Good morning

Your high-signal briefing for today. The most consequential items come first.

1. EXCLUSIVE: Anthropic quietly sets up biology lab as it ramps AI drug…

Reuters reports that Anthropic has quietly established an internal biology laboratory to support an expanding AI-driven drug discovery program. The move signals a strategic shift from pure AI development toward direct biotech application, raising questions about dual-use risks alongside commercial ambitions. It follows public statements by CEO Dario Amodei about AI's potential to compress decades of biomedical progress into a few years.

Why it matters: A leading AI safety company entering wet-lab biology directly reshapes the competitive and regulatory landscape for AI in medicine.

Reuters · Read original →

2. Virtual biotech company puts thousands of AI scientist agents to work…

A Stanford Medicine-affiliated virtual biotech has deployed thousands of autonomous AI scientist agents running in parallel to accelerate drug discovery workflows. The approach attempts to replace or augment human researchers at scale across hypothesis generation, experimental design, and analysis. The project represents one of the most ambitious deployments of agentic AI in a scientific setting to date.

Why it matters: Mass deployment of agentic AI in drug discovery could fundamentally alter research productivity benchmarks and the economics of pharmaceutical R&D.

Stanford Medicine · Read original →

3. Artificial intelligence now beats some of the best human forecasters

The Economist reports that large language model-based systems now outperform some of the best human superforecasters on structured prediction tasks. The finding emerges from competitive forecasting benchmarks where AI models were evaluated against elite human teams. The result challenges assumptions about where human judgment retains comparative advantage over AI.

Why it matters: Demonstrating AI superiority in calibrated probabilistic forecasting has direct implications for economic modeling, risk assessment, and decision-making research.

The Economist · Read original →

4. Brad Setser (@Brad_Setser)

Brad Setser argues that Japan's reflationary experiment is approaching its limits, with rising yields threatening debt sustainability and…

6:18 PM, September 17 · View original → · Profile on X

5. Japan raises interest rate to new 31-year high to curb rising prices

Japan's central bank raised its benchmark interest rate to its highest level in 31 years, citing persistent inflationary pressure driven partly by elevated energy costs. The decision continues the Bank of Japan's gradual exit from its decades-long ultra-loose monetary policy. Global markets are watching closely given Japan's role as a major source of carry-trade financing.

Why it matters: A sustained shift in Japanese monetary policy has significant spillover effects on global capital flows, bond markets, and currency dynamics.

BBC · Read original →

6. AI Governance Needs a Macroprudential Turn - promarket.org

A ProMarket essay argues that AI governance should borrow the macroprudential framework used in financial regulation, focusing on systemic risk rather than firm-level compliance. The author contends that current AI policy treats risks as isolated and firm-specific, missing interconnected failure modes analogous to those that caused the 2008 financial crisis. The piece calls for stress-testing regimes and concentration limits applied across the AI sector.

Why it matters: Applying macroprudential logic to AI regulation is a conceptually novel framing with practical policy implications relevant to both finance and technology governance.

promarket.org · Read original →

7. Researchers have created a mouse with a partly human brain

The Economist reports that researchers have engineered a mouse carrying a partially human brain, incorporating human neural cells to create a chimeric model. The organism is intended to provide a more accurate biological substrate for studying neurological diseases and testing drug candidates. The advance raises both scientific promise and bioethical questions about the boundaries of human-animal chimeras.

Why it matters: Chimeric brain models could substantially improve the translational validity of preclinical neuroscience and drug development research.

The Economist · Read original →

An expert-level generalist AI for abdominal CT diagnosis

A study published in Science presents a generalist AI system that achieves expert-level diagnostic performance across a broad range of conditions on abdominal CT scans, evaluated against specialist radiologist benchmarks.

Why it matters: Expert-level generalist diagnostic AI validated in a high-impact peer-reviewed journal marks a meaningful clinical milestone with direct implications for radiology practice and healthcare economics.

Science · Read original →

Ranked: Where GDP Per Capita Grew Fastest Since 1990

Guyana’s real GDP per capita has surged 1,549% since 1990. See which economies grew fastest, adjusted for inflation and purchasing power.

Source: Visual Capitalist · View source · Found September 18, 2026

Source notes: NEJM: HTTPError; arXiv: HTTPError