Macro FAQ

On June 12, 2026, a US export-control directive required Anthropic to suspend Fable 5 and Mythos 5 for foreign nationals. Anthropic could not check nationality in real time, so it disabled the models worldwide. Access remained off through June 21. I treat this as the first clear use of an AI 'kill switch'.

A US export-control order required the suspension after the government cited national security and a claimed jailbreak. A jailbreak uses prompts to bypass a model's safeguards. Anthropic disputed the basis. It called the flaw a 'narrow, non-universal jailbreak' available from other models. The company complied because it could not screen users by nationality in real time.

The January 2025 US rule divided countries into three tiers for access to AI computing power. It placed European Union members in two different tiers. The government rescinded it in May 2025. The official notice said the rule had 'downgraded' allies to second-tier status. Allied access was therefore a discretionary privilege.

On living standards, mostly no. Paul Krugman argued that much of the measured productivity decline came from price-index methods. A price index tracks changes in prices over time. The median living-standard gap remained roughly stable. The strategic problem is different: Europe uses frontier technology that it neither builds nor controls.

It means Europe can control essential chips, cloud services, AI models, data, and infrastructure without revocable dependence on foreign providers. The June 2026 European Union proposals included Chips Act 2.0 and the Cloud and AI Development Act.

ASML is the only maker of extreme-ultraviolet (EUV) lithography systems. Chipmakers use these machines to produce leading-edge chips. ASML is therefore Europe's strongest technology chokepoint. Yet Europe still depends on foreign graphics processors, design software, cloud services, and frontier models. One chokepoint does not create autonomy.

Henry Farrell and Abraham Newman use this term for power created by control of economic networks. When an economy relies on a few hubs, a hub owner can cut off downstream users. They call this the chokepoint effect. The US-controlled AI stack of chips, cloud services, and models fits that pattern.

Partly. Open weights are model parameters that users can download and run. They are difficult to embargo once distributed. However, cloud services, application programming interfaces, and advanced chips cannot be downloaded with the model. These dependencies preserve the chokepoint.

The economy feels bad because people experience price levels, not rates of change. Inflation fell from 9% to 2.4%, but cumulative prices rose roughly 25% since 2020, with groceries up 29.4% and housing affordability at its lowest since the 1980s. Economists celebrate the rate normalizing; consumers live with the permanent level shift. This levels-vs-rates disconnect is the structural explanation for the vibecession.

The levels-vs-rates problem describes a disconnect between how economists measure inflation (year-over-year rate of change) and how consumers experience it (cumulative price level). Inflation falling to 2.4% means prices are rising slowly again. It does not mean the 25% cumulative increase since 2020 reverses. A grocery bill that cost $150 in 2020 costs $194 today and will never cost $150 again.

On February 4, 2026, Stewart hosted Nobel laureate Thaler on 'The Weekly Show' to discuss behavioral economics. In the conversation Stewart argued, in his own framing, that economics functions to maximise shareholder value and characterised nudge theory as inadequate for systemic problems. He then sketched a market-based mechanism for carbon mitigation that, as several economists noted publicly afterward, resembled the carbon tax Thaler had outlined moments earlier. The episode generated public commentary from economists including Jason Furman and Jerusalem Demsas.

Vibepression is a term coined by Charles Schwab's Kevin Gordon in December 2025 to describe the deepening of the vibecession, a concept Kyla Scanlon introduced in June 2022. As of February 2026, the University of Michigan consumer sentiment index sits at 57.3, the 3rd percentile of its historical range, despite GDP growth of 4.4% and unemployment of 4.3%. The vibecession never resolved; it got a bleaker name.

In aggregate, yes, since June 2023. But the distribution is K-shaped: high earners held 4.5% wage growth while the bottom quartile fell from 7.5% to roughly 3.5%. The Economic Policy Institute reported in February 2026 that low-wage workers' real wages actually declined in 2025, reversing pandemic-era compression that had closed up to one-third of the post-1979 wage gap.

No. Prices will not return to pre-pandemic levels. Cumulative CPI is up roughly 25% since early 2020, with food-at-home up 29.4% and housing costs up 30-45%. Reversing this would require sustained deflation, which central banks actively prevent because falling price levels cause recessions. The inflation rate has normalized at 2.4%, but the level shift is permanent.

Less effective than widely believed. A meta-analysis by Maier et al. found that after correcting for publication bias, real-world nudges increase desired behavior by just 1.4 percentage points, compared to 8.7 in lab settings. A 2025 second-order meta-analysis by Hu found nudge effects drop to near zero (d=0.004) after full bias correction. Specific applications like pension auto-enrollment show larger effects, but the average impact is far smaller than proponents claim.

The Wero payment system is a digital wallet built by the European Payments Initiative (EPI), a consortium of 16 major European banks. It uses Single Euro Payments Area (SEPA) Instant Credit Transfer infrastructure. Wero moves money directly between bank accounts in under 10 seconds. Users identify the recipient with a phone number, email address, or QR code. The payment bypasses card networks. By February 2026, EPI had committed roughly €500 million in capital and had over 1,100 member institutions.

EPI, Bancomat, Bizum, SIBS, and Vipps MobilePay signed a Memorandum of Understanding on February 2, 2026, formalizing the EuroPA alliance. Its hub connects Wero with Spain's Bizum, which had 30.6 million users at the time of publication. It also connects Italy's Bancomat, Portugal's SIBS, and Nordic wallet Vipps MobilePay, which had 12.5 million users. The alliance created a network of 130 million users across 13 countries overnight. It covers 72% of the EU population and could make merchant support more likely.

A Visa or Mastercard transaction can cost a European merchant up to 2%. That price includes interchange, scheme, and processing fees. Interchange is the fee that the merchant's acquiring bank pays to the cardholder's issuing bank. Scheme fees are card-network charges, while processing fees pay for transaction handling. Wero's proposed German price is 0.77% plus fees charged by the merchant's checkout gateway. The difference is about 100 to 120 basis points per transaction. One basis point is 0.01 percentage points. This cost advantage exists because account-to-account (A2A) payments bypass the card network.

The 2015 Interchange Fee Regulation capped consumer debit interchange at 0.2% and credit interchange at 0.3%. Visa and Mastercard responded by increasing unregulated scheme fees by 33.9% between 2018 and 2022. The average net merchant service charge nearly doubled from 0.27% to 0.44%, which neutralized the regulatory benefit. The cap also reduced the revenue available to fund new payment networks, which made entry harder for competitors.

India's Unified Payments Interface (UPI) processed 228.3 billion transactions worth $3.6 trillion in 2025. Brazil's Pix reached 175 million users and processed $4.6 trillion in 2024. Both reached high transaction volumes within a few years. India had low card penetration, so UPI filled a vacuum instead of displacing incumbents. Brazil's central bank required institutions to participate in Pix. Europe has high card penetration, established consumer habits, and 27 regulatory jurisdictions. Those differences make direct comparison misleading.