The pivot from payroll to compute is structural, not cyclical.


Bit Tech commits $725bn capex; simultaneously cuts headcount at scaleMeta, Amazon, Microsoft, and Alphabet have collectively signalled approximately $725 billion in capital expenditure for 2026 — up more than 75% year-over-year — almost entirely for data centres, custom chips, and AI infrastructure. Concurrently: Meta cutting 8,000 jobs in May, Amazon down ~30,000 roles, Microsoft offering voluntary buyouts to ~125,000 staff. • The pivot from payroll to compute is structural, not cyclical. • Source: Tech Startups, May 8 Continue reading The pivot from payroll to compute is structural, not cyclical.

Hantavirus – what you need to know


Prompt (Claude AI) Does hantavirus only occur in ships? Are there moreexamples in say buidlings with poor pest control Output No, hantavirus is not specific to ships at all. Ships are just one environment where it can occur. Hantavirus transmission happens wherever infected rodents (primarily deer mice, white-footed mice, and rats) are present and their droppings, urine, or nesting material become disturbed and aerosolized. Common settings where outbreaks have occurred: Rural cabins and vacation homes — closed up over winter, mice move in, people stir up dried droppings when reopening Barns and outbuildings — classic agricultural exposure Construction sites — disturbing rodent-infested spaces Warehouses … Continue reading Hantavirus – what you need to know

Risks Cannot Be Mitigated Without Proper Risk Definition Based on Factual Structure — e.g. Hormuz


Colin Henderson | Bankwatch.ca Historical research and analysis – Claude.ai (analysis posted separately) The Structural Thesis The liberal rules-based order masks but does not correct the natural hegemonic propensity of powerful states. Institutional frameworks — democratic checks, multilateral structures, constitutional constraints — act as the brake on that propensity. Remove the institutional framework through autocratic or highly concentrated leadership, and the hegemonic drive expresses itself without the liberal mask. This is not a policy choice. It is a structural shift. And standard risk frameworks are not built to model it. Hormuz is the current operational instance of that thesis in … Continue reading Risks Cannot Be Mitigated Without Proper Risk Definition Based on Factual Structure — e.g. Hormuz

Liberal Values in the Post-WW2 Institutional Architecture – research and analysis


Research and analysis with Claude AI. Backgrounder to Risk Mitigation post: TITLE: Risks cannot be mitigated without proper risk definition based on factual structure – e.g. Hormuz Liberal Values in the Post-WW2 Institutional Architecture Prompt: Diplomacy and Values since WW2. What are the liberal values used in construction of post WW2 institutional architecture including but not limited to,  NATO, world bank, Marshall plan Output The post-WW2 order was a deliberate construction, driven by a diagnosis of what caused the catastrophe: nationalism, autarky, power politics without rules, and the failure of collective security in the 1930s. The architects — primarily American … Continue reading Liberal Values in the Post-WW2 Institutional Architecture – research and analysis

OpenAI, Anthropic ramp up enterprise push


OpenAI and Anthropic are both partnering with private equity firms in a bid to deploy their AI products to more businesses.  OpenAI is forming a $10 billion venture, raising funding from investors including Brookfield and Bain Capital, Bloomberg reported. Anthropic on Monday announced a $1.5 billion joint venture with Wall Street firms that is expected to act as a consulting arm for Anthropic.  The rival AI startups are racing to win over more enterprise customers — 20% of US businesses have adopted AI tools, mostly to “supplement a small number of employee work tasks,” Goldman Sachs analysts wrote. Both companies are focusing on a push … Continue reading OpenAI, Anthropic ramp up enterprise push

AI in financial services: regulators 2 years behind banks


This post is targeted at both banks and regulators (OSFI in Canada ). More specifically Agentic AI is the risk frame here. Agentic AI is defined here in banking context, goes exponentially beyond automation. Take mortgage sourcing. Banks have multiple and defined process for mortgages that sources, adjudicates, processes and funds. Each step is calibrated and defined and understood by regulators and those mortgages go on to be bundled and sold as tranches with known, defined risks. Where Agentic AI takes over the agent determines the best and most effective process to follow and will establish improvements which could bundle … Continue reading AI in financial services: regulators 2 years behind banks

Top 20 Arms Sales / Country 2023/4


Source: STOCKHOLM INTERNATIONAL PEACE RESEARCH INSTITUTE The independent resource on global security ____________________________________________ South Korea, Russia, Germany show significant increases. When we get 2025 data this trend will show even more of a shift to Europe, Asia reflecting NATO fracturing, and countries that support all growth (Korea). Why it matters: This will create shifts in investment choices and NATO structure, so worldwide implications. This is largely driven by US focus on own self interests and increasingly Israel lobby (AIPAC) influence on foreign policy with intended and untended consequences this brings. It remains to be seen whether this structural shift holds up … Continue reading Top 20 Arms Sales / Country 2023/4

LLM inefficiency and what could be the better model than next-token prediction


A discussion I had with Anthropic Claude verbatim. (Bloomberg: circular support amongst small group of individual players and the inherent financial risk if current frontier models for AI cannot provide the expected value) Prompt: Re the financial risk referred to in this Bloomberg piece, is there any evidence of work on a better inference and reasoning model that doesn’t drive exponential requirements in data centre capacity to support the model. It seems the current model requires is highly inefficient requiring infinite capacity to support. # There’s substantial work underway — but with an important caveat that’s directly relevant to the … Continue reading LLM inefficiency and what could be the better model than next-token prediction

Orbital data centres: power constraint forces compute off-planet


Orbital data centres: power constraint forces compute off-planet Worth highlighting from today’s briefing. New & Emerging Orbital data centres: power constraint forces compute off-planetAI hyperscale demand has hit a hard ceiling: a single facility now requires 100–500 MW, and US grid capacity and permitting timelines (5–7 years) cannot keep pace. SpaceX filed FCC plans in January for up to one million data-centre satellites. Starcloud filed for 88,000. Google’s Project Suncatcher is developing radiation-hardened TPUs for orbital deployment. Nvidia’s GTC 2026 launched the Space-1 Vera Rubin Module for in-orbit AI compute; Starcloud has already trained an LLM in space on Nvidia … Continue reading Orbital data centres: power constraint forces compute off-planet

EU AI Office shut out of Mythos as UK AISI leads


Resultsense Industry News 17 April 2026 3 min readResultsense via POLITICO Europe EU AI Office locked out of Mythos as UK keeps edge  AI safety groups tell the European Commission its AI Office lacks Mythos access and the staff to evaluate it, while the UK AI Security Institute published technical analysis within a week.  TL;DR: The European Commission is not among the 40 organisations Anthropic has granted access to its Mythos model, a Commission spokesperson has confirmed. Eight AI safety groups have written to Brussels demanding that the EU AI Office’s safety unit quadruple to 160 staff by 2030. The UK’s AI Security Institute, by contrast, gained … Continue reading EU AI Office shut out of Mythos as UK AISI leads