U.S. — Robert Shiller, a Nobel laureate and economics professor at Yale University, published a guest essay on June 22, 2026, titled "This Doommaxxing Has Got to Stop." The essay argued that negative expectations about AI could contribute to their own realization.

Shiller stated, "When millions of people make millions and millions of decisions based upon negative expectations, there is a risk that fear can actually help birth the reality." He noted a recent Quinnipiac poll from March 2026, where 70% of Americans anticipated AI would lead to fewer available jobs. This percentage increased from 56% in a Quinnipiac poll conducted in March 2025.

Further survey data indicated that 40% of Americans believed AI would negatively impact society over the next two decades, according to a Pew survey. Only 16% of respondents in the same survey believed AI would have a positive societal impact during that period. Despite these concerns, the Yale Budget Lab reported no change in the occupational mix for jobs most exposed to AI since ChatGPT's launch in late 2022.

Shiller referenced historical instances of job displacement fears, such as the 1957–58 economic downturn, which journalists termed the "Automation Recession." He also mentioned initial predictions from figures like Anthropic CEO Dario Amodei, who suggested AI could eliminate half of entry-level white-collar jobs within five years, and Microsoft AI chief Mustafa Suleyman, who estimated most white-collar automation could occur within 12 to 18 months. Both Amodei and Suleyman have since modified their timelines regarding AI job displacement.

Shiller observed that government intervention has limits in addressing these narratives. He shared, "There's only so much Washington can do about these narratives. And, suffice to say, Donald Trump is no Franklin Roosevelt." He referred to President Franklin Roosevelt's 1935 fireside chat, which a recent paper found measurably increased spending in cities with higher radio exposure.

Why It Matters

Shiller's essay addresses concerns about the potential economic and social impacts of artificial intelligence. The piece notes growing public anxiety regarding AI's effect on employment, as indicated by recent polls showing a large percentage of Americans expect job losses. Shiller's argument suggests a mechanism by which these widespread negative expectations could influence actual economic outcomes, drawing parallels to historical events where public sentiment played a role in economic shifts, such as the 1929 stock market crash when approximately 2% of American households owned stock.