Fri 17 April 2026
Daily Brief — Curated and contextualised by Best Practice AI
Investors Channel Billions into AI, OpenAI Splurges on Chips, and Banks Face Model Risks
TL;DR Ericsson warns of rising semiconductor costs driven by AI demand, squeezing network equipment margins. Investors direct billions into AI via ETFs and elite funds like Iconiq, while defense stocks rise on global tensions. OpenAI commits over $20 billion to Cerebras chips amid data center delays affecting 40% of US builds, including Microsoft and OpenAI projects; UK banks prepare to deploy Anthropic's Claude Mythos despite warnings of cyber vulnerabilities.
The stories that matter most
Selected and contextualised by the Best Practice AI team
The 3 forces quietly dismantling the business model that made enterprise software fabulously profitable
Roundtables with senior business leaders in San Francisco and New York revealed a consensus: AI isn't just threatening SaaS, but the entire industry.
Ericsson Targets Networks Growth Despite Caution Over Rising Costs
Chief Executive Borje Ekholm said it was facing increasing input costs, especially in semiconductors, caused in part by AI demand.
AI Dominates ETF Flows
Investors are getting picky, and ETFs reveal exactly where the money is going. Seana Smith, Senior Investment Strategist at Global X, joins Bloomberg Open Interest to break down why AI still dominates, how defense stocks are surging on global tensions, and why only a handful of companies will truly win the AI race. (Source: Bloomberg)
Iconiq, Go-To Wealth Adviser for Tech’s Elite, Is Putting Billions Into AI
Last year, Anthropic PBC chief Dario Amodei and a handful of executives traveled 8,000 miles from San Francisco to the Middle East. They were there to meet with some of the world’s most deep-pocketed investors, including Qatar’s sovereign wealth fund and Abu Dhabi-based MGX.
The Tyranny of AI Everywhere
Sneakers? Why stop there?
Economics & Markets
AI Dominates ETF Flows
Investors are getting picky, and ETFs reveal exactly where the money is going. Seana Smith, Senior Investment Strategist at Global X, joins Bloomberg Open Interest to break down why AI still dominates, how defense stocks are surging on global tensions, and why only a handful of companies will truly win the AI race. (Source: Bloomberg)
Iconiq, Go-To Wealth Adviser for Tech’s Elite, Is Putting Billions Into AI
Last year, Anthropic PBC chief Dario Amodei and a handful of executives traveled 8,000 miles from San Francisco to the Middle East. They were there to meet with some of the world’s most deep-pocketed investors, including Qatar’s sovereign wealth fund and Abu Dhabi-based MGX.
US companies that jumped ship to tech, AI businesses over the years | Reuters
Sneaker company Allbirds' pivot to AI computing has revived a Wall Street trend in recent years: smaller companies leaning into tech-focused transformations to win over AI-hungry investors.
These 3 Charts Show How Venture Capital Has Concentrated At The Top In 2026
In the first quarter of 2026, a handful of large, well-funded AI companies, almost all based in the U.S., captured the vast majority of venture dollars, even as global startup deal count fell, Crunchbase data shows.
How the post-boom correction reshaped venture capital firms, now disrupted by AI — TFN
Artificial intelligence has breathed a new lease of life into an otherwise sluggish venture capital market, and investors are back to cutting cheques. UK firms have shed headcount, reflecting a post-boom correction, as the industry more broadly turns to AI to aid investing.
AI in Drug Discovery Market: Share, Growth and Outlook Driven by AI-Enabled Pharmaceutical Innovation
The AI in Drug Discovery Market size was valued at USD 2 34 Billion in 2025 and is projected to reach USD 17 43 Billion by 2033 growing at a CAGR of 28 5 during the forecast period 2026 2033 ...
Lam Research (LRCX) and the AI Semiconductor Boom: Analyst Insights – ICO Optics
This blog post digs into Lam Research’s recent share moves and what’s driving them. I’m coming at this from the […]
Helical Builds Reproducible AI Platform for Drug Discovery
Helical raises $10M to advance reproducible AI workflows, accelerating drug discovery with scalable, decision-ready platforms.
Irish-founded Ulysses raises $46m in rounds featuring A16Z
The San Francisco-based start-up is building networked autonomous vehicles that operate above and below the surface of the ocean, ‘Earth’s last frontier’. Read more: Irish-founded Ulysses raises $46m in rounds featuring A16Z
The 3 forces quietly dismantling the business model that made enterprise software fabulously profitable
Roundtables with senior business leaders in San Francisco and New York revealed a consensus: AI isn't just threatening SaaS, but the entire industry.
RiskWebWorld: A Realistic Interactive Benchmark for GUI Agents in E-commerce Risk Management
arXiv:2604.13531v1 Announce Type: new Abstract: Graphical User Interface (GUI) agents show strong capabilities for automating web tasks, but existing interactive benchmarks primarily target benign, predictable consumer environments. Their effectiveness in high-stakes, investigative domains such as authentic e-commerce risk management remains underexplored. To bridge this gap, we present RiskWebWorld, the first highly realistic interactive benchmark for evaluating GUI agents in e-commerce risk management. RiskWebWorld features 1,513 tasks sourced from production risk-control pipelines across 8 core domains, and captures the authentic challenges of risk operations on uncooperative websites, partially environmental hijackments. To support scalable evaluation and agentic reinforcement learning (RL), we further build a Gymnasium-compliant infrastructure that decouples policy planning from environment mechanics. Our evaluation across diverse models reveals a dramatic capability gap: top-tier generalist models achieve 49.1% success, while specialized open-weights GUI models lag at near-total failure. This highlights that foundation model scale currently matters more than zero-shot interface grounding in long-horizon professional tasks. We also demonstrate the viability of our infrastructure through agentic RL, which improves open-source models by 16.2%. These results position RiskWebWorld as a practical testbed for developing robust digital workers.
Mozilla throws Thunderbolt at enterprise AI providers
Client connects to deepset's Haystack platform Mozilla has declared war on OpenAI, Microsoft, and other firms flogging enterprise AI platforms with an open-source alternative it says provides data privacy guarantees proprietary products never could. …
AI companies are rethinking how they charge to win a bigger slice of business spending
AI firms are altering software pricing models, moving from per user fees to work-based charges.
Personalized pricing a ‘top priority,’ US FTC commissioners tell Congress
US FTC commissioners identified personalized pricing as a top priority during testimony before Congress.
The Next Evolution of the Agents SDK
OpenAI's updated Agents SDK introduces new primitives like configurable memory and native sandbox execution, signaling a move to compete with independent agent infrastructure vendors.
Foundational AI patent commons may add fuel to race for market leadership
Anthropic, Meta, Microsoft, Genentech (Roche Group) and IBM have established a shared commons for foundational artificial intelligence patents with the goal of reducing legal disputes and speeding development.
Big Tech generative AI practices emerge as antitrust concerns, Japan study finds
Japan's antitrust agency warned that generative AI integration into digital services and cloud provider licensing practices could shut out competition, citing an ongoing probe into Microsoft.
Scoop: BNY tests new OpenAI, Anthropic models
BNY has early access to OpenAI's and Anthropic's advanced cyber capability models, making the bank one of the few vetted enterprises with early access to these tools.
Antitrust on Aisle Five: How Well Do Divestiture Remedies Work?
arXiv:2604.15045v1 Announce Type: new Abstract: Antitrust authorities frequently rely on structural divestitures to address competitive concerns raised by mergers. Using census-level establishment data and proprietary transaction records from the U.S. grocery sector, we provide systematic evidence on the long-run effects of such remedies. Divested stores experience an average 31 percent decline in employment over five years, driven by elevated exit rates and persistent contraction among surviving establishments. Sales similarly decline. Transaction-level evidence indicates that divested assets are systematically weaker and are often transferred to lower-capability buyers. These findings suggest that structural remedies may be less effective when the implementation of divestitures allows merging parties substantial discretion over the assets and buyers involved.
Claude's Word Push Challenges Microsoft Copilot Adoption in 2026 - Windows News
Enterprise IT departments report ... appeals to organizations wary of major platform migrations or complex deployment processes. Early 2026 data shows a surprising pattern in enterprise AI adoption....
Monarch Tractors collapse ends in with an acquisition by Caterpillar
Caterpillar has acquired Monarch Tractors following the latter's business collapse.
Adobe embraces conversational AI editing, marking a ‘fundamental shift’ in creative work
Adobe is integrating conversational AI into its editing tools, signaling a major change in how creative professionals work.
Data Engineering Patterns for Cross-System Reconciliation in Regulated Enterprises: Architecture, Anomaly Detection, and Governance
arXiv:2604.15108v1 Announce Type: cross Abstract: Regulated enterprises in the United States--banks, telecommunications providers, large technology companies--operate across heterogeneous systems that were rarely designed to interoperate. ERP platforms, billing engines, supply chain tools, and financial reporting infrastructure coexist within the same organization, but they do not talk to each other well. The resulting fragmentation produces familiar problems: transactions recorded in one system but unreconciled in another, asset inventories drifting from their systems of record, and audit-readiness that depends on manual effort. The PCAOB's 2024 inspection cycle put a number on the consequences: a 39% aggregate Part I.A deficiency rate across all inspected firms. This paper introduces the GERA Framework (Governed Enterprise Reconciliation Architecture)--a vendor-neutral, four-layer data architecture that integrates deterministic cross-system reconciliation, statistical anomaly detection (baseline Z-Score with robust alternatives), governed semantic standardization, and NIST CSF 2.0-aligned security controls into a single methodology. The architecture spans four layers (ingestion, staging, core models, and semantic serving), following the multi-layer pattern now common in modern data platforms. The patterns are demonstrated through U.S. broadband operations--where billing reconciliation, inventory aging, and governance are tightly coupled--and draw on the author's implementation experience across three regulated enterprise environments: a regional bank, a national broadband provider, and a Fortune 500 technology company's central finance organization. This is a practitioner reference--an architectural framework paper documenting field-tested patterns--not a controlled experiment or benchmark study. No proprietary systems, datasets, or internal implementations are disclosed.
Pentagon Seeks Help From Ford and G.M.
Concerned about the slow pace and high cost of weapons production, Pentagon officials have begun talks with General Motors and Ford Motor about producing certain parts.
ReSS: Learning Reasoning Models for Tabular Data Prediction via Symbolic Scaffold
arXiv:2604.13392v1 Announce Type: new Abstract: Tabular data remains prevalent in high-stakes domains such as healthcare and finance, where predictive models are expected to provide both high accuracy and faithful, human-understandable reasoning. While symbolic models offer verifiable logic, they lack semantic expressiveness. Meanwhile, general-purpose LLMs often require specialized fine-tuning to master domain-specific tabular reasoning. To address the dual challenges of scalable data curation and reasoning consistency, we propose ReSS, a systematic framework that bridges symbolic and neural reasoning models. ReSS leverages a decision-tree model to extract instance-level decision paths as symbolic scaffolds. These scaffolds, alongside input features and labels, guide an LLM to generate grounded natural-language reasoning that strictly adheres to the underlying decision logic. The resulting high-quality dataset is used to fine-tune a pretrained LLM into a specialized tabular reasoning model, further enhanced by a scaffold-invariant data augmentation strategy to improve generalization and explainability. To rigorously assess faithfulness, we introduce quantitative metrics including hallucination rate, explanation necessity, and explanation sufficiency. Experimental results on medical and financial benchmarks demonstrate that ReSS-trained models improve traditional decision trees and standard fine-tuning approaches up to $10\%$ while producing faithful and consistent reasoning
The Token Reckoning
This article examines how flat-rate AI coding subscriptions are colliding with agentic workloads that consume significantly more tokens per task, particularly in compute-constrained markets.
Labor & Society
Startup Offers AI Tools to Find New Jobs for Workers Displaced by AI
Pelgo says it can provide placement services quicker and cheaper using AI
AI is destroying jobs – and the energy crisis could make that much worse | Larry Elliott | The Guardian
Every wave of new tech has come with a doomsday scenario. But governments just aren’t planning a human response on the scale required, says Guardian columnist Larry Elliott
India's IT jobs face AI challenges | The Straits Times
Explore how AI is transforming India's tech workforce, leading to job shifts and the urgent need for new skills. Read more at straitstimes.com.
Reveal Survey Identifies the Most In-Demand Technology Jobs and Skills
AI, Cybersecurity, and Data Analysts Are Most In-Demand Jobs as Talent Shortages Impact Technology Leaders...
Humanoids turn AI’s grand ambition physical
The race to give AI a body is accelerating as humanoid robotics technology advances.
A real issue with the current state of our knowledge on the work implications of AI is that there was a genuine discontinuity in AI ability with the rise of practical agentic systems in 2026.
A real issue with the current state of our knowledge on the work implications of AI is that there was a genuine discontinuity in AI ability with the rise of practical agentic systems in 2026. We were starting to get a picture of the impact of chatbots, no real data on agents.
As AI job replacement timeline speeds up, Cornell experts temper estimations
The 2026 Transform AI conference revealed new insights and fast-tracked projections for AI’s shift in the labor market.
Agentic Microphysics: A Manifesto for Generative AI Safety
arXiv:2604.15236v1 Announce Type: new Abstract: This paper advances a methodological proposal for safety research in agentic AI. As systems acquire planning, memory, tool use, persistent identity, and sustained interaction, safety can no longer be analysed primarily at the level of the isolated model. Population-level risks arise from structured interaction among agents, through processes of communication, observation, and mutual influence that shape collective behaviour over time. As the object of analysis shifts, a methodological gap emerges. Approaches focused either on single agents or on aggregate outcomes do not identify the interaction-level mechanisms that generate collective risks or the design variables that control them. A framework is required that links local interaction structure to population-level dynamics in a causally explicit way, allowing both explanation and intervention. We introduce two linked concepts. Agentic microphysics defines the level of analysis: local interaction dynamics where one agent's output becomes another's input under specific protocol conditions. Generative safety defines the methodology: growing phenomena and elicit risks from micro-level conditions to identify sufficient mechanisms, detect thresholds, and design effective interventions.
CAMO: An Agentic Framework for Automated Causal Discovery from Micro Behaviors to Macro Emergence in LLM Agent Simulations
arXiv:2604.14691v1 Announce Type: cross Abstract: LLM-empowered agent simulations are increasingly used to study social emergence, yet the micro-to-macro causal mechanisms behind macro outcomes often remain unclear. This is challenging because emergence arises from intertwined agent interactions and meso-level feedback and nonlinearity, making generative mechanisms hard to disentangle. To this end, we introduce \textbf{\textsc{CAMO}}, an automated \textbf{Ca}usal discovery framework from \textbf{M}icr\textbf{o} behaviors to \textbf{M}acr\textbf{o} Emergence in LLM agent simulations. \textsc{CAMO} converts mechanistic hypotheses into computable factors grounded in simulation records and learns a compact causal representation centered on an emergent target $Y$. \textsc{CAMO} outputs a computable Markov boundary and a minimal upstream explanatory subgraph, yielding interpretable causal chains and actionable intervention levers. It also uses simulator-internal counterfactual probing to orient ambiguous edges and revise hypotheses when evidence contradicts the current view. Experiments across four emergent settings demonstrate the promise of \textsc{CAMO}.
The Missing Knowledge Layer in AI: A Framework for Stable Human-AI Reasoning
arXiv:2604.14881v1 Announce Type: cross Abstract: Large language models are increasingly integrated into decision-making in areas such as healthcare, law, finance, engineering, and government. Yet they share a critical limitation: they produce fluent outputs even when their internal reasoning has drifted. A confident answer can conceal uncertainty, speculation, or inconsistency, and small changes in phrasing can lead to different conclusions. This makes LLMs useful assistants but unreliable partners in high-stakes contexts. Humans exhibit a similar weakness, often mistaking fluency for reliability. When a model responds smoothly, users tend to trust it, even when both model and user are drifting together. This paper is the first in a five-paper research series on stabilising human-AI reasoning. The series proposes a two-layer approach: Parts II-IV introduce human-side mechanisms such as uncertainty cues, conflict surfacing, and auditable reasoning traces, while Part V develops a model-side Epistemic Control Loop (ECL) that detects instability and modulates generation accordingly. Together, these layers form a missing operational substrate for governance by increasing signal-to-noise at the point of use. Stabilising interaction makes uncertainty and drift visible before enforcement is applied, enabling more precise capability governance. This aligns with emerging compliance expectations, including the EU AI Act and ISO/IEC 42001, by making reasoning processes traceable under real conditions of use. The central claim is that fluency is not reliability. Without structures that stabilise both human and model reasoning, AI cannot be trusted or governed where it matters most.
CoopEval: Benchmarking Cooperation-Sustaining Mechanisms and LLM Agents in Social Dilemmas
arXiv:2604.15267v1 Announce Type: cross Abstract: It is increasingly important that LLM agents interact effectively and safely with other goal-pursuing agents, yet, recent works report the opposite trend: LLMs with stronger reasoning capabilities behave _less_ cooperatively in mixed-motive games such as the prisoner's dilemma and public goods settings. Indeed, our experiments show that recent models -- with or without reasoning enabled -- consistently defect in single-shot social dilemmas. To tackle this safety concern, we present the first comparative study of game-theoretic mechanisms that are designed to enable cooperative outcomes between rational agents _in equilibrium_. Across four social dilemmas testing distinct components of robust cooperation, we evaluate the following mechanisms: (1) repeating the game for many rounds, (2) reputation systems, (3) third-party mediators to delegate decision making to, and (4) contract agreements for outcome-conditional payments between players. Among our findings, we establish that contracting and mediation are most effective in achieving cooperative outcomes between capable LLM models, and that repetition-induced cooperation deteriorates drastically when co-players vary. Moreover, we demonstrate that these cooperation mechanisms become _more effective_ under evolutionary pressures to maximize individual payoffs.
Microsoft, Salesforce Copilot/Agentforce prompt injection
Security researchers have identified vulnerabilities in Microsoft and Salesforce AI agents related to prompt injection.
Layered Mutability: Continuity and Governance in Persistent Self-Modifying Agents
arXiv:2604.14717v1 Announce Type: cross Abstract: Persistent language-model agents increasingly combine tool use, tiered memory, reflective prompting, and runtime adaptation. In such systems, behavior is shaped not only by current prompts but by mutable internal conditions that influence future action. This paper introduces layered mutability, a framework for reasoning about that process across five layers: pretraining, post-training alignment, self-narrative, memory, and weight-level adaptation. The central claim is that governance difficulty rises when mutation is rapid, downstream coupling is strong, reversibility is weak, and observability is low, creating a systematic mismatch between the layers that most affect behavior and the layers humans can most easily inspect. I formalize this intuition with simple drift, governance-load, and hysteresis quantities, connect the framework to recent work on temporal identity in language-model agents, and report a preliminary ratchet experiment in which reverting an agent's visible self-description after memory accumulation fails to restore baseline behavior. In that experiment, the estimated identity hysteresis ratio is 0.68. The main implication is that the salient failure mode for persistent self-modifying agents is not abrupt misalignment but compositional drift: locally reasonable updates that accumulate into a behavioral trajectory that was never explicitly authorized.
AI has an Awful Image problem
Modern-day Luddites are gaining ground because tech titans haven’t shown people how innovation will improve their lives
Claude Opus wrote a Chrome exploit for $2,283
Pause your Mythos panic because mainstream models anyone can use already pick holes in popular software Anthropic withheld its Mythos bug-finding model from public release due to concerns that it would enable attackers to find and exploit vulnerabilities before anyone could react.…
Seeking Help, Facing Harm: Auditing TikTok's Mental Health Recommendations
arXiv:2604.14832v1 Announce Type: cross Abstract: Recommender systems on social media increasingly mediate how users encounter mental health content, yet it remains unclear whether they distinguish help-seeking from distress expression. We conduct a controlled 7-day audit of TikTok's "For You" page using 30 fresh accounts and LLM-guided agents that vary initial search framing (distress- vs. help-initiated) and interaction strategy (engaged, avoidant, passive). Across 8,727 recommended videos, interaction behavior dominates exposure outcomes: engagement rapidly saturates feeds with mental health content (~45% of daily recommendations), while avoidance and passive viewing reduce but do not eliminate exposure (~11-20%). Search framing mainly shifts composition rather than volume--help-initiated searches yield more potentially supportive material, yet potentially harmful content persists at low but non-zero levels, including content in the Suicide/Self-Harm category. These findings suggest limited sensitivity to user intent signals in TikTok's recommendations and motivate context-aware safeguards for sensitive topics.
Listening Alone, Understanding Together: Collaborative Context Recovery for Privacy-Aware AI
arXiv:2604.13348v1 Announce Type: new Abstract: We introduce CONCORD, a privacy-aware asynchronous assistant-to-assistant (A2A) framework that leverages collaboration between proactive speech-based AI. As agents evolve from reactive to always-listening assistants, they face a core privacy risk (of capturing non-consenting speakers), which makes their social deployment a challenge. To overcome this, we implement CONCORD, which enforces owner-only speech capture via real-time speaker verification, producing a one-sided transcript that incurs missing context but preserves privacy. We demonstrate that CONCORD can safely recover necessary context through (1) spatio-temporal context resolution, (2) information gap detection, and (3) minimal A2A queries governed by a relationship-aware disclosure. Instead of hallucination-prone inferring, CONCORD treats context recovery as a negotiated safe exchange between assistants. Across a multi-domain dialogue dataset, CONCORD achieves 91.4% recall in gap detection, 96% relationship classification accuracy, and 97% true negative rate in privacy-sensitive disclosure decisions. By reframing always-listening AI as a coordination problem between privacy-preserving agents, CONCORD offers a practical path toward socially deployable proactive conversational agents.
AI Is Weaponizing Your Own Biases Against You
New research from MIT and Stanford suggests that AI systems can be used to exploit and amplify individual human biases.
Quantifying and Understanding Uncertainty in Large Reasoning Models
arXiv:2604.13395v1 Announce Type: new Abstract: Large Reasoning Models (LRMs) have recently demonstrated significant improvements in complex reasoning. While quantifying generation uncertainty in LRMs is crucial, traditional methods are often insufficient because they do not provide finite-sample guarantees for reasoning-answer generation. Conformal prediction (CP) stands out as a distribution-free and model-agnostic methodology that constructs statistically rigorous uncertainty sets. However, existing CP methods ignore the logical connection between the reasoning trace and the final answer. Additionally, prior studies fail to interpret the origins of uncertainty coverage for LRMs as they typically overlook the specific training factors driving valid reasoning. Notably, it is challenging to disentangle reasoning quality from answer correctness when quantifying uncertainty, while simultaneously establishing theoretical guarantees for computationally efficient explanation methods. To address these challenges, we first propose a novel methodology that quantifies uncertainty in the reasoning-answer structure with statistical guarantees. Subsequently, we develop a unified example-to-step explanation framework using Shapley values that identifies a provably sufficient subset of training examples and their key reasoning steps to preserve the guarantees. We also provide theoretical analyses of our proposed methods. Extensive experiments on challenging reasoning datasets verify the effectiveness of the proposed methods.
European banks can withstand current shocks, watchdog head says | Reuters
European lenders are resilient enough to absorb current financial and geopolitical shocks but need to prepare for future uncertainties including cybersecurity risks from AI, François-Louis Michaud, the new head of the European Banking Authority, said.
Liz Kendall urges UK public to embrace AI as government makes first £500m fund investment
Technology secretary plays down fears over jobs and cyber security as stake taken in British startup The UK technology secretary has urged the country to “make AI work for Britain”, brushing off fears about its impact on jobs and cybersecurity as the government announced its first investment under a £500m sovereign AI fund. Liz Kendall said the UK had to “seize” the opportunity offered by AI despite concerns underlined this month when US startup Anthropic revealed it had developed an AI model that posed a potentially significant cyber threat. Asked how the government makes the case for embracing a technology that could disrupt jobs and now cybersecurity, Kendall said: “We have to seize this to make it work, for Britain, for our jobs, for solving the biggest challenges we face as a world.” Speaking on Thursday as the government unveiled its first investment in a UK company as part of a £500m sovereign AI fund, Kendall acknowledged “people are worried about the risks and what it means for their jobs”, but AI entrepreneurs also believed they can “make it work … they can create jobs”. Continue reading...
It’s clear we won’t regulate AI for safety’s sake
As usual, governments will barely affect the trends that most affect our lives
European consumer protection chief brings digital regulation message to Silicon Valley
Michael McGrath, the European Commission's consumer protection chief, visited Silicon Valley to discuss the EU's planned Digital Fairness Act and potential cooperation with the California Privacy Protection Agency.
Cyber rules shift as geopolitics & AI reshape policy
NCC Group says geopolitics, digital sovereignty and AI are driving tougher cyber rules, with boards facing greater accountability and scrutiny.
Research reveals that EU AI rules stop at its borders with little accountability for human rights impacts abroad · Global Voices
On paper, many of these technologies are described as civilian. In practice, the line between civilian and military use is almost always blurred, particularly in authoritarian contexts and conflict zones.
South Korea moves to rationalize regulatory system under presidential body
South Korea is overhauling its regulatory framework by elevating its decision-making body to a presidential-level committee to better manage AI and other strategic industries.
Europe moves to restrict WhatsApp use for officials over AI, security concerns
In a recent update Europe is moving to ban WhatsApp for officials. Meta platforms has been threatened with an interim European Union ban on policies that allegedly prevent rival artificial...
What the EU AI Act requires for AI agent logging - Help Net Security
What the EU AI Act requires for AI agent logging: the four articles that matter, key deadlines, and where the compliance gaps are.
US FTC shouldn’t be lead AI enforcer, Chairman Ferguson says
US Federal Trade Commission Chairman Andrew Ferguson stated the agency should not become the general all-purpose AI regulator.
Udio will continue to face DMCA claim over YouTube scraping, US judge orders
A US judge declined to dismiss claims that AI music company Udio violated the Digital Millennium Copyright Act by scraping songs from YouTube.
Vietnam's draft copyright decree risks undermining AI training exception
Vietnam's draft copyright decree could render the amended IP law's new text-and-data mining exception ineffective, creating uncertainty for AI developers and raising risks for domestic innovation.
AI agents covered by EU rules 'mainly' as general-purpose AI systems, Virkkunen says
EU technology chief Henna Virkkunen stated that AI agents fall under the AI Act as general-purpose AI systems, with the European Commission continuing to assess associated risks.
Lawmakers gathered quietly to talk about AI. Angst and fears of 'destruction' followed - Newsday
Lawmakers on Capitol Hill conducted a roundtable with leading AI executives and academics to discuss the potential transformative impacts of the technology on American society.
These companies are actually upskilling their workers for AI - here's how they do it | ZDNET
Business leaders explain why they're still hiring for entry-level jobs, how they're investing in employees, and what's working so far.
Stanford AI Index 2026 shows AI adoption pressure on education and jobs | ETIH EdTech News — EdTech Innovation Hub
AI in education, edtech AI tools, and AI skills demand surge in Stanford AI Index 2026. Up to 84 percent of students use AI, while jobs, infrastructure, and policy struggle to keep pace. ETIH edtech news covers digital learning, workforce skills, and AI adoption impact
Technology & Infrastructure
Apple Upgraded as BNP Paribas Sees Path to Increase Market Share
Apple Inc. was upgraded to outperform from neutral by BNP Paribas, which said the iPhone maker can use the recent spike in memory prices as a way to increase its share of smartphone market share.
Ericsson Targets Networks Growth Despite Caution Over Rising Costs
Chief Executive Borje Ekholm said it was facing increasing input costs, especially in semiconductors, caused in part by AI demand.
Swedish semiconductor startup AlixLabs closes €15 million Series A to scale atomic-level etching technology
AlixLabs, a Lund-based semiconductor startup specialising in Atomic Layer Etching (ALE) technology, has announced the closing of its €15 million Series A in the first quarter of 2026, following a strategic investment from Stephen Industries, a Finnish investment company. In November 2025, AlixLabs announced that Global Brain and key institutional investors had subscribed to its […]
Interview: Semidynamics expands from SoC to rack-level solutions, targeting memory-intensive AI inference
Spain-based AI chip startup Semidynamics has recently seen a wave of positive developments. Its first chip, fabricated on TSMC's 3nm process, has successfully completed tape-out, and just weeks ago, the company secured an investment commitment from SK Hynix. Both significantly enhance its ...
TSMC's 2nm Supply Crunch Is Forcing Smartphone Makers to Reserve Top Chipsets for 'Ultra' Models Only, as DRAM Shortage Piles On
Even though 2nm chipset launches are expected in Q4 2026, one tipster claims that TSMC will be partly responsible for smartphone companies making downgrades
Nous Research introduces Tool Gateway
Nous Research has rolled out a unified system that integrates scraping, automation, and generation tools for model access.
Mythos cyber scare signals the economics of AI scarcity
As capabilities of frontier models advance, gaining access to technology could become critically important
AI Memory Products Market to Reach USD 92.4 Billion by 2031, Driven by Explosive Demand for HBM and AI Servers | Valuates Reports
/PRNewswire/ -- What is the Market Size of AI Memory Products? The Global market for AI Memory Products was valued at USD 3079 Million in the year 2024 and is...
Prefill Is Compute-Bound. Decode Is Memory-Bound. Why Your GPU Shouldn’t Do Both.
Inside disaggregated LLM inference — the architecture shift behind 2-4x cost reduction that most ML teams haven't adopted yet.
Rethinking AI Hardware: A Three-Layer Cognitive Architecture for Autonomous Agents
arXiv:2604.13757v1 Announce Type: new Abstract: The next generation of autonomous AI systems will be constrained not only by model capability, but by how intelligence is structured across heterogeneous hardware. Current paradigms -- cloud-centric AI, on-device inference, and edge-cloud pipelines -- treat planning, reasoning, and execution as a monolithic process, leading to unnecessary latency, energy consumption, and fragmented behavioral continuity. We introduce the Tri-Spirit Architecture, a three-layer cognitive framework that decomposes intelligence into planning (Super Layer), reasoning (Agent Layer), and execution (Reflex Layer), each mapped to distinct compute substrates and coordinated via an asynchronous message bus. We formalize the system with a parameterized routing policy, a habit-compilation mechanism that promotes repeated reasoning paths into zero-inference execution policies, a convergent memory model, and explicit safety constraints. We evaluate the architecture in a reproducible simulation of 2000 synthetic tasks against cloud-centric and edge-only baselines. Tri-Spirit reduces mean task latency by 75.6 percent and energy consumption by 71.1 percent, while decreasing LLM invocations by 30 percent and enabling 77.6 percent offline task completion. These results suggest that cognitive decomposition, rather than model scaling alone, is a primary driver of system-level efficiency in AI hardware.
Nvidia Helps Make Quantum Computing CEO a Billionaire in Days
The founder of quantum computing startup Xanadu Quantum Technologies Ltd. became a billionaire in a matter of days after Nvidia Corp. threw its weight behind the emerging technology.
IOWN Global Forum targets datacenter interconnects to scatter AI infrastructure
Fast WAN consortium thinks neoclouds are ripe for hookups The IOWN Global Forum will likely focus on datacenter interconnect use cases in the, to help diverse providers of AI infrastructure ply their trade.…
AlphaCNOT: Learning CNOT Minimization with Model-Based Planning
arXiv:2604.13812v1 Announce Type: new Abstract: Quantum circuit optimization is a central task in Quantum Computing, as current Noisy Intermediate Scale Quantum devices suffer from error propagation that often scales with the number of operations. Among quantum operations, the CNOT gate is of fundamental importance, being the only 2-qubit gate in the universal Clifford+T set. The problem of CNOT gates minimization has been addressed by heuristic algorithms such as the well-known Patel-Markov-Hayes (PMH) for linear reversible synthesis (i.e., CNOT minimization with no topological constraints), and more recently by Reinforcement Learning (RL) based strategies in the more complex case of topology-aware synthesis, where each CNOT can act on a subset of all qubits pairs. In this work we introduce AlphaCNOT, a RL framework based on Monte Carlo Tree Search (MCTS) that address effectively the CNOT minimization problem by modeling it as a planning problem. In contrast to other RL- based solution, our method is model-based, i.e. it can leverage lookahead search to evaluate future trajectories, thus finding more efficient sequences of CNOTs. Our method achieves a reduction of up to 32% in CNOT gate count compared to PMH baseline on linear reversible synthesis, while in the constraint version we report a consistent gate count reduction on a variety of topologies with up to 8 qubits, with respect to state-of-the-art RL-based solutions. Our results suggest the combination of RL with search-based strategies can be applied to different circuit optimization tasks, such as Clifford minimization, thus fostering the transition toward the "quantum utility" era.
Users complain that UK Azure is having capacity problems
We hear Sweden is lovely place for workloads to visit Microsoft Azure capacity woes are back, and worse than ever, judging by the complaints of UK users.…
Volt announces ‘AI Gigafactory’ in Rotterdam, Netherlands
Dutch AI data center company Volt has announced the development of an “AI Gigafactory” in Rotterdam. Details of the project are sparse, but the company said it will offer “compute at industrial scale .
Is Energy the Real AI Bottleneck? What Investors Need to Know | The Motley Fool
U.S. AI data centers could consume 9–17% of U.S. electricity by 2030, up from 4–5% today. Here's what that means for investors
AI scaling now depends on efficiency
Scaling AI requires efficient compute that can deliver both performance and scale. As demand grows, data centers must do more within fixed power and space constraints.
Hong Kong to host China's chip hub as Beijing seeks tech edge
Hong Kong is set to host a national manufacturing-innovation center focused on power semiconductors to support China's drive for semiconductor self-sufficiency.
Towards Scalable Lightweight GUI Agents via Multi-role Orchestration
arXiv:2604.13488v1 Announce Type: new Abstract: Autonomous Graphical User Interface (GUI) agents powered by Multimodal Large Language Models (MLLMs) enable digital automation on end-user devices. While scaling both parameters and data has yielded substantial gains, advanced methods still suffer from prohibitive deployment costs on resource-constrained devices. When facing complex in-the-wild scenarios, lightweight GUI agents are bottlenecked by limited capacity and poor task scalability under end-to-end episodic learning, impeding adaptation to multi-agent systems (MAS), while training multiple skill-specific experts remains costly. Can we strike an effective trade-off in this cost-scalability dilemma, enabling lightweight MLLMs to participate in realistic GUI workflows? To address these challenges, we propose the LAMO framework, which endows a lightweight MLLM with GUI-specific knowledge and task scalability, allowing multi-role orchestration to expand its capability boundary for GUI automation. LAMO combines role-oriented data synthesis with a two-stage training recipe: (i) supervised fine-tuning with Perplexity-Weighted Cross-Entropy optimization for knowledge distillation and visual perception enhancement, and (ii) reinforcement learning for role-oriented cooperative exploration. With LAMO, we develop a task-scalable native GUI agent, LAMO-3B, supporting monolithic execution and MAS-style orchestration. When paired with advanced planners as a plug-and-play policy executor, LAMO-3B can continuously benefit from planner advances, enabling a higher performance ceiling. Extensive static and online evaluations validate the effectiveness of our design.
Data Farms To Power AI Are A New 'Not In My Back Yard' Flashpoint | Across America, US Patch
In town halls across the country, residents push back against data centers as some state legislatures question if tax breaks went too far.
AI Business | LinkedIn
Coverage includes developments and trends in generative AI , large language models, machine learning, regulation, governance, cybersecurity and ethics.
Nervous Indian Fintechs Push Anthropic for Access to Mythos
India’s major financial technology firms are pushing Anthropic PBC to give them early access to Mythos, the artificial intelligence model that has sparked global fears about a new era of cyberattacks.
AI’s Next Frontier: People Skills
Imagine a chatbot that actually knows how to talk to you.
Frontier models are failing one in three production attempts and getting harder to audit
A report highlights the reliability issues and auditing challenges facing current frontier AI models in production environments.
The cognitive companion: a lightweight parallel monitoring architecture for detecting and recovering from reasoning degradation in LLM agents
arXiv:2604.13759v1 Announce Type: new Abstract: Large language model (LLM) agents on multi-step tasks suffer reasoning degradation, looping, drift, stuck states, at rates up to 30% on hard tasks. Current solutions include hard step limits (abrupt) or LLM-as-judge monitoring (10-15% overhead per step). This paper introduces the Cognitive Companion, a parallel monitoring architecture with two implementations: an LLM-based Companion and a novel zero-overhead Probe-based Companion. We report a three-batch feasibility study centered on Gemma 4 E4B, with an additional exploratory small-model analysis on Qwen 2.5 1.5B and Llama 3.2 1B. In our experiments, the LLM-based Companion reduced repetition on loop-prone tasks by 52-62% with approximately 11% overhead. The Probe-based Companion, trained on hidden states from layer 28, showed a mean effect size of +0.471 at zero measured inference overhead; its strongest probe result achieved cross-validated AUROC 0.840 on a small proxy-labeled dataset. A key empirical finding is that companion benefit appears task-type dependent: companions are most helpful on loop-prone and open-ended tasks, while effects are neutral or negative on more structured tasks. Our small-model experiments also suggest a possible scale boundary: companions did not improve the measured quality proxy on 1B-1.5B models, even when interventions fired. Overall, the paper should be read as a feasibility study rather than a definitive validation. The results provide encouraging evidence that sub-token monitoring may be useful, identify task-type sensitivity as a practical design constraint, and motivate selective companion activation as a promising direction for future work.
Meta researchers introduce Hyperagents to unlock self-improving AI for non-coding tasks
Meta has unveiled 'Hyperagents,' a new approach designed to enable AI to improve itself across various non-programming domains.
Numerical Instability and Chaos: Quantifying the Unpredictability of Large Language Models
arXiv:2604.13206v1 Announce Type: new Abstract: As Large Language Models (LLMs) are increasingly integrated into agentic workflows, their unpredictability stemming from numerical instability has emerged as a critical reliability issue. While recent studies have demonstrated the significant downstream effects of these instabilities, the root causes and underlying mechanisms remain poorly understood. In this paper, we present a rigorous analysis of how unpredictability is rooted in the finite numerical precision of floating-point representations, tracking how rounding errors propagate, amplify, or dissipate through Transformer computation layers. Specifically, we identify a chaotic "avalanche effect" in the early layers, where minor perturbations trigger binary outcomes: either rapid amplification or complete attenuation. Beyond specific error instances, we demonstrate that LLMs exhibit universal, scale-dependent chaotic behaviors characterized by three distinct regimes: 1) a stable regime, where perturbations fall below an input-dependent threshold and vanish, resulting in constant outputs; 2) a chaotic regime, where rounding errors dominate and drive output divergence; and 3) a signal-dominated regime, where true input variations override numerical noise. We validate these findings extensively across multiple datasets and model architectures.
Exploration and Exploitation Errors Are Measurable for Language Model Agents
arXiv:2604.13151v1 Announce Type: new Abstract: Language Model (LM) agents are increasingly used in complex open-ended decision-making tasks, from AI coding to physical AI. A core requirement in these settings is the ability to both explore the problem space and exploit acquired knowledge effectively. However, systematically distinguishing and quantifying exploration and exploitation from observed actions without access to the agent's internal policy remains challenging. To address this, we design controllable environments inspired by practical embodied AI scenarios. Each environment consists of a partially observable 2D grid map and an unknown task Directed Acyclic Graph (DAG). The map generation can be programmatically adjusted to emphasize exploration or exploitation difficulty. To enable policy-agnostic evaluation, we design a metric to quantify exploration and exploitation errors from agent's actions. We evaluate a variety of frontier LM agents and find that even state-of-the-art models struggle on our task, with different models exhibiting distinct failure modes. We further observe that reasoning models solve the task more effectively and show both exploration and exploitation can be significantly improved through minimal harness engineering. We release our code \href{https://github.com/jjj-madison/measurable-explore-exploit}{here}.
AI Isn't Read In - by Soban, The Stranger
AI is primarily trained on written data from the internet. AI , however, also is primarily used to generate written data to put on the internet. This means that with every new round of AI training, the input data that the AI is being trained on includes more and more of the output from previous AI models.
Weight Patching: Toward Source-Level Mechanistic Localization in LLMs
arXiv:2604.13694v1 Announce Type: new Abstract: Mechanistic interpretability seeks to localize model behavior to the internal components that causally realize it. Prior work has advanced activation-space localization and causal tracing, but modules that appear important in activation space may merely aggregate or amplify upstream signals rather than encode the target capability in their own parameters. To address this gap, we propose Weight Patching, a parameter-space intervention method for source-oriented analysis in paired same-architecture models that differ in how strongly they express a target capability under the inputs of interest. Given a base model and a behavior-specialized counterpart, Weight Patching replaces selected module weights from the specialized model into the base model under a fixed input. We instantiate the method on instruction following and introduce a framework centered on a vector-anchor behavioral interface that provides a shared internal criterion for whether a task-relevant control state has been formed or recovered in open-ended generation. Under this framework, the analysis reveals a hierarchy from shallow candidate source-side carriers to aggregation and routing modules, and further to downstream execution circuits. The recovered component scores can also guide mechanism-aware model merging, improving selective fusion across the evaluated expert combinations and providing additional external validation.
Adoption & Impact
GeoAgentBench: A Dynamic Execution Benchmark for Tool-Augmented Agents in Spatial Analysis
arXiv:2604.13888v1 Announce Type: new Abstract: The integration of Large Language Models (LLMs) into Geographic Information Systems (GIS) marks a paradigm shift toward autonomous spatial analysis. However, evaluating these LLM-based agents remains challenging due to the complex, multi-step nature of geospatial workflows. Existing benchmarks primarily rely on static text or code matching, neglecting dynamic runtime feedback and the multimodal nature of spatial outputs. To address this gap, we introduce GeoAgentBench (GABench), a dynamic and interactive evaluation benchmark tailored for tool-augmented GIS agents. GABench provides a realistic execution sandbox integrating 117 atomic GIS tools, encompassing 53 typical spatial analysis tasks across 6 core GIS domains. Recognizing that precise parameter configuration is the primary determinant of execution success in dynamic GIS environments, we designed the Parameter Execution Accuracy (PEA) metric, which utilizes a "Last-Attempt Alignment" strategy to quantify the fidelity of implicit parameter inference. Complementing this, a Vision-Language Model (VLM) based verification is proposed to assess data-spatial accuracy and cartographic style adherence. Furthermore, to address the frequent task failures caused by parameter misalignments and runtime anomalies, we developed a novel agent architecture, Plan-and-React, that mimics expert cognitive workflows by decoupling global orchestration from step-wise reactive execution. Extensive experiments with seven representative LLMs demonstrate that the Plan-and-React paradigm significantly outperforms traditional frameworks, achieving the optimal balance between logical rigor and execution robustness, particularly in multi-step reasoning and error recovery. Our findings highlight current capability boundaries and establish a robust standard for assessing and advancing the next generation of autonomous GeoAI.
The Tyranny of AI Everywhere
Sneakers? Why stop there?
AI News Digest, April 16: AI Agents Are Now Running Payroll. The Governance Gap Is Real.
AI payroll agents are now live across 40+ countries. Apr 16: enterprise agent sprawl, the HR readiness gap, and EU AI Act's political fight ahead.
WebXSkill: Skill Learning for Autonomous Web Agents
arXiv:2604.13318v1 Announce Type: new Abstract: Autonomous web agents powered by large language models (LLMs) have shown promise in completing complex browser tasks, yet they still struggle with long-horizon workflows. A key bottleneck is the grounding gap in existing skill formulations: textual workflow skills provide natural language guidance but cannot be directly executed, while code-based skills are executable but opaque to the agent, offering no step-level understanding for error recovery or adaptation. We introduce WebXSkill, a framework that bridges this gap with executable skills, each pairing a parameterized action program with step-level natural language guidance, enabling both direct execution and agent-driven adaptation. WebXSkill operates in three stages: skill extraction mines reusable action subsequences from readily available synthetic agent trajectories and abstracts them into parameterized skills, skill organization indexes skills into a URL-based graph for context-aware retrieval, and skill deployment exposes two complementary modes, grounded mode for fully automated multi-step execution and guided mode where skills serve as step-by-step instructions that the agent follows with its native planning. On WebArena and WebVoyager, WebXSkill improves task success rate by up to 9.8 and 12.9 points over the baseline, respectively, demonstrating the effectiveness of executable skills for web agents. The code is publicly available at https://github.com/aiming-lab/WebXSkill.
We tested Anthropic's redesigned Claude Code desktop app
An evaluation of the updated Claude Code desktop application and its implications for enterprise workflows.
AI For Smarter Regulatory Filings And Pharma Factories – blog.aimactgrow.com
How AI Makes Regulatory Filings and Pharma Factories SmarterSynthetic intelligence is reshaping how medicines are developed, manufactured, and accredited, but
Low-Cost System for Automatic Recognition of Driving Pattern in Assessing Interurban Mobility using Geo-Information
arXiv:2604.15216v1 Announce Type: cross Abstract: Mobility in urban and interurban areas, mainly by cars, is a day-to-day activity of many people. However, some of its main drawbacks are traffic jams and accidents. Newly made vehicles have pre-installed driving evaluation systems, which can prevent accidents. However, most cars on our roads do not have driver assessment systems. In this paper, we propose an approach for recognising driving styles and enabling drivers to reach safer and more efficient driving. The system consists of two physical sensors connected to a device node with a display and a speaker. An artificial neural network (ANN) is included in the node, which analyses the data from the sensors, and then recognises the driving style. When an abnormal driving pattern is detected, the speaker will play a warning message. The prototype was assembled and tested using an interurban road, in particular on a conventional road with three driving styles. The gathered data were used to train and validate the ANN. Results, in terms of accuracy, indicate that better accuracy is obtained when the velocity, position (latitude and longitude), time, and turning speed for the 3-axis are used, offering an average accuracy of 83%. If the classification is performed considering just two driving styles, normal and aggressive, then the accuracy reaches 92%. When the geo-information and time data are included, the main novelty of this paper, the classification accuracy is improved by 13%.
The adoption trap
But they are also not yet sufficient ... we are deploying. The Deloitte report, perhaps unintentionally, makes this point with its own language. The most common responses Indian organisations cite for building AI capability include upskilling and reskilling programs at 61%, incentives to drive adoption at 59%, and ...
Its noticeable how much of the whole practice of working with AI - the prompts, the skill files, the connectors, retrieval work, the markdown files, etc. - is a substitute for the real problem of continual learning. If that ends up being solved, a lot of things will change fast.
Its noticeable how much of the whole practice of working with AI - the prompts, the skill files, the connectors, retrieval work, the markdown files, etc. - is a substitute for the real problem of continual learning. If that ends up being solved, a lot of things will change fast.
Most people are only getting a fraction of the value they could be getting from AI tools. It's a huge "capabilities overhang". Along with the amazing Cat Goetze and Parth Patil, I've created a @MasterClass that quickly shows you how to operate at the scale of a small team with a set of easy-to-use AI tools. Check out the link below.
Most people are only getting a fraction of the value they could be getting from AI tools. It's a huge "capabilities overhang". Along with the amazing Cat Goetze and Parth Patil, I've created a @MasterClass that quickly shows you how to operate at the scale of a small team with a set of easy-to-use AI tools. Check out the link below.
How Guidesly Built AI-Generated Trip Reports for Outdoor Guides on AWS
Guidesly's Jack AI leverages AWS services like Amazon Bedrock and SageMaker to automate the creation of marketing-ready trip reports and social content for outdoor guides.
SciFi: A Safe, Lightweight, User-Friendly, and Fully Autonomous Agentic AI Workflow for Scientific Applications
arXiv:2604.13180v1 Announce Type: new Abstract: Recent advances in agentic AI have enabled increasingly autonomous workflows, but existing systems still face substantial challenges in achieving reliable deployment in real-world scientific research. In this work, we present a safe, lightweight, and user-friendly agentic framework for the autonomous execution of well-defined scientific tasks. The framework combines an isolated execution environment, a three-layer agent loop, and a self-assessing do-until mechanism to ensure safe and reliable operation while effectively leveraging large language models of varying capability levels. By focusing on structured tasks with clearly defined context and stopping criteria, the framework supports end-to-end automation with minimal human intervention, enabling researchers to offload routine workloads and devote more effort to creative activities and open-ended scientific inquiry.
Synthetic Reflections on Resource Extraction
arXiv:2602.09299v2 Announce Type: replace Abstract: This paper describes how AI models can be augmented and adapted to interpret landscapes. We present the technical framework of a Sentinel-2 satellite asset interpretation pipeline that combines statistical operations, human judgment, and generative AI models to produce succinct commentaries on industrial mining sites across the planet. To this end we introduce a novel bespoke landscape descriptor, the Urban Dwelling and Mining Index, and discuss how this metric can improve the performance of a multimodal language model in assessing the spatial distribution of mining operations.
Optimizing Earth Observation Satellite Schedules under Unknown Operational Constraints: An Active Constraint Acquisition Approach
arXiv:2604.13283v1 Announce Type: new Abstract: Earth Observation (EO) satellite scheduling (deciding which imaging tasks to perform and when) is a well-studied combinatorial optimization problem. Existing methods typically assume that the operational constraint model is fully specified in advance. In practice, however, constraints governing separation between observations, power budgets, and thermal limits are often embedded in engineering artefacts or high-fidelity simulators rather than in explicit mathematical models. We study EO scheduling under \emph{unknown constraints}: the objective is known, but feasibility must be learned interactively from a binary oracle. Working with a simplified model restricted to pairwise separation and global capacity constraints, we introduce Conservative Constraint Acquisition~(CCA), a domain-specific procedure designed to identify justified constraints efficiently in practice while limiting unnecessary tightening of the learned model. Embedded in the \textsc{Learn\&Optimize} framework, CCA supports an interactive search process that alternates optimization under a learned constraint model with targeted oracle queries. On synthetic instances with up to 50~tasks and dense constraint networks, L\&O improves over a no-knowledge greedy baseline and uses far fewer main oracle queries than a two-phase acquire-then-solve baseline (FAO). For $n\leq 30$, the average gap drops from 65--68\% (Priority Greedy) to 17.7--35.8\% using L\&O. At $n{=}50$, where the CP-SAT reference is the best feasible solution found in 120~s, L\&O improves on FAO on average (17.9\% vs.\ 20.3\%) while using 21.3 main queries instead of 100 and about $5\times$ less execution time.
Spatiotemporal Analysis of VIIRS Satellite Observations and Network Traffic During the 2025 Manitoba Wildfires
arXiv:2604.14392v1 Announce Type: new Abstract: Climate change has intensified extreme weather and wildfire conditions globally. Canada experienced record-breaking wildfires in 2023 and 2025, burning millions of hectares and severely impacting the Prairie provinces, with Manitoba facing its worst season in 30 years. These events highlight the urgent need to understand and mitigate escalating fire risks. While existing research largely focuses on wildfire management approaches, few studies have explored the relationship between user network traffic and wildfire activity, despite the potential of such correlations to provide valuable spatiotemporal insights into wildfire dynamics. This paper investigates the relationship between wildfire intensity and network performance during the 2025 Manitoba wildfire season, using Visible Infrared Imaging Radiometer Suite (VIIRS) satellite-derived Fire Radiative Power data and large-scale Speedtest measurements. We found statistically significant correlations between wildfire intensity and several network performance metrics in both the province-wide and region-wide case studies, as measured by Spearman's correlation coefficients ($\rho$) and corresponding p-values. Throughput-related metrics showed inverse correlations with wildfire intensity (e.g., download speed: $\rho = -0.214$, $p\_value = 0.004$), whereas latency-related metrics showed positive correlations (e.g., round-trip time latency: $\rho = 0.162$, $p\_value = 0.0308$). The findings suggest satellite fire indicators and network performance metrics together can reveal vulnerabilities during extreme environmental events and support diaster response and recovery efforts.
Enhancing Large Language Model-Based Systems for End-to-End Circuit Analysis Problem Solving
arXiv:2512.10159v2 Announce Type: replace Abstract: LLMs have demonstrated strong performance in data-rich domains such as programming, yet their reliability in engineering tasks remains limited. Circuit analysis--requiring multimodal understanding and precise mathematical reasoning--highlights these challenges. Although Gemini 2.5 Pro shows improved capabilities in diagram interpretation and analog-circuit reasoning, it still struggles to consistently produce correct solutions when given both textual problem descriptions and circuit diagrams. Meanwhile, engineering education demands scalable AI tools capable of generating accurate solutions for applications such as automated homework feedback. This paper presents an enhanced end-to-end circuit problem-solving framework built upon Gemini. We first conduct a systematic benchmark on undergraduate circuit problems and identify two key failure modes: 1) circuit-recognition hallucinations, particularly incorrect source polarity detection, and 2) reasoning-process hallucinations, such as incorrect current direction assumptions. To address recognition errors, we integrate a fine-tuned YOLO detector and OpenCV-based processing to isolate voltage and current sources, enabling Gemini to accurately re-identify source polarities from cropped images. To mitigate reasoning errors, we introduce an ngspice-driven verification loop, in which simulation discrepancies trigger iterative solution refinement with optional HITL feedback. Experimental results demonstrate that the proposed pipeline achieves 97.59% accuracy, substantially outperforming Gemini's baseline of 79.52%. Furthermore, on four variations of hand-drawn circuit diagrams, accuracy improves from 56.06%--71.21% to 93.94%--95.45% with statistically significant gains. These results highlight the robustness, scalability, and practical applicability of the proposed framework for engineering education and real-world circuit analysis tasks.
AI tech collab takes on SA’s energy distribution crisis | ITWeb
Energy tech firm Plentify partners with local residential developer to roll out AI-driven home energy management system.
Novartis CEO joins Anthropic board, embedding AI in the heart of biopharma - PharmaLive
In recent months, Anthropic has been building more and more ties with the biopharma industry, including partnerships with Big Pharma companies such as Sanofi, Novo Nordisk and AbbVie.
Filmmakers defend Val Kilmer movie made with AI | Reuters
The makers of a new film with an AI-generated performance by late actor Val Kilmer defended their work on Thursday and said they believed their approach demonstrated an ethical path to future use of the technology.
Geopolitics
China's AI Chip Self-Sufficiency Hits 41%, Korea Slips to 5th in Semiconductors - Seoul Economic Daily
Chinese firms captured 41% of their AI chip market, with self-sufficiency projected at 76% by 2030. Korea slipped to 5th in semiconductor competitiveness as Samsung and SK hynix secure two-thirds of ASML's EUV shipments.
Algorithms At War: Why The World Needs Rules Before AI Writes The Next Battlefield | Outlook India
Failure to secure influence over AI ecosystems risks forfeiting control over not just technology, but also economic competitiveness, governance systems, and geopolitical standing
Get the full executive brief
Receive curated insights with practical implications for strategy, operations, and governance.