Think American medicine is a mess now?
Wait until Google and OpenAI try to fix it.
The billionaire bros of software and social media covet the $5 trillion the United States spends on healthcare every year. They promise artificial intelligence will remake drug discovery and all medicine.
Yet their interest is personal too. The boys seem almost offended their fortunes cannot keep them alive forever, or at least a lot longer than poor people. They have fallen hard for lifehacks and longevity extension, often with a strong psychedelic flavor.
Tech bros want to upgrade their biological software. But biology and computing are a terrible match, and the puffery-heavy culture of Silicon Valley makes the fit even worse.
—
(Thinking small — and big. With your help.)
—
As the story of Elizabeth Holmes proves.
You may remember Holmes, would-be wunderkind who dropped out of Stanford University as a 19-year-old in 2003, promising to revolutionize blood testing and medical diagnostics.
Spoiler alert: she didn’t. She so didn’t that she now spends her days at a federal prison “camp” in Texas, following convictions in 2022 for fraud and conspiracy at Theranos, her company. She should be out around 2031 or so.
That’s the 50,000-foot view.
The details tell an even more troubling story — as I realized when I finally read Bad Blood, the definitive account of Theranos’s rise and fall. John Carreyrou wrote the book in 2018, drawing on his work investigating Theranos for The Wall Street Journal. (Carreyrou, a first-rate investigative reporter, is now at the New York Times.)
As Carreyrou makes clear, Holmes knew almost nothing about the complex biology underlying blood testing. Yet she insisted Theranos could produce reliable laboratory results using samples that were a fraction of the size of conventional blood draws. She seemed to believe that through sheer willpower she could bend the laws of physics and chemistry — and wouldn’t listen to anyone she hired who tried to warn her she couldn’t.
Though Holmes was not much of a scientist, she had two world-class skills.
The first: charming wealthy, successful old men. She was not sexy, or even conventionally pretty, but her intensity and focus made her catnip to them. Theranos’s board grew to include Henry Kissinger, George Shultz, and Jim Mattis, a retired four-star Marine general.
The second: telling lies without blinking. (One could argue those two skills are actually one.)
And so Holmes convinced investors to pour about $1 billion into Theranos over a decade — at ever-higher valuations. At its peak in 2014, Theranos was worth about $10 billion. Holmes’s share of the company was worth over $5 billion, and “media darling” doesn’t begin to describe the publicity she received. “This CEO is out for blood,” Fortune magazine proclaimed in a cover story about Holmes. She was the great white whale for financial and technology reporters, a young female executive who had started her own company.
—
(Wait, wrong sassy female billionaire!)
—
As Carreyrou makes clear in Bad Blood, Holmes was ambitious and greedy and lied constantly and would likely score high on a psychopathy checklist.
She thought of herself as the next Steve Jobs and intentionally modeled herself on him. Her obsession with guarding her public image grew alongside her fame. Beyond that, though, Carreyrou does not really get close to her or figure out what makes her tick.
Instead, he focuses on Theranos itself, and the story he tells is both hilarious and horrifying.
For Theranos failed from its first day to its last. The company spent massively on research and development and hired scientists with impeccable credentials. But they were hamstrung by Holmes. She cared more about taking the smallest possible draws of blood and producing a small, attractive, iPhone-like device than producing a machine that would work. Holmes was also obsessed with corporate secrecy and insisted on tight internal information silos, preventing Theranos’s engineers, chemists, and designers from working together effectively.
As a result, even in 2015, 12 years after Holmes dropped out of Stanford, its blood testers were basically prototypes, unable to produce reliable results even for basic tests like blood potassium levels.
By then, Holmes had boxed herself into a corner. The Walgreens pharmacy chain had paid Theranos $140 million for the right to offer its tests (somehow never bothering to check if Theranos’s technology actually worked). Theranos had missed multiple deadlines with Walgreens; had she told the truth and admitted failure, Theranos might have collapsed.
Instead, she decided to use standard blood-testing machines from Siemens to run Theranos’s tests, hiding them in a second laboratory on a different floor from Theranos’s own machines. But the Siemens readers didn’t work well with the tiny blood samples Theranos had collected either. So the company diluted the samples with saline before running them through the Siemens equipment.
The results were predictably disastrous. Theranos’s findings were unpredictable — and often flatly wrong. Yet the company managed to keep federal oversight at bay by exploiting a regulatory loophole between the Food and Drug Administration and Medicare.
Finally enough people both inside and around Theranos became worried about what they saw that they started talking, and Carreyrou got on the story. Theranos, and David Boies, its outside counsel, did everything possible to keep him from writing, including trying to intimidate his sources and threatening lawsuits against him and the Journal.
They failed, luckily.
By 2017 Theranos was effectively out of business, and in 2018 it officially filed for bankruptcy.
—
(Don’t do the crime if you can’t do the time… Elizabeth Holmes walks the exercise yard in August 2025.)
—
No one should be surprised Theranos blew up.
Medicine is deeply frustrating from bench to bedside. Big gains are rare. On the basic research side, scientists can spend decades trying to unlock the role a single protein or gene plays in an illness, much less trying to fix it.
On the other end of the spectrum, physicians face cancer and other relentless diseases — and patients whose capacity for self-destruction is far greater than their willingness to change. Better sanitation and nutrition, not medical care, drove the great 19th- and 20th-century gains in life expectancy.
Computer science faces the opposite problem. At this moment the human ability to speed progress in hardware and software appears to be outrunning our ability to manage it.
Computers and software barely existed two generations ago. Today they are the most powerful force in the world. OpenAI and Anthropic, two companies that did not even exist when John Carreyrou started writing about Theranos, are now among the most valuable in the world.
—
(I don’t work for Silicon Valley. Or pharma. I don’t work for anyone except you.)
—
Of course tech bros believe instant success is possible. Instant success defines them.
So Silicon Valley rewards puffery and hates regulation, which slows down progress — and profit. Sell it now, design it later. Ship it now, fix it after that. Who cares? Everything will get worked out. And once it does, it will work forever. Computers offer replication at scale in a way that messy biology cannot.
No one should be surprised technology companies and venture capitalists see the bloated and expensive American healthcare industry as ripe for disruption. And medicine itself is increasingly dependent on biological information in all its forms, from DNA to diagnostic imaging to electronic medical records.
But Holmes’s story serves as a dark forecast of the pressure that medicine will soon face as the most powerful companies and richest people in the world try to reinvent biology and ultimately to remake humanity.
Theranos is the most vivid example so far of the risks Silicon Valley poses as it tries to bring a culture of speed and secrecy and anti-regulatory hype to medicine.
It won’t be the last.
Here’s a little known or recognized tidbit that became the key to the inevitable conclusion of Theranos:
The whistleblower was none other than George Schultz’s grandson while George was on the Theranos BOD.
Tyler Shultz was George Shultz’s grandson.
George Shultz, the former Secretary of State, joined the Theranos board in 2011 and became one of Elizabeth Holmes’s strongest supporters. He encouraged Tyler to work at Theranos.
Once inside the laboratory, Tyler discovered serious discrepancies between what Theranos claimed its technology could do and what the testing data actually showed.
Tyler went to his grandfather first.
George didn’t believe him. He sided with Holmes.
Tyler persisted, eventually contacting regulators and becoming a confidential source for Wall Street Journal reporter John Carreyrou.
It got extraordinarily ugly. Theranos lawyers even confronted Tyler at George Shultz’s own home, and the dispute became so severe that grandfather and grandson were essentially communicating through lawyers.
That may be the most remarkable lesson of Theranos:
A member of the board had a whistleblower inside the laboratory—his own grandson—personally warning him that something was seriously wrong.
And he believed Elizabeth Holmes instead.
George and Tyler eventually reconciled before George’s death in 2021. Tyler later said his grandfather never formally apologized, but eventually acknowledged that Tyler had been right.
I couldn't help thinking about the Covid testing as I read this article. The entire world's lives and livelihoods depending on the test kits. Makes me wonder how much money the makers of those tests got.