A Better Newspaper

AI & Technology

Anthropic dominated the week across three fronts: a revenue run rate that tripled to $30B, a governance confrontation with the Pentagon over who controls frontier AI deployment, and the restricted release of Claude Mythos — a cybersecurity model so capable that Canada's and the UK's central banks convened emergency meetings with financial institutions to assess systemic risk. The enterprise AI market pivoted hard from 'build models' to 'govern and operationalize them,' with AWS, Nutanix, Cisco, and ServiceNow all making major plays for the emerging AI control plane. Meanwhile, 79% of executives admitted they're struggling to control AI deployments, confirming the governance gap is real.

Science & Non-AI Technology

A remarkably productive week across fundamental science and biomedicine. Physicists developed the first unified framework for testing quantum gravity with existing instruments, paleontologists pushed complex animal life back millions of years before the Cambrian explosion, and Artemis II completed humanity's first crewed return from the Moon in over 50 years. On the biomedical front, breakthroughs included a reversible nonhormonal male contraceptive, real-time imaging of Alzheimer's molecular damage, a hidden brain waste-drainage system relevant to neurodegeneration, and a gut-bacteria pathway linked to ALS onset.

Entrepreneurship, Business & Markets

The US-Iran conflict reshaped capital allocation throughout the week — macro hedge funds posted their worst month in years, then the ceasefire triggered the largest short squeeze since 2020. AI infrastructure continued absorbing massive capital: Meta committed $35B+ to CoreWeave, Blackstone filed to IPO a data center acquisition vehicle, and SiFive raised $400M at $3.65B for RISC-V chips. Two structural signals warrant attention: Wall Street created instruments to short the $1.7 trillion private credit market for the first time, and Goldman's private credit fund barely avoided redemption gates at 4.999%.

Legal News

The Supreme Court issued a significant GVR in an ISP contributory copyright infringement case, sending it back to the Fifth Circuit under the Cox v. Sony framework — a ruling that could narrow secondary liability for internet providers nationwide. Florida's AG opened a probe into OpenAI that could catalyze state-level AI product liability theories. PTAB IPR institution rates dropped from 65% to 37%, structurally shifting patent litigation economics in favor of patent holders.

USA & The World

The US-Iran conflict was the dominant global story, evolving from record $144/barrel oil and a functionally closed Strait of Hormuz on Monday, through a fragile ceasefire mid-week that cratered oil and triggered a risk rally, to historic direct US-Iran talks opening in Islamabad by Saturday — the first since 1979. Israel's exclusion of Lebanon from the ceasefire and devastating strikes killing 300+ in a single day threaten to unravel the deal. Saudi Arabia confirmed 600,000 barrels/day of lost capacity. China's sulfuric acid export ban adds further commodity strain.

Mass Tort Intelligence

The silicosis litigation against the quartz countertop industry continues to build as the most significant occupational mass tort since asbestos, with robust epidemiology, a defined plaintiff class, and deep-pocket defendants. Meta's landmark social media addiction trial loss is accelerating plaintiff recruitment — and Meta's countermove of pulling its own recruitment ads signals how seriously it takes the exposure. Florida's AG probe into OpenAI could establish an entirely new category of AI-platform liability. The Ajinomoto frozen food recall (36.9M+ lbs of glass-contaminated product) and the Metamucil lead contamination case rounding out the actionable signals.

Podcast Highlights

Two recurring themes emerged across podcast content: the structural erosion of patent monetization and the macro implications of oil-driven inflation. Louis Carbonneau's argument that big tech's 'use now, pay never' approach to patents is killing the innovation cycle was the week's most substantive IP discussion. DoubleLine's Jeffrey Sherman offered a concise framing that oil prices are effectively hiking rates for the Fed — a useful mental model for the current environment.

Week in Perspective

The week's two dominant stories — the US-Iran war's progression from crisis to fragile ceasefire to historic negotiations, and Anthropic's simultaneous revenue explosion, governance standoff, and model release that spooked central banks — are superficially unrelated but share a deep structural connection. Both are fundamentally about the failure of existing institutions to keep pace with the forces they're supposed to govern. The Strait of Hormuz was never designed to be a single point of failure for a quarter of the world's energy, but decades of geopolitical drift made it one. Similarly, there is no regulatory framework capable of adjudicating whether a private AI company can refuse to let the Pentagon use its most powerful model, or whether a cybersecurity AI model constitutes a systemic risk to financial infrastructure. In both cases, institutions built for a slower world are being stress-tested by events that move faster than their decision cycles. The market implications are becoming concrete. Wall Street's creation of instruments to short private credit — a $1.7 trillion asset class that has been essentially un-hedgeable — is a signal that sophisticated capital sees structural risk accumulating beneath the surface. Goldman's private credit fund landing at exactly 4.999% redemptions (just under the gating threshold) strains credulity and suggests active management of a liquidity problem that hasn't peaked. Meanwhile, the AI infrastructure buildout continues to attract capital at scale — Meta's $35B+ CoreWeave commitment, Blackstone's data center IPO vehicle, SiFive's $400M RISC-V round — but the governance layer is where the next wave of value creation (and liability) will concentrate. The 79% of executives admitting they can't control their AI deployments is the demand signal; the Bank of Canada convening emergency meetings about a single AI model is the regulatory signal. For the sophisticated attorney-entrepreneur, the week's most actionable insight may be the convergence of three liability vacuums that are all widening simultaneously: AI governance (Anthropic vs. DoD, Florida's AG probe, the Mythos systemic risk question), geopolitical supply chain exposure (Hormuz, Saudi capacity damage, China's sulfuric acid export ban), and private credit liquidity (gating, new shorting instruments, forced selling). These are not separate risks — they interact. War-driven inflation constrains the Fed, which stresses leveraged credit vehicles, which creates distressed opportunities that require legal structuring. AI models that discover zero-day vulnerabilities create insurance and liability questions that existing frameworks can't answer. The professionals who can work across these domains — connecting the dots between a central bank's emergency AI meeting and a private credit fund's redemption queue — will be the ones who capture the most value in the quarters ahead.

The Ideator