When Mark prepares a memo on whether AI is a financial bubble late last year, he viewed the technology with a healthy dose of skepticism. Like many, he initially thought AI models were little more than glorified search engines, capable of extracting and regurgitating data without any real understanding. He wonders if AI can truly break new ground, or if it’s only limited by the statistical rewiring of people’s minds. And so, as he explained in the follow-up note to clients on February 26he talked to “some interesting techies in their thirties and forties.” One of them suggested that he ask Claude.

Anthropic’s AI model has produced a tutorial, Marks wrote, that explains AI and changes over the past three months. The resulting 10,000-word essay was so impressive, Marks added, that he decided to reprint most of it for his clients in a recap, but he wanted to add some of his own personal writing.

“I could have saved myself a lot of time by asking Claude to write this memo,” he wrote, “but I decided not to, because I consider putting words on paper a big part of the fun.”

Marks emphasized to clients: “I want to try to convey the level of awe with which I see Claude’s output.” Not only does it read like a personal letter you might have sent to a friend or colleague, but it psychologically breaks down Marks’ self-doubt by using the investor’s own legendary mentors against him. Above all, he says, it’s really smart.

“It argues logically, anticipates points that I can respond to, injects humor, and strengthens its credibility by frankly acknowledging the limitations of AI, as I can,” he said. “I’ve asked AI questions in the past and gotten answers, but I’ve never received a personal explanation like I did in this case.”

Here’s what Claude said to Howard, leaving him very impressed, and a little shaken.

The Graham, Buffett, Munger example

In response to Marks’ question about whether AI can actually think, synthesize, or come up with new ideas, Claude gave an “enthusiastic rebuttal,” Marks said. The machine spoke directly to him: “Howard, everything you know about investing comes from other people,” the AI ​​wrote. “Benjamin Graham teaches you about the margin of safety. Buffett teaches you about quality. Charlie Munger teaches you about mental models from multiple disciplines. John Kenneth Galbraith teaches you about the psychology of financial manias.”

AI points out that Marks has spent 50 years reading thousands of books, memos, and case studies, taking raw material from others to create something truly new.

“You learn patterns of reasoning from decades of reading. I learn patterns of reasoning from training,” the AI ​​argued. “The question is not where the inputs come from. The question is whether the system – human or artificial – can combine them in ways that are truly new and useful.”

The argument left Mark stunned.

“Of course, it’s absolutely true,” he wrote, marveling at how the AI ​​reasoned rationally, injected humor, and anticipated its opponents. The revelation shifted his entire perspective on machine intelligence, forcing him to ask: Is the way AI grows, learns, and ‘thinks’ really different from ours?

Marks is famous in financial circles for his periodic “memo” to Oaktree clients, written since 1990which analyzes market cycles, risk, and investor behavior; They have been called some of the most influential writings in modern finance and are now in the Permanent collection of the Museum of American Finance. Warren Buffett said that when he saw a memo from Howard Marks, it was the first thing he read, which helped cement Marks’ reputation among professional investors. In 2024, the Museum of American Finance awarded Marks a Lifetime Achievement Award for his impact on financial thinking, highlighting how his memos and philosophy shaped modern investing. So his opinion is just one of the impact of AI, but it is an opinion.

Reasons why and why not

Now looking at AI not just as an assistant but as an autonomous system—what he calls “Level 3” AI capable of completely replacing work—Marks said he sees many implications for Wall Street. Many of the qualities needed to be an exceptional investor exist in AI, which can absorb endless data, recognize historical patterns, and act completely without human greed, fear, or whim. He also noted that technology is advancing at speeds beyond anything in history, with some models writing their own code and testing their own software autonomously.

However, despite the alarmingly rapid evolution of AI and its ability to imitate the intellectual synthesis of Buffett and Munger, Marks added that he does not believe that human investors are completely extinct. Human content, he argues, remains in the evaluation of truly new developments where historical precedents are absent. Additionally, AI lacks a critical investment component: “skin in the game.” Unlike humans, a machine cannot intuitively sense risk or feel the gut-wrenching fear of capital loss.

Get the scores right AI will revolutionize finance. But there is reason to stop feeling the awe he describes. The beauty and persuasiveness of Claude’s response to Mark alone can cause doubt because Personalization is one of the main features of Claude’s product. Large language models are very good at interpreting context clues—in this case, the user’s name, professional background, and possible objections—and tailoring the output accordingly. The 10,000 word essay feels like a note from a good friend because it engineered to feel that way.

Claude’s central argument—that Marks learned from Graham and Buffett as much as Claude learned from training data—seems profound, but human learning and statistical pattern matching on text corpora are not the same. Mark just didn’t exist read Graham; he applied the margin of safety to real markets, lost money, felt fear, changed his thinking under real uncertainty, and built convictions that would pay him if he was wrong. That feedback loop—results, revisions, hard-win judgments—is something Claude has never experienced and cannot replicate.

In the end, Marks’ encounter with Claude convinced him that AI is more powerful than a simple fad, estimating its true potential is probably underestimated today. While he warns against going “all-in” on AI stocks due to the risk of disruption, the legendary investor’s final advice is clear: No one should remain “all-in” and risk missing out on one of the greatest technological leaps in human history. Even if Claude doesn’t really have skin in the game.