Uber’s president says AI spending is “harder to justify” after the company reportedly exhausted its annual AI budget just four months into 2026, questioning whether rising token consumption for Claude Code has translated into meaningful returns. The episode matters because it shifts the debate on AI from capability to accountability, and signals that even large deployers are pressing for tighter links between agentic tooling, measurable productivity, and cost control. Source
AI warfare is no longer a theoretical problem at the UN: the Convention on Certain Conventional Weapons, focused on lethal autonomous systems, is hosted twice a year at the United Nations in Geneva, underscoring how quickly “killer robots” have moved from speculation to policy agenda. The relevance for global politics is straightforward—once autonomy enters the kill chain, red lines, verification, and command-and-control rules become harder to enforce, raising the stakes for arms-race dynamics even before new laws are agreed. Source
The tech world’s “biggest transformation possibly ever” is being traced to Claude Code and OpenClaw, which the story says kicked off computing’s most destabilising wave of agentic automation and helped plunge development into chaos. This matters because agentic systems change not only what software does but how it is built—by letting tools take initiative, coordinate work, and trigger actions at scale—so failures propagate differently from conventional model outputs. Source
US law enforcement warns of “Anti-Tech Extremism” as AI hatred grows, with documents obtained by WIRED describing a new category of threat tied to Americans’ anger over job-stealing AI and the siting of data centers. The political implication is that AI economics can translate into physical risk: as communities contest infrastructure and employment impacts, radicalisation can attach itself to technology targets, forcing a security response that may in turn shape public acceptance of AI deployment. Source
AI has not yet produced a clean story of mass unemployment, with aggregate employment in developed countries remaining broadly stable and assessments finding limited evidence that AI has shifted headline numbers. But the story flags a subtler labour-market deterioration: the quiet weakening of entry-level work, implying that AI’s impact may show up first in access to the job ladder rather than total job counts, with long-run consequences for skills formation and social mobility. Source