AI Scale

Google has been told to temporarily halt construction of its datacentres in Finland while environmental impact assessments are carried out. The order reflects growing scrutiny of the water, energy and land demands of large-scale datacentre projects in the Nordic region, where tech companies have invested heavily to take advantage of cool climates and low-carbon power. The move signals tighter oversight of big tech infrastructure projects in Finland.

A professor at the Indian Institute of Science, Bengaluru, is drawing wide attention after describing the rise of artificial intelligence as 'a civilisational crisis'. In a widely shared interview, he discusses what AI means for education, the future of work and how universities should respond, arguing that teaching and employment face disruption on a scale society is not prepared for.

SoftBank is seeking to raise $100 billion from investors in the Gulf to fund an expanded push into artificial intelligence, according to a Financial Times report. The Japanese tech conglomerate, led by Masayoshi Son, has been aggressively repositioning itself around AI, and sovereign wealth funds in the region have become major backers of large-scale AI ventures.

A commentary in the National Catholic Reporter argues that theological education is becoming more important, not less, in an era shaped by artificial intelligence. The piece contends that questions of meaning, ethics and human purpose raised by AI require the kind of formation that theology programs provide, and that religious institutions should invest in such education rather than scale it back.

QuEra Computing has detailed how it uses AI to automatically lock and tune the lasers that control its quantum computers. Keeping hundreds of lasers precisely aligned is a major operational burden in neutral-atom quantum systems, and automating the process with machine learning reduces manual intervention and drift-related downtime. The technical write-up has drawn attention among quantum computing and engineering communities discussing the practical challenges of scaling quantum hardware.

A new commentary argues that the United Kingdom's push to adopt artificial intelligence in its armed forces is moving faster than the oversight meant to govern it. The piece says Britain's military AI ambitions require clearer guardrails on accountability, safety and use of force before the technology is deployed at scale, adding to a wider debate over how democracies regulate AI in defence.

Carbon-A, a DNA annotation model, has been used to identify 566 million previously unlogged gene candidates, with the work documented in a new technical write-up covering genome annotation at GenBank scale. The result suggests large-scale machine learning can flag vast numbers of potential genes that standard pipelines missed. Discussion is centered on how reliable these candidates are and what validating them at such volume would require.

Anthropic says code produced by its Claude AI assistant outgrew git: daily merges in its monorepo reportedly jumped from 2,500 to over 30,000 between May and September 2026. The company built Scion, an in-house version control system developed in five months, which does without a full copy of the repository anywhere. The move is being discussed as a striking example of AI-generated code straining developer infrastructure.

Biotech startup Giraffe Bio has raised a $6.4 million seed round to help miners extract more copper, aiming bespoke biomolecules at an anticipated shortage of the metal. Demand for copper is climbing as AI data centers and electric vehicles scale up, and investors are betting biology-based extraction methods could unlock more supply from existing mines.

SpaceX is preparing a giant fundraising round aimed at securing Nvidia chips and expanding its artificial intelligence capabilities. The move, tied to Elon Musk's broader AI ambitions, would give the company more computing muscle for data centers, reportedly including infrastructure linked to projects like Colossus. The story is drawing attention on Wall Street given the scale of the financing and the intensifying race for semiconductors.

The Wall Street Journal examines whether politicians in Ohio and Michigan will face voter backlash over data center development, as large-scale projects spark concerns about energy costs, water use, and strain on local infrastructure. The piece asks whether the rapid expansion of data centers, driven by the AI boom, is becoming a political liability in key Midwest states ahead of future elections.

Chipmakers racing to power artificial intelligence are increasingly turning to advanced packaging as a key source of competitive advantage. As traditional transistor scaling slows, companies are focusing on how chips are assembled and connected, using techniques that stack and link components to boost performance and efficiency in AI processors.

AI company Anthropic has issued a warning that botnets powered by artificial intelligence could spread across internet-connected devices. The concern is that AI could automate and scale the hijacking of everyday connected hardware, turning large numbers of devices into coordinated networks used for attacks or other malicious activity, raising new cybersecurity risks.

A new argument is making the rounds in AI circles: curating high-quality training data delivers more performance gains than simply adding compute. The claim, laid out in a recent blog post from data curation company Datology AI, has drawn attention on Hacker News, where commenters are debating whether better data selection really outperforms brute-force scaling in model training.

ASML, the Dutch maker of advanced lithography machines used to produce cutting-edge chips, has warned it will not be able to meet demand for its equipment in 2027 or 2028. The statement highlights the scale of the global semiconductor build-out, as chipmakers race to add capacity for AI-driven demand while ASML's production of its complex tools remains limited.

Character.AI has filed for Chapter 11 bankruptcy protection, according to court papers, despite having around 12 million active users. The company has never been profitable and was reportedly losing more than $8 million a month. Disney has emerged as the opening bidder for some of its assets at $20 million, while the app remains operational during the sale process. Commenters note the case shows that a large user base alone does not guarantee a viable business.

A new report has found that artificial intelligence was involved in 84% of ransomware attacks, a finding flagged by CBS News. The figure suggests attackers are increasingly using AI tools to craft phishing lures, evade detection or scale their operations. The report is likely to fuel debate over the security risks of rapidly adopted AI technology.

A new technical guide outlines the leading tools for serving open-source AI models on Kubernetes, covering how teams can deploy and scale model inference in containerised environments. It examines the trade-offs between different serving frameworks and orchestration setups, reflecting growing interest in running large language models on self-managed infrastructure rather than proprietary cloud services.

Google's push to expand artificial intelligence is colliding with a fundamental constraint: electricity. Data centres powering AI models require enormous amounts of energy, and the US power grid is struggling to keep up with demand, raising questions about how quickly Google and other tech firms can scale their AI infrastructure.

China's exports surged in September, growing 36.3 percent year on year, the biggest leap since the COVID-19 pandemic. The jump is being attributed largely to booming demand linked to artificial intelligence, and the figures drew immediate attention from economists and commentators tracking the global AI build-out and China's trade position.

Federal agencies are moving into a new phase of artificial intelligence adoption, one that analysts and officials say could bring a very different budget reality. As AI shifts from small pilots to large-scale deployments, agencies face growing costs for infrastructure, licensing, and workforce changes that early experimental budgets did not account for. Commentators are weighing how Congress and agency leaders will fund sustained AI operations at scale.

Nebius has added NVIDIA Run:ai to its AI Cloud Applications catalog, making the open-source scheduling and orchestration layer available for customers running GPU workloads. The integration lets teams manage and optimize AI training and inference clusters directly through Nebius's cloud platform. The announcement points to tightening cooperation between Nebius and NVIDIA as demand for AI infrastructure keeps climbing.

Nscale, the AI cloud infrastructure company, is being discussed in connection with energy and power, reflecting growing attention to the electricity demands of large-scale data centres used for artificial intelligence workloads. Observers are weighing how much power Nscale's operations require and what that means for its expansion plans, though no specific announcement has been confirmed.

Microsoft has announced plans to scale its custom Azure Maia AI accelerator chips to multi-gigawatt capacities, marking a major shift in its cloud infrastructure strategy. The move reduces the company's heavy reliance on Nvidia's GPUs for AI workloads, and observers see it as part of the broader race among hyperscalers to build in-house silicon for AI datacentres.

University of Pittsburgh researchers, in a study published in May, concluded that workplace productivity did not rise after companies invested in AI and cut staff. The findings challenge the overwhelmingly optimistic outlook executives have expressed about AI's potential for years, with the researchers noting that management optimism does not match measured outcomes. The study is drawing fresh attention as firms continue large-scale AI spending and layoffs.