From the <New York Times< bestselling author of <The New Great Depression< and <Currency Wars<, a telling prediction for how AI will endanger global economic markets and security

In November 2022, OpenAI released GPT-4 in a chatbot form to the public. In just two months, it claimed 100 million users&mdash;the fastest app to ever reach this benchmark. Since then, AI has become an all-consuming topic, popping up on the news, in ads, on your messenger apps, and in conversations with friends and family. But as AI becomes ubiquitous and grows at an ever-increasing pace, what does it mean for the financial markets?

In <MoneyGPT<, Wall Street veteran and former advisor to the Department of Defense James Rickards paints a comprehensive picture of the danger AI poses to the global financial order, and the insidious ways in which AI will threaten national security. Rickards shows how, while AI is touted to increase efficiency and lower costs, its global implementation in the financial world will actually cause chaos, as selling begets selling and bank runs happen at lightning speed. AI further benefits malicious actors, Rickards argues, because without human empathy or instinct to intervene, threats like total nuclear war that once felt extreme are now more likely. And throughout all this, we must remain vigilant on the question of whose values will be promoted in the age of AI. As Rickards predicts, these systems will fail when we rely on them the most.

<MoneyGPT< shows that the danger is not that AI will malfunction, but that it will function <exactly as intended<. The peril is not in the algorithms, but in ourselves. And it&rsquo;s up to us to intervene with old-fashioned human logic and common sense before it&rsquo;s too late.

Autorentext

James Rickards is the Editor of Strategic Intelligence, a financial newsletter, and the bestselling author of The New Great Depression, Aftermath, The Road to Ruin, and many more. He is an investment advisor, lawyer, inventor, and economist, and has held senior positions at Citibank, Long-Term Capital Management, and Caxton Associates.

Klappentext

From the New York Times bestselling author of The New Great Depression and Currency Wars, a telling prediction for how AI will endanger global economic markets and security

In November 2022, OpenAI released GPT-4 in a chatbot form to the public. In just two months, it claimed 100 million users—the fastest app to ever reach this benchmark. Since then, AI has become an all-consuming topic, popping up on the news, in ads, on your messenger apps, and in conversations with friends and family. But as AI becomes ubiquitous and grows at an ever-increasing pace, what does it mean for the financial markets?

In MoneyGPT, Wall Street veteran and former advisor to the Department of Defense James Rickards paints a comprehensive picture of the danger AI poses to the global financial order, and the insidious ways in which AI will threaten national security. Rickards shows how, while AI is touted to increase efficiency and lower costs, its global implementation in the financial world will actually cause chaos, as selling begets selling and bank runs happen at lightning speed. AI further benefits malicious actors, Rickards argues, because without human empathy or instinct to intervene, threats like total nuclear war that once felt extreme are now more likely. And throughout all this, we must remain vigilant on the question of whose values will be promoted in the age of AI. As Rickards predicts, these systems will fail when we rely on them the most.

MoneyGPT shows that the danger is not that AI will malfunction, but that it will function exactly as intended. The peril is not in the algorithms, but in ourselves. And it’s up to us to intervene with old-fashioned human logic and common sense before it’s too late.

Leseprobe
1

The End of Markets

So, are Alexa and Siri conspiring to take over Earth? Maybe. But if they are, it's not personal. It's just gradients.

-Kenneth Wenger, Is the Algorithm
Plotting against Us? (2023)

AI in the Agora

This is how markets end:

December 2, 7:00 a.m. ET | Dow Jones Industrial Average 34,210 (prior trading day's close)

Nick Mera entered the trading floor of his family office, turned to his assistant, Sara, and said, "Good morning, Sara. What's new?"

Sara replied, "Not much. Long-term interest rates are still high but moving sideways. There's some speculation they may be at a peak. Short-term rates are up slightly; the Fed has not budged on its tight money crusade. Inflation appears constant; there's no sign that it's moving down to the Fed target. Dollar index is up to 106.52; sterling, euro, and Swiss franc all down slightly. Yuan is down again to 7.67; yen is joined at the hip to yuan-it's up to 155.78. Oil's flat, stuck around $82.50 per barrel. Gold's boring, still in a narrow range around $2,300 per ounce. Industrial commodities are mostly down on the continued slowdown in China; copper's $3.40. Ag commodities all down, but not by much. Chinese stocks down 1 percent; Japanese stocks are right with them. S&P futures up 0.50 percent, Nasdaq up 1 percent both on good earnings and on steady interest rates despite recession signs. Anything else you'd like to know?"

"That's fine for now, thanks."

Nick could see all the information Sara reported with a glance at his trading screens. Still, he preferred his dialogue with Sara to start the day. She could read millions of data points, analyst reports, news articles, press releases, and financial statements in minutes, in fact, she did so continually and was always up-to-date. With that data and background information, she could quickly determine what markets were trending, what if anything had changed materially, and what was most interesting to Nick (which she inferred by speaking with him every day). She could form tentative estimates of the reasons behind the price levels such as the yuan-yen linkage and the connection between steady interest rates and higher stock prices. She didn't report on the Russian ruble because she knew Nick didn't care; he wasn't in that market. If he expressed interest, she would adapt instantly and begin reporting on that as well. Sara was the ideal digital assistant, with a four-layer-deep neural network and a proprietary state-of-the-art version of GPT. Nick liked the fact that he didn't have to think about market data and was free to muse on other matters, including his next trade. Sara could handle that too.

"All right, OK, let's get to work. I agree long-term rates are ready to fall. They've been held up by momentum traders and the arb crowd who short Treasury notes against swaps. The problem is the banks are losing their appetite for swaps even if they're off-balance-sheet. It's hard to find swap hedges due to collateral shortages; swap spreads require owning Treasuries, which take up balance sheet. That game is over. Once Treasury rates fall, the momentum boys will fold like a cheap suitcase and we'll be on our way to a rally. Call Goldman and Citi-buy ten million ten-years and finance them overnight."

Nick didn't have to explain his views to Sara; he could just place the order. But the explanation was part of the training set. By learning the reasoning, Sara learned new patterns to go into future analyses. Besides, Nick liked having someone to talk to.

Sara said, "OK," paused for about thirty seconds, and said, "Done." It was not clear if there was a robot on the other side of Sara's order; it didn't matter. Nick was now long $10 million of ten-year Treasury notes. He had negative carry since overnight financing rates were higher than the yield-to-maturity on the notes, but Nick was betting the notes themselves would gain 20 percent or more as rates declined. His repo collate…