A 1929-Style Crisis in 2026?

New York Times journalist Andrew Ross Sorkin draws parallels between the bubble of the 1920s and today’s AI moment.

mission [EXP+PROS]

New York Times journalist Andrew Ross Sorkin sees several points of convergence between the financial bubble of the 1920s and the current moment around artificial intelligence, though he avoids making direct comparisons. In his new book, 1929: Inside the Greatest Crash in Wall Street History, he prefers to focus on how people reacted to that period and leave the conclusions to readers.

After publishing the bestseller Too Big to Fail, about the 2008 financial crisis, Sorkin was often asked about comparisons with 1929 and realized he did not have a satisfactory answer. That prompted him to study the period more deeply, especially the human decisions that helped shape the collapse.

The 1920s were a time of explosive innovation, with the rapid expansion of radio and the automobile. Between 1928 and 1929, the stock market rose by about 90%. The dominant sentiment resembled today’s FOMO (Fear of Missing Out): everyone wanted to take part in the boom and get rich quickly.

The stock market became part of popular culture. Financial titans began appearing on magazine covers, and the stock exchange became a daily topic of conversation. “John Raskob, a GM executive, was the Elon Musk of that era,” Sorkin notes.

Parallels with AI

Sorkin sees similarities between the technological enthusiasm of the 1920s and the current boom in artificial intelligence. The difference, he argues, is that today enthusiasm is already mixed with fear about the economic consequences, whereas in the 1920s optimism largely dominated.

He also points to a possible mismatch between the hundreds of billions of dollars invested in AI and the sector’s short-term revenue potential. At some point, he suggests, “the math may stop adding up.”

Professional investors often follow herd behavior because market incentives reward those who stay with the trend, the journalist explains. Being wrong together with everyone else tends to be less risky for one’s career than exiting too early.

In his own daily work, Sorkin says he uses AI for administrative tasks and research, but not for writing.

While he believes the technology can accelerate intellectual production, he raises a question: if books can be produced much faster, will people still read them—or will they simply ask their AI agents to summarize them?