What Would John Maynard Keynes Say About A. I.? The Keynes Bot Told Us. To understand how A.

I. will affect our future, we’re looking to the past. Nearly a century ago, the legendary economist John Maynard Keynes predicted that technology would significantly increase productivity and reshape how we live. Today, as A.

I. capabilities accelerate, technologists have similar claims. But what would Keynes say about this moment? The Princeton University economist Markus Brunnermeier set out to answer this very question. Using an A.

I. agent trained on more than 12,000 pages of Keynes’s writings and transcripts, along with the few available recordings of him, Brunnermeier and his team built a bot designed to answer questions in Keynes’s voice and style. He then had an in-depth conversation with the Keynes bot about the future of work and what happens to society when technology changes what people do for a living. Times Opinion, in collaboration with Brunnermeier, distilled those conversations into a short exchange.

The result is a revealing conversation with a Keynes who never lived to see the A. I. age — and a trenchant warning from his A. I. bot._(The New York Times has sued Open AI and Microsoft, claiming copyright infringement.

The companies have denied those claims.)_How the Keynes bot Was Created The economist Markus Brunnermeier and the researcher and Ph. D. student Pablo Balsinde built a bot using Chat GPT to create the Keynes A. I.

The bot was instructed to draw primarily from Keynes’s seminal essay “Economic Possibilities for Our Grandchildren” (though the model was also provided with his larger body of work) and to answer questions in his style, which the pair described as “first-person, early 20th century British cadence, restrained wit.”According to the team behind the bot, its initial responses were often long-winded and unclear. To improve them, the researchers “asked the GPT to rephrase, prompted it to reply using famous quotes from Keynes and cut parts of responses.” While the language was generated by the model, the responses were edited by the research team for clarity and brevity. For the visuals, the team uploaded a public domain image of Keynes and used Deep Seek to alter the environment, placing him behind a desk.

To create audio resembling Keynes’s voice, the team used a recording of Keynes from the BBC as a reference._Below is an edited transcript of this video. We recommend watching it in its original form for the full effect. You can do so using the video player above.