Michael Miebach, Mastercard chief executive, noted that the same consumer protection questions the payments industry has addressed over the past two decades are now resurfacing in the agentic commerce space. Professor Andrew W. Lo added: "One of the things about large language models that I find particularly concerning is that no matter what you ask it, it'll always come back with an answer that sounds authoritative, even if it's not."

Professor Andrew W. Lo said that AI struggles with tax optimization, doesn't understand regulatory nuance and — unlike a human financial adviser — isn't subject to legal requirements, such as acting in a client's best interest. He stressed that it's important to ask critical questions when using AI for retirement advice, such as prompting an AI to say where it might be wrong and to list its assumptions and uncertainties.

Professor Andrew W. Lo expects AI models to quickly advance and be able to serve as fiduciaries in coming years. As of now, he cautioned, they aren't ready for prime time. "We believe that it is possible to train an LLM, just like we train humans, to provide fiduciary duty," Lo said. "But they don't have it right now, and the guardrails aren't there to protect individuals."

"I think that there's a real art and science to prompt engineering," said professor Andrew W. Lo. "When it comes to specific calculations of your own personal situation, that's where you have to be very, very careful," he said. AI can also sometimes provide wrong answers due to so-called "hallucination" of the algorithm. "One of the things about large language models that I find particularly concerning is that no matter what you ask it, it'll always come back with an answer that sounds authoritative, even if it's not," Lo said.

"The problem that we have to solve is not whether AI has enough expertise," said professor Andrew W. Lo. "The answer right now is, clearly, AI has the financial expertise. What they don't have is that fiduciary duty. They don't have the ability to suffer consequences if they make a mistake to the same degree that a human advisor does." The notion of putting a client's interest ahead of yours "has no teeth" without responsibility or legal liability, he said.

Professor Andrew W. Lo and co-author wrote: "Across pharma and academia, an estimated 5,000+ shelved drug candidates were discontinued for reasons unrelated to safety or efficacy. Each represents a potential therapy for conditions that, in many cases, have no approved treatment at all. Industry stakeholders have a unique opportunity to collaborate on identifying these compounds. Aligning these assets with capable and motivated partners will benefit both drug developers and patients."

Professor Andrew W. Lo has described today's AI chatbots as the "digital equivalent of sociopaths." While the language may sound harsh, he describes them as smooth, persuasive and devoid of empathy. They can present both good and bad advice with the same confident tone.

The internet can be filled with unsympathetic commenters, criminals and terrible advice. Look for online groups that screen new members and have active moderators, suggested professor Andrew W. Lo.

Professor Andrew W. Lo argues that today's AI chatbots are fundamentally unsuited to serve as financial advisers. In his view, they resemble "digital sociopaths," fluent and convincing, yet lacking empathy and moral judgment. Even so, Lo believes large language models could eventually become valuable tools — particularly for small investors with limited experience.

Should you use AI for financial advice? Professor Andrew W. Lo says not yet. Large language models like Copilot or ChatGPT aren’t suited to being used as financial advisers because they are the digital equivalent of sociopaths — smooth, persuasive and devoid of empathy.

Professor Andrew W. Lo, senior lecturer Egor Matveyev, and co-author wrote: "In our paper, we compile the first comprehensive dataset covering the full universe of U.S. nonprofit endowments. Using IRS Form 990 filings from 2008 to 2020, we study nearly 375,000 nonprofit organizations, including about 40,000 that maintain endowment assets. We combine detailed investment disclosures on Schedule D with governance, compensation, and financial information on the main form to examine how endowment strategy and performance relate to organizational structures and oversight practices."

Investors should ask broad and unbiased questions when interacting with large language models in order to get the most accurate answers, professor Andrew W. Lo told Axios.

"Even before the change in policies that are being implemented now, vaccines were a difficult business," said professor Andrew W. Lo. "It's just become that much harder." Smaller biotechnology companies, which already tend to operate on thin margins, "will get out of the business," Lo said. If a new pandemic were to ignite in the current climate, "Operation Warp Speed Part 2 would not operate at warp speed," he said.

Professor Andrew W. Lo said: "I think that a large number of individuals, who are currently not getting any financial advice and badly need it, will have access to pretty good financial advice at no cost. That's the promise of AI over the course of the next few months, not to mention years."

"All financial innovation is generally there to solve a problem," says professor Andrew W. Lo. "If you've got some long-dated obligations, you want long-dated assets to mature and pay them off." Lo, speaking generally and not about Fox Hedge, adds that complex ways of engineering long-dated investments could end up creating their own problems if they become common.