/
© 2026 RiffOn. All rights reserved.

Get your free personalized podcast brief

We scan new podcasts and send you the top 5 insights daily.

  1. Moody's Talks - Inside Economics
  2. ENCORE AI Series: Anthropic's Peter McCory
ENCORE AI Series: Anthropic's Peter McCory

ENCORE AI Series: Anthropic's Peter McCory

Moody's Talks - Inside Economics · Sep 22, 2026

Anthropic's Peter McCory discusses AI's massive economic potential, forecasting a possible 1.8 ppt annual boost to labor productivity growth.

Q1's Weak 2% GDP Growth Masks a Strong 2.5% Rise in Private Demand

The headline Q1 GDP figure was disappointing. However, a key underlying metric, final demand from private consumption and investment, grew a robust 2.5%. This points to an ongoing investment boom that suggests more economic resilience than the top-line number indicates.

ENCORE AI Series: Anthropic's Peter McCory thumbnail

ENCORE AI Series: Anthropic's Peter McCory

Moody's Talks - Inside Economics·2 months ago

AI's Primary Productivity Gain Is Broadening Skill Sets, Not Just Accelerating Tasks

While AI makes existing tasks faster, its most significant impact, according to Anthropic's user data, is enabling employees to perform functions outside their core expertise. An economist, for example, can now build interactive data dashboards without coding knowledge, effectively broadening their role and capabilities.

ENCORE AI Series: Anthropic's Peter McCory thumbnail

ENCORE AI Series: Anthropic's Peter McCory

Moody's Talks - Inside Economics·2 months ago

Economic Downturns Could Trigger Rapid AI-Driven Restructuring and Job Displacement

The biggest threat of AI on jobs may not be during stable economic times. Historical precedent suggests businesses use downturns as an opportunity to restructure and adopt new technologies. A future recession could see firms rapidly automate cognitive work, amplifying the shock and prolonging the downturn.

ENCORE AI Series: Anthropic's Peter McCory thumbnail

ENCORE AI Series: Anthropic's Peter McCory

Moody's Talks - Inside Economics·2 months ago

Anthropic's Microdata Points to an AI Productivity Boost Double the Macro-Economic Consensus

Mainstream economists forecast AI will add ~0.8 percentage points to annual labor productivity. However, Anthropic's internal data, based on observed task-level time savings, suggests a potential boost of 1.8 percentage points. This indicates a significant gap between micro-level efficiencies and macro-level forecasts.

ENCORE AI Series: Anthropic's Peter McCory thumbnail

ENCORE AI Series: Anthropic's Peter McCory

Moody's Talks - Inside Economics·2 months ago

AI's First Labor Market Impact Is Weakened Hiring for Young Workers, Not Mass Layoffs

While aggregate unemployment shows no AI impact, a subtle trend is emerging. Anthropic's research finds suggestive evidence that hiring rates for younger workers in AI-exposed roles have weakened. This implies firms are using AI to augment existing teams, thus reducing the need to hire new junior talent.

ENCORE AI Series: Anthropic's Peter McCory thumbnail

ENCORE AI Series: Anthropic's Peter McCory

Moody's Talks - Inside Economics·2 months ago

Users With 6+ Months of AI Experience Shift to Using It as a Collaborative Partner

New users tend to treat AI as a simple tool to complete specific tasks. However, Anthropic's research shows that after six months, experienced users are more likely to engage with AI as a thought partner. This suggests that unlocking AI's full potential requires developing a more collaborative, iterative skill set over time.

ENCORE AI Series: Anthropic's Peter McCory thumbnail

ENCORE AI Series: Anthropic's Peter McCory

Moody's Talks - Inside Economics·2 months ago

Generative AI Creates a "Ship of Theseus" Problem for Expert Authorship

When an expert repeatedly uses an LLM to write in their specific style, and that output enters the public domain, the AI is effectively trained on its own previous outputs. This creates a feedback loop that raises a philosophical question: at what point does the expert's voice cease to be authentic and become a reflection of the AI?

ENCORE AI Series: Anthropic's Peter McCory thumbnail

ENCORE AI Series: Anthropic's Peter McCory

Moody's Talks - Inside Economics·2 months ago

Prompting an LLM to Critique Its Own Output Is a Powerful Way to Uncover Flaws

Instead of accepting an AI's initial output, a power-user technique is to ask it to "step back" and act as an external expert reviewing its own work. This prompt often causes the model to identify its own errors and logical flaws, leading to a much more accurate and refined final product.

ENCORE AI Series: Anthropic's Peter McCory thumbnail

ENCORE AI Series: Anthropic's Peter McCory

Moody's Talks - Inside Economics·2 months ago

AI Is Diffusing Through the Economy 5-10 Times Faster Than Previous Technologies

Unlike the internet or electricity, which required massive infrastructure buildouts, AI can be adopted almost instantly. Data shows late-adopting regions are catching up on AI usage 5 to 10 times faster than they did with past consequential technologies, suggesting a compressed timeline for economic impact.

ENCORE AI Series: Anthropic's Peter McCory thumbnail

ENCORE AI Series: Anthropic's Peter McCory

Moody's Talks - Inside Economics·2 months ago