Tyler Shultz warns that AI gives fraudsters room to deceive

The film, and the person in it

“You Can See Everything” follows Elizabeth Holmes in the weeks before she went to prison. Directors Nathan Fielder and Lance Oppenheim do not present her simply as a master manipulator or a victim. Instead, the film shows how she can elicit sympathy even from people who know about her crimes.

The documentary has become one of the highest-grossing documentaries in history and prompted wide discussion in the United States. Shultz, a former Theranos employee who became one of the company’s most important whistleblowers, told The San Francisco Standard that Holmes is discussed now only because she wants to remain in the spotlight.

The film’s opening question is why a mother of two young children would invite a crew to document her last days with her family before beginning a sentence of more than 11 years. The film leaves much unanswered. The article’s interpretation is that Holmes wants people to feel sorry for her — a response that could be more useful to her future ambitions than ridicule or hatred.

The Theranos test for AI investors

Theranos was valued at about $10 billion at its peak, despite never making its technology work. The company ran only some tests on its own devices; the rest were performed using equipment from other manufacturers. Its machines reached Walgreens stores, and real patients received dangerously inaccurate results.

Shultz argues that investors should move beyond the idea that a half-hour coffee meeting is enough to justify putting in $100,000. He said Holmes scaled that model up to $100 million. His grandfather, George Shultz, served as U.S. secretary of state and secretary of the treasury, was a prominent investor and sat on Theranos’s board.

Holmes raised money from individual investors, including the DeVos and Walton families, rather than relying mainly on venture funds. Theranos did not show prospective investors scientific data or proof that its devices worked, citing the risk of exposing trade secrets. That secrecy put off many leading Silicon Valley venture capitalists.

Shultz believes Holmes raised the largest angel-investment round in history. But the larger lesson, he says, is about the power investors hand over when they place large sums in one person’s hands. The collapse of Theranos was a sharp reminder to check investments carefully; he worries that AI investors may forget it.

He also says diligence is harder in AI: it can be difficult to tell genuinely new technology from a shell built around ChatGPT. I think that is the most useful part of the comparison. “AI” is not evidence that a company has built something distinctive, just as a promise about trade secrets was not evidence that Theranos’s machines worked.

The next chapter

In the film, Holmes and her partner, hotel-business heir Billy Evans, are working on a diagnostics company called Haemanthus. They call it “Theranos 2.0.” The company has reportedly raised about $20 million.

The documentary does not portray Holmes favorably, but it sidesteps some of Theranos’s most serious conduct: the dangerous medical information given to patients and the company’s targeted pursuit of Shultz as he tried to expose his former employer. That omission matters. Holmes remains more a figure of public mockery than someone most people want to do business with.

Shultz said he is worried about what comes next. He thinks Holmes is preparing a second act, but does not know what it will bring. My guess is that the harder test for any new venture will not be whether it can attract attention or capital, but whether outsiders can verify what its technology does.