OpenNashPodcasts

Topic

Larry Summers

Larry Summers is an economist whose views on artificial intelligence, the economy and his own career are collected here from podcast interviews, mainly an NEJM AI Grand Rounds conversation recorded in 2024. Other speakers, including Sam Altman, Bret Taylor and Tim Geithner, also discuss him.

On that show, Summers said AI differs from earlier technologies such as electricity because it can improve itself. He also described joining the OpenAI board after the 2023 leadership change and recalled meeting Jeff Bezos at the US Treasury in 1998. Geithner told Odd Lots that Summers believed every policy was short of the frontier of knowledge.

26 facts from 5 episodes, in date order. These are what speakers said on the shows, each linked to the moment it was said. They aren't independently verified.

1998

Larry Summers recalls that he and Treasury Secretary Robert Rubin remarked after the Bezos lunch that mall owners should watch out.

“And I remember then Secretary Rubin and I saying to each other that the owners of malls had better watch out.”

Larry Summers · The Economics of AI: A Conversation with Larry Summers, 0:08:08

2023

Larry Summers says he was approached by those forming OpenAI's new governance arrangements to join the board after the leadership transition.

“I was approached by the people who were involved in forming new arrangements to ask whether I would take on a position on the board.”

Larry Summers · The Economics of AI: A Conversation with Larry Summers, 0:23:43

2023

Summers says the new OpenAI board commissioned a review in which millions of dollars were spent, tens of thousands of documents were reviewed and a major law firm questioned witnesses.

“Millions of dollars were spent. Tens of thousands of documents were reviewed. Witnesses were questioned for many hours by a major law firm that had substantial experience with this kind of thing.”

Larry Summers · The Economics of AI: A Conversation with Larry Summers, 0:25:38

2024

Summers says the best current guess is that diminishing returns from scaling arrive somewhat faster than expected, offset by larger gains from longer-chain, more human-like reasoning.

“diminishing returns is likely to come somewhat faster from scaling, but that modifications of the technology to make possible longer change of reasoning and more human-like reasoning processes”

Larry Summers · The Economics of AI: A Conversation with Larry Summers, 0:13:40

21st century

Larry Summers says it is an even-money chance that historians will treat Russia-Ukraine, Donald Trump and Xi Jinping as secondary to the rise of human-like machine intelligence.

“I think that it's an even-money chance that when historians write about the second-fifth of the 21st century, Russia and Ukraine and Donald Trump and Xi Jinping will be secondary stories”

Larry Summers · The Economics of AI: A Conversation with Larry Summers, 0:00:00

Undated

Summers says the insight about malls and online retail did not become financially important until 15 to 20 years later.

“But that wouldn't have been a financially important legitimate insight for another 15 or 20 years after we had it.”

Larry Summers · The Economics of AI: A Conversation with Larry Summers, 0:08:08

Undated

Summers recalls hearing about a decade earlier that automated vehicles had driven from San Jose to San Francisco along Route 101 with no driver.

“I remember hearing now a decade ago about these automated vehicles that had driven up Route 101 from San Jose to San Francisco with no driver.”

Larry Summers · The Economics of AI: A Conversation with Larry Summers, 0:08:08

Undated

Larry Summers argues that AI differs from earlier general-purpose technologies such as electricity because it can improve itself, whereas electricity did not generate more electricity.

“Electricity was great, but electricity didn't self-generate more electricity.”

Larry Summers · The Economics of AI: A Conversation with Larry Summers, 0:10:30

Undated

Summers notes that the first moving picture cameras were placed at the back of theaters to record plays before better uses were found.

“The first cameras, moving picture cameras, were put at the back of theaters to record plays.”

Larry Summers · The Economics of AI: A Conversation with Larry Summers, 0:14:59

Undated

Larry Summers argues that obesity rose even though the upper 75 percent of Americans could already afford enough calories 50 years ago, so he does not blame calorie abundance.

“I think if you look at the upper 75 % of the American population, there was no substantial difficulty in affording an adequate caloric intake 50 years ago, and yet levels of obesity have risen.”

Larry Summers · The Economics of AI: A Conversation with Larry Summers, 0:19:12

Undated

Larry Summers says the film The Social Network's portrayal of his meeting with the Winklevoss twins is not literally accurate but conveys the nature of the meeting.

“It is not literally accurate, but it conveys the nature of that meeting.”

Larry Summers · The Economics of AI: A Conversation with Larry Summers, 0:27:58

Undated

Larry Summers says being president of Harvard University is harder than being US Treasury Secretary because of more politics and the extreme decentralization of universities.

“President of Harvard University, it's got a lot more politics in it than working at the treasury in Washington, given the extreme decentralization of Harvard and of universities in general.”

Larry Summers · The Economics of AI: A Conversation with Larry Summers, 0:29:56

Undated

Larry Summers says he would choose economist John Maynard Keynes as his dinner companion because Keynes embodied a practical, intellectual, economics-oriented life.

“Probably John Maynard Keynes, because he embodied the practical, intellectual, economics-oriented life that I have tried to lead.”

Larry Summers · The Economics of AI: A Conversation with Larry Summers, 0:30:20

Undated

Larry Summers explains that the efficient market hypothesis holds that future speculative prices cannot be predicted, because predictable moves would already be priced in.

“the efficient market hypothesis, essentially has as its central idea that you can't predict the evolution of a future speculative price, because if you could predict it, it already would have moved.”

Larry Summers · The Economics of AI: A Conversation with Larry Summers, 0:32:01

Undated

Larry Summers says AI and digital technology promote economies of scale and non-convexities, which make economic analysis harder and may weaken the pure invisible hand.

“AI and digital technology more generally promote economies of scale. They promote what economists call non-convexities.”

Larry Summers · The Economics of AI: A Conversation with Larry Summers, 0:32:41

Undated

Tim Geithner attributes to Larry Summers the idea that a financial system should be safe for failure rather than aim to make failure inconceivable.

“The test of a system is, I think Larry Summers said, it should be safe for failure.”

Tim Geithner · Tim Geithner on How to Fight the Next Financial Crisis, 0:25:01

Undated

Tim Geithner says Larry Summers, when he arrived to run the Treasury's international side, believed every policy and practice was short of the frontier of knowledge.

“he came in and he had this deep conviction that anything we were doing, any type of policy, any type of practice was short of the frontier of knowledge.”

Tim Geithner · Tim Geithner on How to Fight the Next Financial Crisis, 0:39:42

Episodes