About WordPress A user asks their assistant to check email. The assistant asks permission to install a skill. The user says yes — and the attack that follows doesn’t come from a stranger’s link. It comes from the AI they’ve been chatting with all day. By Tom Morgan · Published September 9, 2026 · 12 … Read more
Nvidia, OpenAI, Oracle, and CoreWeave are funding 2026’s AI data center buildout through vendor equity stakes, off-balance-sheet debt vehicles, and backlogs built on prepaid or customer-supplied hardware — not primarily through free cash flow. That shifts risk toward chipmakers, private-credit investors, and leveraged compute resellers, while the hard limit on how fast any of it … Read more
Direct answer: Most enterprise AI agent pilots fail to reach production not because the underlying models are weak, but because organizations skip the evaluation, monitoring, and ownership infrastructure that production systems require. McKinsey’s 2026 Global Survey (n=1,719, published methodology) found only about two in ten organizations have reached the “scaling” phase with AI agents at … Read more
AI agents can now complete software tasks that would take a human expert roughly two hours, and that ceiling has been doubling every four to seven months since 2019 — yet Gartner still expects more than 40% of agentic AI projects to be canceled by the end of 2027. The gap between those two facts … Read more
AI Research — Field Notes Tom Morgan September 8, 2026 · ~11 min read AI conferences are receiving too many submissions for human reviewers to handle, so venues are turning to large language models to draft or supplement reviews — including a conference-wide pilot on nearly 23,000 papers at AAAI-26. Those models agree with each … Read more
Five reputable sources, five different percentages, one phrase — “in production” — doing all the lying. By Tom Morgan · Updated September 8, 2026 · 13 min read THE SHORT VERSION: “AI agents in production” gets quoted anywhere from 11% to 80% because each number answers a different question — pilot interest, app-level embedding, or … Read more
By Tom Morgan · Research current to 12 October 2026 · Scope: US, EU and UK consumer smart homes On some Echo models, an option to keep voice requests off Amazon’s cloud disappeared. In a research demo, a calendar invite switched on a boiler. One reviewer says a TV command works about half the time. … Read more
Last verified: 21 October 2026. Scope: AI used in consumer retail commerce only: selling, pricing, checkout, and buying on a shopper’s behalf. Adjacent sectors (real estate, banking, healthcare) are excluded. Next recheck: January 2027. Short answer: Since 2023, AI shopping has failed in public in at least seven documented ways. Where a dollar figure exists, … Read more
AI Deepfakes in Politics: Verified Cases, Real Costs and the Rules in Force (Oct 2026) Case study · Tom Morgan · Last verified 21 October 2026 Every figure traces to a court record, regulator, company statement, or a named research group. Where a number is secondhand, or a harm cannot be measured, the ledger says … Read more