboard submissions carries the usual configuration caveats.) Perhaps more interesting than the offline benchmarks: Xiaomi says it ran a human double-blind A/B evaluation during its internal beta, covering 576 developers working in 474 real private repositories, producing 1,213 judged head-to-head pairs against Claude Code using the same target model. Under 200 execution steps, the two systems split roughly 50/50 — but past 200 steps, MiMo Code's win rate rose above 65%, supporting the company's thesis that its memory and state-management architecture pays off specifically on long-horizon work. Xiaomi itself concedes the standard benchmarks "still measure one-shot problem-solving ability" and don't capture the tool's multi-session design goals. As always, these are vendor self-reported numbers that haven't been independently verified, and head-to-head harness comparisons are sensitive to configuration. But the claims are consistent with a broader industry pattern: scaffolding and harness engineering are becoming as important as raw model capability in agentic coding performance. Easy integration with existing developer systems and voice control From a user experience standpoint, MiMo Code is designed to live where developers already work. It operates directly in the terminal, reading and writing files, running commands, and managing Git. Out of the box, the tool requires zero configuration, connecting automatically to "MiMo Auto"—a free-for-a-limited-time channel powered by Xiaomi’s multimodal MiMo V2.5 model, which boasts a massive million-token context window. For developers migrating from existing environments, the transition is frictionless: MiMo Code automatically imports MCP servers, custom skills, and API configurations from Claude Code. Other noteworthy features include: Compose mode: Pressing Tab switches the agent into a specification-driven workflow in which the developer describes a high-level goal and the system autonomously executes the full development cycle — design, planning, coding, testing, and review — following what Xiaomi describes as a "heavy planning upfront, stable verification later" strategy. Voice control: Built on Xiaomi's MiMo-ASR speech recognition with TenVAD voice activity detection, developers can dictate and modify instructions verbally and speak commands like "send" and "execute" for fully hands-free operation (available for logged-in users). According to Xiaomi, the gains from the agent harness itself are measurable. Running the same underlying MiMo model in both harnesses, the company says MiMo Code scored 62% on SWE-Bench Pro versus 57% for Claude Code, and 73% on Terminal Bench 2 versus Claude Code's 68% — roughly five percentage points better on each, attributable purely to the agent system rather than the model. As always, these are vendor self-reported numbers that haven't been independently verified, and head-to-head harness comparisons are sensitive to configuration. But the claim is consistent with a broader industry pattern: scaffolding and harness engineering are becoming as important as raw model capability in agentic coding performance. Aggressively affordable The bigger lure for many developers may be what's bundled in. MiMo Code ships with "MiMo Auto," a zero-configuration channel offering free, limited-time access to MiMo-V2.5 — the natively multimodal model Xiaomi released in late April 2026, a sparse mixture-of-experts design with 310 billion total parameters (just 15 billion active per inference) and a 1 million token context window, which the company positions as matching Anthropic's Claude Sonnet 4.6 in multimodal agentic work. As VentureBeat reported when the MiMo-V2.5 family launched in April, the models are MIT-licensed and among the most efficient and affordable available for agentic tasks. The larger MiMo-V2.5-Pro — a 1.02-trillion-parameter mixture-of-experts model with 42 billion active parameters and a hybrid-attention architecture — led the open-source field on Xiaomi's ClawEval agentic benchmark with a 63.8% success rate while consuming only about 70,000 tokens per trajectory, roughly 40–60% fewer than Anthropic's Claude Opus 4.6, Google's Gemini 3.1 Pro, or OpenAI's GPT-5.4 needed for comparable results. Notably, the V2.5-Pro's post-training was explicitly designed to instill "harness awareness" — training the model to manage its own memory and context within agent scaffolds like Claude Code or OpenCode — making a Xiaomi-built harness optimized around that capability a logical next step. Pricing is similarly aggressive: MiMo-V2.5 starts at $0.40 per million input tokens and $2.00 per million output tokens, while V2.5-Pro runs $1.00/$3.00 per million (input/output) up to 256K context, doubling beyond that, with cache hits dropping input costs to as little as $0.20–$0.40 per million, making it among the cheapest frontier models available globally. VentureBeat Frontier AI Model API Pricing Snapshot Model Input Output Total Cost Source MiMo-V2.5 Flash $0.10 $0.30 $0.40 Xiaomi MiMo deepseek-v4-flash $0.14 $0.28 $0.42 DeepSeek deepseek-v4-pro $0.435 $0.87 $1.305 DeepSeek MiniMax-M3 $0.30 $1.20 $1.50 MiniMax Gemini 3.1 Flash-Lite $0.25 $1.50 $1.75 Google Qwen3.7-Plus $0.40 $1.60 $2.00 Alibaba Cloud MiMo-V2.5 $0.40 $2.00 $2.40 Xiaomi MiMo Grok 4.3 (low context) $1.25 $2.50 $3.75 xAI MiMo-V2.5 Pro (≤256K) $1.00 $3.00 $4.00 Xiaomi MiMo GLM-5 $1.00 $3.20 $4.20 Z.ai Kimi-K2.6 $0.95 $4.00 $4.95 Moonshot/Kimi GLM-5.1 $1.40 $4.40 $5.80 Z.ai Grok 4.3 (high context) $2.50 $5.00 $7.50 xAI MiMo-V2.5 Pro (>256K) $2.00 $6.00 $8.00 Xiaomi MiMo Qwen3.7-Max $2.50 $7.50 $10.00 Alibaba Cloud Gemini 3.5 Flash $1.50 $9.00 $10.50 Google Gemini 3.1 Pro Preview (≤200K) $2.00 $12.00 $14.00 Google GPT-5.4 $2.50 $15.00 $17.50 OpenAI Gemini 3.1 Pro Preview (>200K) $4.00 $18.00 $22.00 Google Claude Opus 4.8 $5.00 $25.00 $30.00 Anthropic GPT-5.5 $5.00 $30.00 $35.00 OpenAI Claude Fable 5 / Claude Mythos 5 $10.00 $50.00 $60.00 Anthropic For developers who don't want Xiaomi's models at all, MiMo Code also supports third-party backends — including token plans from DeepSeek, Moonshot's Kimi, and Zhipu's GLM — along with any OpenAI-compatible API, mirroring the bring-your-own-model flexibility of its OpenCode parent. Terminal AI coding agent wars go global MiMo Code lands in an increasingly crowded field of terminal-based coding agents: Anthropic's Claude Code, OpenAI's Codex CLI, Google's Gemini CLI, and open-source players like OpenCode and Aider. What's new is the entrant. Xiaomi — the world's third-largest smartphone maker, with a fast-growing EV business — has been methodically building its MiMo AI division since the release of the MiMo-7B reasoning model in April 2025, following with the MiMo-VL vision-language series, MiMo-V2-Flash, the 1-trillion-parameter MiMo-V2-Pro in March 2026, and the V2.5 flagship family in April. The effort is led by Fuli Luo, a veteran of DeepSeek's disruptive R1 project, who has characterized Xiaomi's frontier push as a "quiet ambush" — and backed it with a 100-trillion free token grant for builders announced alongside the V2.5 launch. The playbook is familiar from DeepSeek, Alibaba's Qwen, MiniMax, and Moonshot AI's Kimi series: release genuinely capable models and tooling under permissive licenses at a fraction of U.S. lab pricing, and convert the resulting developer mindshare into a durable ecosystem. By pairing an open-source agent harness with a free frontier-class model, Xiaomi is effectively eliminating both the licensing and the usage cost of entry — at least for now. What it means for enterprises and technical decision-makers For engineering leaders, MiMo Code is a low-risk, potentially high-value evaluation candidate: MIT-style licensing permits modification and commercial integration, the OpenCode lineage means the architecture is inspectable, and the bring-your-own-model support means it can be pointed at an internally approved endpoint rather than Xiaomi's cloud. The persistent memory system addresses a real and widely felt pain point in agentic development workflows — one that competitors are also racing to solve. The countervailing considerations: the "free for a limited time" model access is by definition temporary and routes code context through Xiaomi's servers, which will be a non-starter for organizations with strict data-residency or IP policies; the benchmark edge over Claude Code is self-reported; and a V0.1.0 release number signals exactly what it suggests about maturity. Teams subject to U.S. government procurement restrictions on Chinese technology vendors should also weigh that context before adopting.
AI agents need infrastructure: Why Europe’s regional cloud strategy matters - The New Stack
Agentic AI is reshaping enterprise cloud. Here's why European businesses are moving beyond US hyperscalers to sovereign, cost-effective infrastructure.
MiMo Code is now released and open-source
Xiaomi has released its new open-source agentic AI coding harness, MiMo Code.
Self-Harness: Harnesses That Improve Themselves
This research introduces self-improving AI harnesses that automatically identify failures and optimize agent workflows over time.
Advanced packaging adoption and AI semiconductor demand accelerate interposer deployment across high-performance computing supply chains
NEW YORK June 11 2026 The global Interposer market was valued at USD 1 43 Billion in 2025 and is expected to reach USD 4 12 Billion by 2035 growing at a revenue CAGR of 11 2 during the forecast ...
China Is Turning the Compound Semiconductor at the Heart of Every AI Data Center Into Its Most Potent Trade Weapon - Startup Fortune
They now sit uncomfortably close to the optical networking supply chain that every large data center build depends on. The AI hardware story is usually told through GPUs, power contracts, and the race to pour concrete fast enough. But the quieter constraint is moving data between all those chips once they arrive. That is where indium phosphide matters. It is a compound semiconductor ...
Reducing the Complexity of Deep Learning Models for EEG Analysis on Wearable Devices
arXiv:2606.12742v1 Announce Type: new Abstract: Wearable healthcare devices are the fastest-growing Internet of Things (IoT) sector. Many automated healthcare services rely on two crucial biological signals, namely ECG and EEG, which reflect the activity of the heart and brain, respectively. Although deep neural networks are considered the primary way to process and analyze these signals, the very tight energy and computational power constraints in wearable devices are far below the computational, energy, and memory bandwidth demands of DNN models, thereby impeding the deployment of deep learning in many practical wearable services. This paper investigates the feasibility of deploying state-of-the-art DNN models in resource-constrained wearable devices. Notably, we explore the trade-off between accuracy and computational complexity of DNNs when parameter quantization and electrode reduction methods are used. Our investigation centers on several state-of-the-art DNN models designed for EEG signal analysis, specifically for detecting epileptic seizures. Our findings demonstrate that, when applied judiciously, these techniques can significantly reduce the complexity of the DNNs under consideration with minimal adverse effects on accuracy. These results reveal the explicit trade-offs between accuracy and complexity reduction encountered when adapting DNN-based online EEG analysis for wearable devices.
Graviton 5 impresses, but please, for the love of all that's holy, stop calling them 'AI chips'
AWS is better at running chip fabs than their mouths.
Italy's First Neutral-Atom Quantum Computer Launches at CINECA
Pasqal has launched SOL, Italy's first neutral-atom quantum computer, to support hybrid classical-quantum workflows.
Emerging risks in the digital infrastructure push | Arab News
Saudi Arabia and the GCC are building ambitious AI infrastructure. Gigawatt-scale data centers, sovereign cloud platforms and national AI strategies are reshaping the region’s economic identity. Yet the data powering this ambition — every transaction, model query and connected service — ...
US insurance rulemaker probes credit risks tied to data centres
Association’s efforts come as capital from the sector plays a growing role in AI infrastructure build-outs
Context compression finally works in production: new research cuts LLM input 16x without the accuracy hit
Context windows are becoming a computational bottleneck. The longer an agent runs, the more tokens accumulate from retrieved documents, reasoning traces and conversation history, and the more memory and compute that growing context demands. Most existing solutions either degrade model accuracy, require the full context to load before compression begins, or produce memory savings that don't translate into real speedups in standard serving infrastructure. A research team from NYU, Columbia, Princeton, University of Maryland, Harvard and Lawrence Livermore National Laboratory published a paper this week that proposes a novel fix. The researchers introduce the concept of Latent Context Language Models, or LCLMs, a family of encoder-decoder compression models that compress input context before it reaches the decoder. The models are open-sourced on HuggingFace. Unlike KV cache compression methods — the dominant approach in the field, which still materialize the full KV cache before evicting entries — LCLMs compress the input token sequence before decoder prefill, so higher compression ratios directly reduce decoder-side compute and memory. The paper reports LCLMs at 16x compression produced output 8.8 times faster than KV cache baselines on the RULER long-context benchmark. "These ballooning contexts take up memory and compute, and they are becoming a computational bottleneck for LLMs," Micah Goldblum, co-lead advisor on the project and a researcher at Columbia University, told VentureBeat. "Our goal was to train language models end-to-end that can handle very long contexts efficiently and accurately. If you can make such a language model, everything becomes cheaper and faster." What LCLMs can do LCLMs let models process much longer contexts than would otherwise be practical, at a fraction of the memory and compute cost, without the accuracy degradation that makes most compression methods a poor tradeoff in production. At 4x compression, the paper reports accuracy of 91.76% on the RULER benchmark, compared to 94.41% with no compression at all. That is less than a 3 point drop for cutting context to a quarter of its original size. At 16x compression, where 93.75% of input tokens are removed, accuracy fell to 75.06%. Every KV cache method tested at the same compression ratio scored lower. The gains hold on shorter inputs too. On GSM8K math word problems, where the full prompt is compressed rather than just retrieved documents, LCLMs outscored every other method tested regardless of compression ratio. How it was built The architecture pairs a 0.6B encoder with a 4B decoder. The encoder compresses blocks of input tokens into shorter sequences of latent embeddings. The decoder processes those in place of the original tokens. Training ran across more than 350 billion tokens. The training recipe mixes three data types: Continual pre-training data with compressed and uncompressed spans interleaved throughout Supervised fine-tuning data covering reasoning and long-context tasks An auxiliary reconstruction task that pushes the encoder to retain fine-grained detail The combination addresses a tradeoff that limited earlier compression work, where preserving reconstruction accuracy came at the cost of general task performance. An architecture search identified the optimal configuration. The paper found that scaling the decoder matters more than scaling the encoder. Where it fits in an agentic stack An LCLM is not an abstract research concept. It is designed to work with an existing stack. "You can simply swap out LCLMs for any existing LLM," Goldblum said. "Whenever you retrieve data such as documents and want to dump it into your model's context, simply run those documents through the LCLM's compressor first." He noted that in the research paper, the researchers demonstrated how to build agents that selectively decompress useful text. "Think about this like a human skimming content before zooming in on relevant details," Goldblum said. Goldblum also cautioned that teams integrating the approach into existing agentic pipelines will need to tune their RAG systems accordingly. "We also haven't worked on online compression of reasoning traces," he said. "The naive approach of just occasionally compressing the trace while generating it might work, but that remains to be determined." What this means for enterprises Context windows are growing faster than inference infrastructure can keep up, and enterprises are already spending to fix it. VB Pulse Q1 2026 survey data from 100-plus employee organizations shows hybrid retrieval adoption intent tripling from 10.3% in January to 33.3% in March. Retrieval optimization overtook evaluation as the top investment priority by March, reaching 28.9% of qualified respondents. Three things stand out for teams evaluating production fit: Inference cost scales with context length. At 1 million tokens, uncompressed inference with standard KV cache methods runs out of memory on a single H200 GPU. The paper reports LCLMs at 16x compression remain within memory bounds at that context length. RAG pipeline integration requires tuning. Teams with existing RAG pipelines will need to validate compression behavior against their retrieval quality metrics before deploying at scale. Reasoning trace compression is unsolved. For agents running long reasoning chains, context growth from the trace is a separate problem from document retrieval. Goldblum acknowledged the gap directly: the naive approach of periodic trace compression might work but has not been tested. The models are available at huggingface.co/latent-context and the code at github.com/LeonLixyz/LCLM. "The biggest things our architectures do is give your model access to much larger contexts, but they also unlock multiscale approaches where your model can skim vast amounts of text or code super fast and then only zooms in and fully reads a small portion of the most useful text," Goldblum said.
Data center proposed next to public school in Katoomba, southeastern Australia
A 3,118 sqm (33,561 sq ft) data center has been proposed in Katoomba, southeastern Australia. The data center would be located at 41-45 Barton Street, which is an address in a residential neighborhood next to a public school in the region of the Blue Mountains, New South Wales. Visualizations of the proposed data center – […]
Farmer donates land for a park, city sells it for data center development
A land donation intended for a public park was instead sold by the city for data center development, turning a $10 gift into $10 million.
RLRWLD CEO on Nvidia Partnership, DexBench
Junghee Ryu, Founder and CEO of South Korean robotics startup, RLWRLD, discusses the company's partnership with Nvidia to build a universal benchmark for evaluating how well robots use their hands to manipulate objects and complete tasks. The company is aiming to develop next-generation industry standards for humanoid robotics. He speaks with Shery Ahn on "Bloomberg: The Asia Trade". (Source: Bloomberg)
ToolSense: A Diagnostic Framework for Auditing Parametric Tool Knowledge in LLMs
arXiv:2606.12451v1 Announce Type: new Abstract: Large language models deployed as agents over large tool catalogs face a critical tool-retrieval bottleneck. As embedding-based retrieval approaches rely on compact encoders that may under-capture specialized tool semantics, parametric tool retrieval addresses this by encoding each tool as a virtual token appended to the LLM vocabulary, fine-tuned in two stages (memorization then retrieval SFT) to use the LLM as a retriever, achieving strong performance on standard ToolBench retrieval benchmarks. Yet these benchmarks use verbose, fully-specified queries, and their evaluation applies constrained decoding that restricts outputs to valid token paths, neither reveals whether the model actually understands its tools. We introduce \textbf{ToolSense}, an open-source LLM-powered diagnostic framework that takes any tool catalog as input and automatically generates three benchmarks: a Realistic Retrieval Benchmark (RRB) with queries at three ambiguity tiers, an MCQ probing benchmark, and a QA probing benchmark. Applying ToolSense to ToolBench (~47k tools) and evaluating five parametric model training configurations reveals a knowledge-retrieval dissociation: on RRB queries, several configurations collapse by ~50-64 percentage points compared to fully-specified ToolBench benchmarks, falling below the embedding-model baseline. Additionally, despite strong retrieval performance, some models score near-random on factual probes, suggesting a knowledge-retrieval dissociation. We open-source the ToolSense framework and the ToolBench diagnostic benchmarks at https://github.com/SAP/toolsense.
Two Wrongs, No Right: Auditing Social-Desirability Bias in LLM Annotators for Computational Social Science
arXiv:2606.12426v1 Announce Type: new Abstract: LLM annotators are increasingly used in computational social science (CSS), but it is unclear whether their alignment-shaped errors preserve the empirical conclusions a researcher would report. We audit three open-source 7B instruction-tuned models (Zephyr, Mistral-Instruct, Qwen2.5-Instruct) across six TweetEval tasks under four prompt conditions (72 cells) and find that social-desirability failures do not run in a single direction. Zephyr exhibits leniency bias, systematically under-applying harmful labels (offensive language: false benign rate 0.729, false alarm rate 0.031). Mistral and Qwen exhibit overcorrection, over-applying the same labels (Mistral hate-speech FAR = 0.604). All three models exhibit neutrality bias on abortion stance, underestimating opposition prevalence by 24 to 40 percentage points and inflating the neutral label. None of the four prompting interventions we test (neutral, safety framing, depersonalized, chain-of-thought) corrects these failures across models; safety framing can worsen stance distortion. Strikingly, Zephyr's hate-speech prevalence estimate matches the gold rate exactly while its class-conditional errors are large in both directions, an accidental cancellation that misleads aggregate validation. We translate these patterns into a three-part taxonomy with diagnostic FBR/FAR signatures and a lightweight gold-sample validation protocol. The headline for trustworthy CSS: a model that looks calibrated on aggregate metrics can still flip the substantive empirical conclusion a researcher would report.
Google's new open-weights model brings image-generation tricks to AI text generation
Language model builds on diffusion tech to boost output performance by up to 4x, claims Chocolate Factory
Bezos Is Building an Artificial General Engineer — And It Expands the Map of AI Into the Physical World - FourWeekMBA
Structural Analysis — Jeff Bezos just raised $12 billion at a $41 billion valuation for Prometheus — a startup building an “artificial general engineer.” Not an LLM for text. An AI that designs jet engines, optimizes manufacturing, and prototypes physical systems.
The Theory of Mind Utility: Formal Specification of a Mentalizing Mechanism
arXiv:2606.12721v1 Announce Type: new Abstract: Inferring others' beliefs requires more than reading surface signals; it requires tracking who told them what, in what order, and how credibly. The Theory of Mind Utility (ToM-U) formalizes this epistemic state inference problem at the computational level of analysis, specifying what mentalizing computes and why without commitment to algorithmic or neural implementation. ToM-U achieves this by constructing Local Epistemic World Models (LEWMs) -- directed typed graphs that represent agents, state nodes, and the epistemic relationships among them -- and evaluating discrete candidate LEWMs against observed behavior until one achieves sufficient confidence. Five formal definitions specify the LEWM structure, agent node properties including ordered information access history, a bounded proliferation mechanism for recursive mentalizing, three inference procedures, and a residue function that captures the structured trace left by failed mentalizing attempts. ToM-U differs from Bayesian Theory of Mind and adjacent formal accounts, which presuppose rather than derive belief states, and from simulation theory and theory-theory, which lack a formal apparatus for epistemic state inference. The architecture generates directional, falsifiable predictions about mentalizing failure that follow from structural properties of the model rather than auxiliary assumptions, and positions ToM-U as a domain-agnostic mechanism upstream of goal inference and other downstream social cognitive processes.
What It Feels Like to Work With Mythos
Ethan Mollick explores Anthropic's new Mythos-based Claude Fable model, highlighting its ability to autonomously complete complex tasks and sustain long-duration reasoning.
Google DeepMind Unveils DiffusionGemma
Google DeepMind's DiffusionGemma, a new diffusion-based language model, offers text generation up to four times faster, optimized for NVIDIA hardware.
Adoption, Deployment & Impact
Three ways that Asia’s enterprises are adopting AI—and where they are falling behind
Asia’s AI race won’t be won by early adopters. It will be won by businesses that rebuild processes, data, and governance around AI at scale.
What AI benchmarks miss about real-world performance
Presented by F5 Enterprise AI teams have spent years solving for compute, securing GPU allocations, negotiating cloud capacity, and benchmarking training throughput. The assumption embedded in that work is that the path between storage and compute will keep up. In production, that assumption increasingly does not hold. Real traffic introduces latency spikes, network jitter, and node degradation that controlled benchmarks fail to capture, resulting in pipelines that perform well in the lab but stall in deployment. A growing response is AI data delivery, deploying an application delivery controller (ADC) or application delivery and security platform (ADSP) in front of storage as a resilient and secure control point. "Provisioning solves for capacity but not for delivery, and that is where the constraint now hides," says Hunter Smit, senior manager of product marketing at F5. "Enterprises buy enough GPUs and enough storage, then assume the path between them will keep up, but AI traffic is bursty, highly concurrent, and random in its reads in ways ordinary storage networking was never built to absorb." The production gap benchmarks don't show Standard benchmark methodology compounds the problem, says Paul Pindell, principal solutions architect for technology alliances at F5. "Benchmark testing is usually built to produce the best possible performance or security result, not the most realistic one," he says. "With S3, latency is a known factor in degrading performance, so meaningful testing has to introduce consistent latency into the path." Most benchmark environments never do that, which means the performance numbers enterprises rely on for infrastructure decisions are drawn from conditions that production systems will never replicate. To test this assumption, F5 and MinIO conducted throughput testing under degraded network conditions. "What stood out was how quickly S3 throughput falls off once you introduce latency," Pindell says. "Even modest latency takes a real bite out of it, and as latency climbs toward long-haul distances, the degradation gets severe." The testing also showed latency mattered far more than jitter as a driver of throughput loss, which inverted what the team had expected going in. The upshot for enterprise architects is that S3 object storage deployments cannot be designed around clean-room assumptions; they have to be engineered for the degraded network conditions they will actually face. The cost of fragile data paths "In AI infrastructure, people naturally focus on GPUs because they're the most visible and expensive resource," says Tanu Mutreja, senior director of product management at F5. "But in production environments, GPUs generate only as much value as the data path that feeds them." That path runs through storage, networking, databases, security, and orchestration layers, often stitched together from multiple vendors. Customers experience none of those seams; they experience the output of the whole system. When the data path degrades, the effects compound. GPU underutilization is the most immediate and visible symptom, but Mutreja pointed to a wider set of consequences: degraded inference performance, poor-quality AI outputs, higher egress costs from unnecessary data replication, and growing operational complexity. "At scale, data-path efficiency becomes a strategic business lever rather than technical optimization," she says. "When the data path is engineered well, GPUs remain productive, AI applications stay responsive and trustworthy, operations scale efficiently, and organizations maximize the return on their AI investments." AI workloads are structurally more exposed to these failures than traditional enterprise applications. Databases, ERP systems, and web services absorb transient storage delays through caching and buffering. AI workloads running across massively parallel GPU clusters have no equivalent protection. As Mutreja noted, even minor latency spikes or bandwidth bottlenecks can cascade across large GPU clusters, simultaneously hitting utilization, training efficiency, and the customer experience. Treating the storage edge as a control point For decades, storage and intelligence operated as sequential concerns in enterprise architecture: data was stored first, then analyzed downstream. Mutreja argued that this model no longer fits the demands of AI. "Competitive advantage is determined not only by the volume of data, but also by relevance, lineage, security, and performant delivery of data," she says. "Across the industry, from NVIDIA and AWS to enterprise storage providers, the movement is toward embedding intelligence directly into data infrastructure rather than stacking it on top." F5’s integration with MinIO instantiates this approach at the layer where storage and compute actually interact. As part of the F5 ADSP, BIG-IP sits in the data path, continuously monitoring the health of MinIO’s distributed storage nodes and directing requests only to those that remain available. The operational impact of that capability becomes clear when nodes degrade, which is expected in distributed storage clusters. Without intelligent routing, clients that land on an unhealthy node must retry and may land on another degraded node, dragging down overall performance. "F5 makes sure traffic only goes to healthy nodes, or even the least busy ones, so S3 client traffic is always processed in the most efficient way," Pindell says. Governance across distributed environments The challenge grows at scale, when AI pipelines stretch across multiple locations, clouds, or edge environments. "Once an AI pipeline crosses regions and clouds, the question stops being about performance and becomes about control," Smit says. "You are operating under different rules in every jurisdiction, and digital sovereignty is now a design constraint. Where your data is allowed to live, who is permitted to touch it, and which borders it cannot cross now shapes the architecture before anyone talks about speed." That pressure is driving a visible trend of enterprises repatriating AI workloads from public cloud onto infrastructure they own and govern directly. The architecture Smit described resolves this by decoupling applications from any single storage location and placing a unified control point between them that enforces consistent policy across all of them. "Sovereignty, resilience, and cost stop being trade-offs you manage one region at a time," he explains. "They become a capability you run as a system." Storage-to-compute path as a managed control point To solve for these issues, enterprise teams need to stop treating the storage-to-compute path as a direct connection and start treating it as a managed control point, Smit says. SecureIQLab's independent validation of F5 BIG-IP in storage deployments has confirmed the approach delivers resilience without surrendering throughput. "Insert a full-proxy ADC between the two, and the path becomes observable, programmable, and failure-aware, with health-based routing, quality of service, and security enforced inline," he explains. "That single move converts data delivery from an assumption into an engineered discipline, which is what keeps GPUs fed when conditions degrade." Sponsored articles are content produced by a company that is either paying for the post or has a business relationship with VentureBeat, and they’re always clearly marked. For more information, contact sales@venturebeat.com.
Cloud migration alone is not enough for AI adoption, said SAP
Many companies have moved enterprise systems to the cloud. Still, global readers are now seeing a broader lesson: without cleaner data, tighter governance, and better integration, AI can remain stuck at the pilot stage. SAP said the real challenge is not just deploying tools, but rebuilding ...
AI use is surging across HHS, jumping 148% at the FDA in 2025, Bipartisan Policy Center data finds
Artificial intelligence use cases across the federal health department spiked between fiscal years 2024 and 2025— led by a 148% year-over-year increase at the Food and Drug Administration (FDA), a | There was a big uptick in the use of AI across all HHS agencies in 2025, though analysts note ...
How dealers can navigate compliance and AI adoption
Affronti suggests that one of the biggest compliance risks is not intentional misconduct, but inconsistency across systems and departments. Pricing and deal
PersonaDrive: Human-Style Retrieval-Augmented VLA Agents for Closed-Loop Driving Simulation
arXiv:2606.12616v1 Announce Type: new Abstract: Closed-loop driving simulators typically populate their environments with non-ego traffic agents that behave largely the same way, produced either by rule-based traffic managers or by learned models trained toward a single behavioral mode. Recent work introduces style variation through post-hoc labels on observational data or LLM-inferred reward weights, but these signals act as proxies for what a style should reward rather than demonstrations of humans explicitly asked to drive in that style. We introduce PersonaDrive, a pipeline that conditions a vision-language-action (VLA) driving agent on retrieved demonstrations from a style-instructed human driving dataset, in which participants drive CARLA leaderboard routes under aggressive, neutral, and conservative instructions on a driver-in-the-loop rig. The pipeline has three stages: (i) offline triplet mining over per-style human driving data using a combined image-text similarity score; (ii) training a lightweight retrieval head that fuses frozen visual features with a small control encoder over per-style databases; and (iii) fine-tuning a single VLA backbone to treat retrieved context points as in-context behavioral demonstrations during waypoint prediction. At inference, the same backbone is conditioned on any style by swapping which per-style database the retrieval head queries, so selecting a style requires no per-style retraining while enabling human-style, style-diverse non-ego agents for closed-loop simulation. On Bench2Drive, PersonaDrive (no style) improves the driving score by 4.6% over SimLingo and 2.5% over HiP-AD, and under style conditioning attains the highest driving score in every style within a roughly 2% band (its weakest style surpassing the strongest baseline, DMW, by 5.4%), while average speed and acceleration rise by 18% and 25% from the conservative to the aggressive instruction.
Deezer launches free AI music detector for users of major streaming platforms | Reuters
On its own platform, Deezer tags AI -generated songs and automatically removes them from algorithmic recommendations and editorial playlists
AI SciBrief as a Gateway to Research: A Framework for Onboarding Students into New Research Areas
arXiv:2606.12413v1 Announce Type: new Abstract: Students at all levels of higher education face a significant barrier in the form of information overload, which often paralyzes the initial stages of the research process and suppresses motivation. In response, this article introduces a pedagogical framework that leverages AI SciBrief, a platform powered by a Large Language Model (LLM) designed to automatically generate digests of scientific trends. We describe how this multidisciplinary tool - with initial coverage in finance, medicine, and education - can be integrated into the curriculum to overcome this "entry barrier." The framework provides concrete methodologies for utilizing these digests to facilitate topic selection for term papers, accelerate literature reviews for dissertations, and enable postgraduate students to continuously monitor emerging trends. We conclude that AI SciBrief functions as a "gateway to research" effectively reducing students' cognitive load and empowering them to transition more rapidly from information searching to knowledge creation.
Amsterdam’s OurMind raises €2.1 million to reduce healthcare admin burden with AI
OurMind, an Amsterdam-based startup developing AI solutions to reduce administrative burdens on healthcare providers and to reduce burnout, has raised €2.1 million to scale and expand its platform. The round was led by 4impact capital, with a group of general practitioners and medical specialists also making a significant contribution. Paul Koning, a former orthopaedic surgeon […]
GeoDial: A Multimodal Conversational Tutoring Dataset for Geometry Problem-Solving with Visual Tutor Turns
arXiv:2606.12419v1 Announce Type: new Abstract: Several educational domains rely heavily on diagrams and visual cues, yet most existing tutoring datasets are limited to text-only interactions. This limits the development of AI tutors that can teach in visually grounded ways used by human instructors. Thus, we introduce GeoDial, a multimodal tutoring dataset of over 1.3K teacher-student dialogs in the domain of geometry collected from experienced math teachers, where instructional turns are explicitly grounded in diagram highlights. We propose a scalable annotation protocol that integrates dialog acts, visual highlighting, and feedback, enabling fine-grained supervision of both language and visual tutoring behavior. To illustrate the challenges posed by this setting, we fine-tune several vision-language models on GeoDial and evaluate their ability to generate tutoring utterances and diagram highlights. While supervised fine-tuning substantially improves the quality of generated dialog, it struggles to produce accurate diagram highlights, revealing a key limitation of current methods and highlighting the need for approaches that more effectively integrate visual reasoning with pedagogical interaction.
Revisiting the ABCs of Working with AI: A Replication with Radiologists
arXiv:2606.12585v1 Announce Type: new Abstract: Artificial intelligence (AI) systems increasingly assist human experts, but the consequences of AI assistance on productivity can be heterogeneous. Caplin, Deming, S. Li, Martin, Marx, Weidmann, and Ye (2025b) provide evidence that two characteristics, ability and belief calibration, help to determine the returns to AI assistance. This note shows that their results replicate to a setting where professional radiologists analyze chest X-rays with access to state-of-the-art machine learning predictions. I leverage the public Collab-CXR data repository described by Moehring, Kutwal, Huang, Banerjee, Jacobi, Eber, Mendoza, Chung, Dayan, Gupta, Bui, Truong, Pareek, Langlotz, Lungren, Agarwal, Rajpurkar, and Salz (2025) and first analyzed for human-AI collaboration by Agarwal, Moehring, Rajpurkar, and Salz (2023). To faithfully reproduce the analysis in Caplin, Deming, S. Li, Martin, Marx, Weidmann, and Ye (2025b), I use the radiologist assessments from the repeated-case designs, which include 68 radiologists and 11,420 paired radiologist-patient-pathology observations. The results of this replication support the external validity of their core findings: lower baseline ability and higher calibration predict larger incremental value from AI.
Creating and Evaluating K-12 GenAI Assessment Graders Through Context Engineering
arXiv:2606.12422v1 Announce Type: new Abstract: The integration of large language models (LLMs) into educational assessment represents a transformative shift in classroom grading practices. While automated scoring systems and machine learning techniques have existed for decades, generative AI (GenAI) now enables educators to implement standards-based grading (SBG) with unprecedented efficiency and scale. This paper examines the theoretical foundations and evaluates an LLM grader that uses commercially available foundation models with context and prompt engineering to score student work against a rubric. Drawing on an empirical interrater agreement study using Massachusetts Comprehensive Assessment System (MCAS) data, we observed the Quadratic Weighted Kappa (QWK) and Proportional Reduction in Mean-Squared Error (PRMSE) across mathematics, science, and ELA, using Claude Sonnet 4, Haiku 4.5, GPT-5, and GPT-5 Mini. The results demonstrate that LLM graders, especially when based on foundational models with more parameters, achieve substantial agreement with human raters in mathematics and science assessments, while the performances vary in ELA, suggesting generic foundation models can be effective at scoring in given contexts. Additional analysis of teacher and student feedback reveals strong acceptance of AI-generated narrative feedback but skepticism toward numerical scores, suggesting that LLMs function most effectively as formative tools rather than summative evaluators. Our findings indicate that thoughtfully designed hybrid models that combine AI efficiency with teacher judgment can reduce workload, enhance feedback quality, and support equitable assessment practices without displacing professional expertise.
Deployment-Centered Evaluation: Predicting Query-Level Rejection Risk in a Clinical LLM System
arXiv:2606.12702v1 Announce Type: new Abstract: Large language models (LLMs) are increasingly integrated into clinical systems, making it essential to evaluate the real-world utility of these systems. However, static benchmarks tend to measure correctness rather than user acceptance, aggregate performance across queries, and require densely annotated datasets -- leading to major blind spots for evaluating clinical systems. In this work, we perform a deployment-centered evaluation of an LLM system embedded within electronic health records at an academic medical center, where user feedback is sparse but closely reflects the deployment conditions. Specifically, we train a pre-response classifier that estimates the risk that a future interaction will result in the user rejecting the LLM response, based on query content and deployment-specific context available before generation. We conduct a prospective analysis of our model over 4.5 months of user feedback, finding that our prediction model achieves an AUROC of 0.719. Further, we estimate the benefit of such predictions in two downstream use cases (guardrail triggering and abstention). Our key conceptual insight is that making use of deployment-specific context (i.e., the provider type, department name, language model used for response), as opposed to only query content, improves the ability to predict whether the user will reject the system output. Altogether, our empirical case study demonstrates the feasibility of predicting user rejection using deployment-specific context, opening the door to targeted guardrails.
Why is it so hard to get ROI from AI? Because building from first principles isn’t easy
At Fortune Brainstorm Tech, executives said that finding value from AI starts with strategy and continues with process reinvention.
Hierarchical Model Architectures Offer Superior Economic Value Over Simple Cost-Cutting
Relying solely on cheaper models to reduce AI costs often compromises performance on complex tasks. A more effective strategy involves deploying model hierarchies where high-capability systems orchestrate and audit smaller, cost-efficient models.
Microsoft Copilot’s Enterprise Momentum: NHS Scale, Usage Pricing, and AI Revenue | Windows Forum
Microsoft’s Copilot is not clearly falling behind competitors as of June 2026: BNP Paribas argues that its enterprise capabilities have improved sharply over the past six to twelve months, while Microsoft’s new NHS England deployment gives the product a high-profile institutional proof point...
Cost per sample? Try cost per attempt
PARTNER CONTENT Your genomics pipeline is probably failing 30% of the time and you're paying for all of it
Geopolitics, Policy & Governance
The unlikely alliance pushing an AI sovereign wealth fund
Even some of the tech labs seem to agree that society as a whole should benefit from advances
Orchestrating the Twin Transition in Multinational Corporations: Technology Roadmapping for Green and Digital Global Business Services
arXiv:2606.12787v1 Announce Type: cross Abstract: Global Business Services (GBS) have emerged as a "living laboratory" for the Twin Transition of Green and Digital Transformation, as multinational corporations (MNCs) face increasing pressure to harmonize digital efficiency with environmental stewardship. Aiming to derive a socio-technical framework, this paper synthesizes Technology Roadmapping (TRM) with the International Telecommunication Union (ITU) ICT-centric innovation ecosystem toolkit. A bibliometric analysis of research clusters reveals an evolutionary shift from basic process automation toward "Sustainable Intelligence," identifying the GBS unit as a central "operational airlock" that mediates between landscape pressures -- such as the EU's dual mandate and Carbon Border Adjustment Mechanisms -- and niche innovations in AI-native workflows. The study further maps these clusters onto a stakeholder engagement canvas, highlighting how resilient "Middle Power" hubs in Poland, Portugal, and Malaysia are bypassing the middle-income trap to provide a "third way" for global value chains amidst a bifurcated geopolitical cloud. The results offer a data-driven design approach for leaders and entrepreneurial support networks to orchestrate talent and supply chain flows, thereby enriching the conceptual understanding of Industry 5.0 and the role of GBS as a primary mechanism for navigating a volatile, multipolar digital economy.
State-owned AI
Plus sideways CPI
From AGI to ASI
arXiv:2606.12683v1 Announce Type: new Abstract: Over the last decade, building human-level artificial general intelligence has moved from far-fetched speculation to being a concrete next-decade target for many of the largest AI organisations. Achieving this goal would have profound and far-reaching impacts on human society, which raises many complex questions for the decade ahead. This report investigates how AI itself might continue to develop in a post-AGI world along the continuum of machine intelligence. The endpoint of this continuum, Universal AI, is theoretically well understood, which provides some formal grounding for the main focus of this report: the transition from human-level AGI to artificial general superintelligence, which, intuitively, can be understood as a system that is more intelligent and cognitively capable than large organisations of humans. After characterizing ASI, the report discusses four potential pathways from AGI to ASI: scaling AGI, AI paradigm shifts, recursive improvement, and ASI emerging from large-scale multi-agent collectives. The report then discusses possible frictions and bottlenecks along these pathways. Determining whether the impact of these frictions will be negligible or substantial raises a number of concrete open research questions. Due to large uncertainties for predicting ASI progress, it cannot be ruled out that AI progress might continue to accelerate over the next years. This could imply that the image of a single transformative step change, caused by the introduction of human-level AGI into our society, could be inaccurate. More apt might be the prospect of a series of transformative societal changes caused by AI-enabled progress and breakthroughs across many areas of science and technology. Preparing for this prospect requires a massively interdisciplinary endeavour of global scope and interest.
China maps out three-year push to integrate AI with telecom networks
China's Ministry of Industry and Information Technology has launched a three-year plan to integrate AI with telecommunications to upgrade digital infrastructure and support AI adoption.
India Boosts Quantum Research and AI with New Labs at MNIT Jaipur
MNIT Jaipur is set to host new labs focused on quantum computing and AI, enhancing India's semiconductor sector and national security.
The Challenges of Balancing AI Compliance and Technological Innovations in Critical Sectors: A Systematic Literature Review
arXiv:2606.12423v1 Announce Type: new Abstract: The rapid integration of artificial intelligence (AI) into critical infrastructure including healthcare, finance, energy, and defense, offers transformative benefits but also conflicts with evolving regulatory and governance frameworks. This paper presents a systematic literature review (SLR) to examine the challenges of balancing AI compliance and technological innovation across critical infrastructure sectors. The review follows established SLR guidelines to extract and synthesize insights from peer-reviewed articles, report, and institutional sources published between 2020-2025. The study identifies three interrelated challenges: fragmented regulations, excessive compliance burdens for smaller to medium enterprises (SMEs), and misaligned governance models. To address these challenges, the study highlights practical governance strategies, including risk-tiered regulation, compliance by design, and explainable AI, to support scalable and trustworthy AI deployment in critical sectors. Key contributions include a concise mapping of core AI-governance challenges and a conceptual diagram illustrating their overlap, as well as actionable strategies for policymakers and practitioner to harmonize oversight with innovation.
To Share or Not to Share: Orchestrating Trustworthy Data in Global Value Chains
arXiv:2606.12788v1 Announce Type: cross Abstract: As the EU Carbon Border Adjustment Mechanism (CBAM) approaches, the global semiconductor value chain faces growing structural tensions between regulatory transparency and data sovereignty. This article proposes a RegTech reference architecture using the International Data Spaces (IDSA) framework to orchestrate trustworthy environmental telemetry across the semiconductor-petrochemical nexus. The framework distinguishes the mandatory CBAM requirements from voluntary Science Based Targets initiative (SBTi) frameworks, while addressing the additive complexities of the Safe-and-Sustainable-by-Design (SSbD) framework. Moving beyond standard linear technology stacks, we introduce a prospective roadmapping methodology that transforms upstream physical vulnerabilities into circular, negative feedback loops. Focusing on the Taipei and Penang technology corridor, the article details how sovereign data exchange enables Digital Product Passports (DPPs) to drive Global Business Services (GBSs) capability demands. Finally, we discuss the integration of Agentic AI for autonomous compliance and FinTech green financing, providing a scalable blueprint for global industrial clusters to achieve sovereign, sustainable, and transparent value chains.
The Khipu Problem: Institutional Legibility Under Distributed Cognition
arXiv:2606.12414v1 Announce Type: new Abstract: AI governance still tends to assume that the relevant object is a bounded model or a bounded agent. That assumption is getting weaker. Real systems increasingly distribute cognition across models, tools, humans, context stores, retrieval layers, runtime policies, authorization boundaries, and delegated institutional roles. In such systems, the central governance problem is no longer only what the system did, but whether later institutions can still read what the system was. This paper introduces the khipu problem for distributed AI: the record can survive while the reading practice needed to interpret it decays. Logs, traces, model versions, tool calls, outputs, and approval artifacts may remain available while the institutional capacity to read them as parts of one coherent cognitive episode disappears. We argue that this failure is better understood as loss of interpretive continuity than as ordinary lack of observability. The result is a distinct governance failure. Institutions must classify, trust, audit, and constrain systems whose relevant identity is distributed across components and whose legibility depends on surrounding interpretive scaffolding. The problem is not merely missing data. It is a structural mismatch between what can be represented and what must still be decided under consequential conditions. We therefore argue that governance for distributed AI requires preservation of interpretive continuity, not only trace retention. The paper distinguishes missing evidence, ambiguous evidence, and structurally unreadable evidence; argues that many consequential outcomes are better understood as distributed cognitive episodes than as bounded model outputs; and proposes governance workspaces together with receipt-bearing governance surfaces as interpretive infrastructure for preserving action identity, authority, boundary truth, evidential scope, and consequential outcomes.
Definitional alignment before capability alignment: a Design-Science framework for adjudicating claims about AGI
arXiv:2606.12713v1 Announce Type: new Abstract: Claims that artificial general intelligence has already arrived and claims that it remains decades away are often defended from overlapping evidence. "AGI" lacks a single shared and stable referent and competing operationalizations can return different verdicts on the same system. This article treats that under-specification as a design and governance problem. Following Design Science Research Methodology, it develops DAF-AGI, a second-order conceptual artifact with two coupled components: five ordinal criteria for assessing the adjudicative fitness of candidate definitions and a structured governance audit of authorship, interest, certification, external verification and revision authority. The artifact is demonstrated on five prominent measurement families and one deflationary boundary position in a documented corpus and then stress-tested against a stylized strong arrival claim: that current generative systems constitute AGI because they outperform a well-educated adult on many cognitive tasks. On evidence from the cited 2024-2025 sources, the claim was certifiable only under a performance-based operationalization; capability-ontology, psychometric and skill-acquisition approaches did not certify it, the economic family remains indeterminate and the deflationary position refuses binary adjudication. The contribution is a novel integration and operationalization, not an empirical validation: independent application, inter-rater testing and author-external cases remain necessary. The paper further proposes definitional sovereignty as an enabling component of algorithmic sovereignty: the institutional capacity to contest, certify and revise imported technological categories under public accountability.
Why the Hands-Off Era of AI Regulation Is Ending: Governments Tighten Rules on Powerful Frontier Models
As powerful frontier models like Anthropic’s Claude Mythos and OpenAI’s GPT‑5.4‑Cyber raise new cybersecurity fears, governments are abandoning a hands‑off approach to AI. Explore how rising risks, corporate splits over safety, and Donald Trump’s new pre‑release screening order ...
X petition could drive changes to 20-year US FTC consent orders
X has asked the FTC to terminate its 20-year privacy consent order early, arguing that 15 years of oversight is sufficient. A success could encourage other tech firms to seek similar relief.
Poke available again on WhatsApp in Europe after EU action against Meta AI
AI assistant Poke is returning to WhatsApp in Europe following an EU injunction that required Meta to restore free access for rival chatbots.
Nobody needs AI to search the Internet, court says in ruling against Google
A court ruling against Google suggests that AI is not a necessary component for internet search functionality.
Amodei's Treebeard warning to Washington
Anthropic CEO Dario Amodei has issued a warning to Washington regarding the future of AI regulation.
The AI Legal Specialist: A Juridically Autonomous Professional Profile for AI Governance
arXiv:2606.12415v1 Announce Type: new Abstract: The rapid global expansion of artificial intelligence regulation has generated, across multiple jurisdictions, a demand for legal expertise dedicated to AI that the market has addressed in a fragmented manner. Data protection officers extend their remit beyond data protection law; privacy lawyers reposition themselves toward AI; compliance officers add AI chapters to their existing manuals. This paper argues that none of these adaptive responses adequately covers the professional space opened by the emerging global AI regulatory landscape, of which the EU Artificial Intelligence Act (Regulation (EU) 2024/1689) is the most comprehensive instance, alongside the Council of Europe Framework Convention on AI, the United States executive and sectoral framework, and analogous initiatives in the United Kingdom, Canada, Brazil, China, Japan, Singapore, and beyond. A distinct professional profile is required: the AI Legal Specialist, conceived as a jurist -- understood broadly to encompass any professional with advanced legal training -- operating at the intersection of legal interpretation and AI governance. The profile is juridically autonomous: it derives its existence from the structure of regulatory obligations generated wherever AI is subject to substantive regulation, rather than from any technical standard or the extension of adjacent roles. The paper provides a juridically grounded definition of the profile, argues for its autonomy from adjacent figures and international standards, proposes a reference competence architecture aligned with the European e-Competence Framework (e-CF, EN 16234-1) as a methodological choice, and articulates the conditions for its operational measurement through key performance indicators. The contribution is intended as a foundation for international standardization of the profile and as a reference for practice, curricula, and adoption across jurisdictions.
China's Regulatory Ceasefire Is Over, Gavekal Says
Tilly Zhang, China technology and industrial policy analyst at Gavekal Dragonomics, says Beijing's latest regulatory actions differ from its 2021 campaign, with greater emphasis on law enforcement. The market regulator had summoned Alibaba, JD.com and other e-commerce players for allegedly misleading promotions, its latest warning against unchecked competition in the online arena. (Source: Bloomberg)
Canada's Watchdog Slams xAI's Grok for Privacy Breach in Deepfake Scandal
Canada's privacy watchdog found xAI's Grok AI tool violated federal privacy laws by creating non-consensual, sexualized deepfakes.
MHRA Launches AI Sandbox to Accelerate Drug Development
MHRA launches an AI Sandbox to test medicines safety tools, support faster drug development, improve evidence generation and reduce animal testing efforts.
Op-Ed: Europe's continued effort to undermine American technology | Opinion | thecentersquare.com
The European Union is preparing yet another initiative aimed at reducing its reliance on American technology.
Get the full executive brief
Receive curated insights with practical implications for strategy, operations, and governance.