You can see every red flag in Theranos 2.0

Column Science Health You can see every red flag in Theranos 2.0 Knowing common health tech grifts comes in handy for assessing the secretive startups in You Can See Everything .

Victoria Song Oct 28, 2026, 6:00 PM UTC Link Share Gift They’re getting the band back together for Theranos 2.0. Supposedly. | Photo: Getty Images Column Science Health You can see every red flag in Theranos 2.0 Knowing common health tech grifts comes in handy for assessing the secretive startups in You Can See Everything .

From afar, it’s easy to look back at the Theranos debacle and scream at Nathan Fielder that, obviously, Elizabeth Holmes is lying in You Can See Everything . In the weeks since the movie’s release, journalists and whistleblowers alike have rushed to remind moviegoers of Holmes’ numerous documented lies. Chief among the facts: The Theranos device never worked as advertised. But Fielder’s aim with this film wasn’t relitigating Theranos. Even the most mind-bending of three-hour documentaries doesn’t have enough time to fully tackle the quagmire that was the secretive biotech startup.

It’s worth remembering that Theranos was a prime example of overinflating what technology can do. It sure would be nice if a single droplet of blood could be run in a magical machine and beep boop — you’ve got accurate results for over a hundred diagnostic tests. It’s a compelling concept, but the existing science doesn’t support the idea. As I wrote in last week’s Optimizer , you simply can’t test for every disease from blood alone. Sometimes, a blood test is merely part of a multistep process to rule out what you don’t have. A finger prick carries a higher risk of rupturing blood cells, which can skew test results. Your finger is more prone to surface contamination because hands are dirty by nature. Compared to drawing blood from a vein, a finger prick may not be purely blood — it often mixes with interstitial fluid, diluting the sample. Different biomarkers can require wholly different testing procedures, making the small volume collected difficult to work with. These are just some of the practical challenges Theranos had to solve to be better than the clinical gold standard. It couldn’t, and that’s why they had to fake it by bringing in Siemens devices that used traditional blood testing methods.

That makes it even more astonishing that Holmes and her partner Billy Evans spend a chunk of the film actively working on what Evans casually calls Theranos 2.0.

This new venture, Haemanthus, involves a similar diagnostic device that uses… lasers. The official line is that Holmes is not helming the effort, but Evans refers to it as “getting the band back together.” The film purposefully blurs the lines between truth and lies. But with nascent health tech, it’s genuinely hard for the average person to distinguish promising medical advancements from snake oil.

If you left You Can See Everything wondering if there’s any truth to using lasers to diagnose conditions — you don’t have to watch microbiology videos at twice the speed to find out. Instead, here are a few extra tips when it comes to spotting shenanigans straight out of the wellness grifter playbook .

In the documentary, Fielder likens Holmes to a magician. Technology, no matter how powerful or well crafted, is never magic.

But going along with Fielder’s metaphor, magicians succeed through misdirection. Evans presents lasers in the film as though they’re a little-known marvel — a key technology Theranos was missing — and that Haemanthus will be the first to make them truly useful. Based on a patent published by The New York Times , Haemanthus uses Raman spectroscopy. For the record, Raman spectroscopy does involve lasers, but it’s not a new technology. It was first discovered in 1928.

In a nutshell, this laser tech is used to identify chemical structures in a molecule. There are myriad biological and medical uses for it, including noninvasively studying tissue samples. But will it, as a New York Times report describes, “identify and qualify biomarkers such as glucose and hormones” and then use AI to “detect cancer and infections”?

Researchers are exploring using this tech to diagnose early-stage cancers . They’re also looking into using it for glucose monitoring . Does that mean this is a better method than what currently exists? Maybe for certain cases, but probably not in every scenario. In medicine, it’s common to require multiple tests for diagnosis. Similarly, there are often multiple ways to treat the same condition. It’s often a game of weighing pros and cons for various approaches. There are too many unknown factors to accurately comment on Haemanthus, but it’s concerning that Evans zeroes in on lasers like they’re a panacea for everything that plagued Theranos.

When investors, journalists, and the public ask hard questions, founders love to sprinkle technology into their rhetoric like it’s fairy dust. You can see it happening right now with AI. But tech is a tool . Any tool will have pros and cons, situations it excels in and scenarios it’s ill suited for. It’s not always a grift, but anytime an executive shows signs of magical thinking or inflates what technology can do, your bullshit alarm should be ringing.

In recent history, I can think of exactly one time the public health sector truly hauled ass to make something happen: the covid-19 vaccine. The vaccine took roughly 10 months to develop, compared to the typical five-to-15-year timeline. But even then, it was a Herculean effort requiring global coordination and tens of billions of dollars , largely from the government. (That’s not including the billions invested in researching mRNA vaccines during the three decades prior to the pandemic.) It was a heartening feat of human ingenuity — but again, it took 10 months built on 30 years of foundational research and several billion dollars.

So it’s an immediate red flag when Evans says in the documentary that Haemanthus would have a novel, revolutionary prototype working in about a year. Case in point, he said that in 2023, and it’s already 2026. But even if lasers were the key to everything, there’s no way in hell something like this would breeze through the regulatory process.

Not much is known about Haemanthus . It has no website, unless you count a sparse LinkedIn page that says the company has two to 10 employees and is working on “AI-native sensors for health,” a technology that “captures thousands of biomarkers.” The bulk of what’s publicly known comes from You Can See Everything and a 2025 New York Times article , which states Haemanthus is building a device that uses a “laser to scan blood, saliva, or urine from pets and analyze the sample on a molecular level.” (The startup is apparently aiming to do veterinary tests first before perhaps eventually graduating to humans.) Its “long-term goal” is a “stamp-size, wearable version” of its product for humans, which will “require three years and $70 million.”

This is an incredibly expeditious timeline for any medical device, let alone a novel one. Not least because the biggest names in wearable technology right now can only detect — not diagnose — signs of perhaps two or three conditions. The holy grail is considered noninvasive continuous glucose monitoring, an endeavor that started in 1975 and that Apple has been rumored to be working on for over 15 years now. In 2023, I did a deep dive into why you shouldn’t expect this to arrive on smartwatches anytime soon. The reasons ranged from weak signals and regulatory mazes to skin color impacting how light reflects and battery life. In 2026, not much has changed on this front.

Meanwhile, take the continuous glucose monitor. Attempts to monitor glucose levels started in the mid-1800s , but a portable way to monitor glucose didn’t emerge until the 1960s. The first FDA-approved CGM was the Medtronic MiniMed in 1999 , but it could only be worn for three days, and data couldn’t be viewed live. A clinician had to review it and inform the patient afterward. The first real-time CGM wasn’t available until 2004, the first over-the-counter CGMs didn’t hit the market until 2024, and only this year did Abbott announce a CGM that can measure two biomarkers .

AI advancements could theoretically help remove some technological roadblocks that may, in turn, speed up the overall process. But you know what AI doesn’t really replace? The absolutely labyrinthine regulatory approval and clearance process. You can have a product made and ready to go, only to have it held up in regulatory limbo for years. It’s happened to several Withings products . The infamous wearable taint zapper took six years , an untold number of testing bunnies, and at least a million dollars to hit the market. It’s why wearable companies are currently lobbying in Washington , likely spending mountains of cash to lower these hurdles.

So the next time a biotech startup promises something revolutionary relatively fast, it’s usually a pretty good sign they’re either delusional or not telling you the whole truth.

Another red flag with Theranos was that the company never published its research or submitted it for peer review. Haemanthus’ pseudo-secrecy — if you really wanted to build in stealth, why would you talk about it at length in a documentary? — is another concerning sign.

Perhaps because of Theranos, health tech companies have begun publishing more of their science. Oura, Apple, and Google are all great examples of companies that have published clinical and peer-reviewed research in reputable journals, as well as white papers explaining their features.

The act of publishing doesn’t get you a hall pass from scrutiny. Science-washing in the wellness and health sectors is rampant, with some companies using consumer perception studies as “evidence” instead of more rigorous studies. Or, if clinical studies are done, they’re often more of a vanity project than a groundbreaking revelation. Lastly, not all peer-reviewed papers are created equal. You have to be able to read and analyze the fine print to tell if the study itself was done well.

However, the absence of any reviewable research should sound the alarm for anyone who values science. In early 2015, months before the damning Wall Street Journal report , Stanford University professor John P.A. Ioannidis penned a critique of Theranos’ lack of peer-reviewed research in JAMA . Theranos responded by siccing its lawyers on Ioannidis in an attempt to get him to retract his opinion. That’s sketchy as hell — especially for any biotech or health tech company.

Based on these criteria, Theranos and Haemanthus set off a cacophony of obvious warning bells. But they’re not the only ones. Every year at CES, there are plenty of health tech startups peddling dubious concepts and vaporware right next to people genuinely doing the work. Health is the most valuable thing anyone has, so it’s not always easy to see false promises when your eyes are clouded by hope. No one wants to be sick, and technological miracles are an easier pill to swallow than the protracted ups and downs of real science. But if there’s one thing I took away from You Can See Everything , it’s that Theranos’ playbook is still, unfortunately, alive and well.

Victoria Song Column Health Optimizer Science Most Popular Most Popular Here’s why the iPhone Duo’s hinge could be cracking the screen All the news from Apple’s ‘Pro and tell’ event Touchscreen MacBook live blog: On the ground at Apple’s launch event Apple announces its first touchscreen MacBook Pro Amazon’s new Echo line is here: the 6 biggest updates Advertiser Content From This is the title for the native ad

Fuente: The Verge