The Final Boss.
We return to the planned third and final item in the entrepreneurship video trilogy: The Inventor: Out for Blood in Silicon Valley. To recap, we began our reviews with a wannabe sultana who didn't know the cost of bread in her own fiefdom, and moved on to a bevy of hardscrabble ladies who at least knew the value of a buck, even if they then went slightly too far in their pharmaceutical dealings, towards goosing their profits. This trend, then, would come to a head with the savviest and sneakiest of them all: Elizabeth Holmes of Theranos.
We've mentioned Theranos in passing back in 2016, and the intervening few years has only added more clarity to the drawn-out fraud saga, so let's begin with the biosketch. Neither pampered Agrabian princess nor downtrodden ethnic exotic dancer (no, not the same), Holmes was about as privileged as any nominally upper-middle-classed American lass could expect to be, being possessed of a generation-suffixed Enron vice president father, which led her to Stanford via exclusive prep school. She had the bandwidth to sell C++ compilers along the way (shades of Bill Gates there), and crossed paths with Singapore as an eighteen year-old, interning at the Genome Institute of Singapore on SARS.
She would then drop out of university like all the cool, well-supported kids, having become enamoured of the idea of performing a battery of tests using merely a few droplets of blood. To this end, she founded Theranos in 2003, which grew into a super-unicorn valued at some US$10 billion by 2014, making her a self-made multi-billionaire and primo feminist tech icon feted in TIME magazine. Zoom ahead a few more years, however, and her net worth's back to approximately zero with Theranos wound up, facing nine Federal counts of wire fraud totalling possibly twenty years in prison, and with her lawyers wanting out due to nonpayment of bills.
How did it all go so wrong?
No Way Back
be first, be smarter, or cheat."
- Margin Call (2011)
Shorn to the essentials, the cause of Holmes' and Theranos' downfall is disarmingly simple: they promised a machine that could perform blood tests on tiny quantities of blood; after a dozen years or so, it transpired that they, in fact, did not have such a machine, and were nowhere close to creating one.
Note, however, that the sin lay not in promising something that they did not yet have (at least in the early days). Heck, that's the entire point of start-ups and Kickstarter campaigns. Returning to the Gates example, he essentially licensed what would become MS-DOS from another Seattle company, after negotiations for CP/M fell through. Holmes' more-visible black-turtlenecked role model Steve Jobs had Woz to do the heavy technical lifting. Prompt outsourcing is certainly one way to be first to market, and has perhaps reached its utmost expression in a recent murder-for-hire case where the assassination got subcontracted out five times, before the last guy met the targeted victim and offered to split the fee if he played dead.
No, the first problem is arguably the choice of domain. Having non-accredited founders in near-pure tech fields such as advertising (Google), hipsterism (Apple), selling books (Amazon) or induced social anxiety (Facebook) is one thing. Biochemistry and medicine, quite another (which is part of why Singapore's slinking away with tail between legs here, what with all the clinical tests, regulatory approvals and ISO certifications involved; it's a pain, from personal experience). Put another way, nobody bothers overmuch if Google misranks or entirely misses out on a website, or Apple omits the headphone jack from their gear; misdiagnose a critical illness, though, and a lot of very scary organizations are gonna get involved.
That said, one might wonder why anybody would bother getting involved with biotech, and especially the hardware side, given the heavy capital requirements (this observation was also made at the Deep Learning Indaba conference I spoke at some months back, where one panellist identified software development as Africa's future for much these reasons). The answer is that the moat works both ways: it may be tough to get inside, but once you're in, the moat protects you from competitors to an extent. For many app-based products, there's always the looming risk of getting cloned down to the last pixel by dev sweatshops (refer to Chapter Two of Kai-Fu Lee's AI Superpowers, for one description)
The half-bearded author of The Upside of Irrationality chips in
The documentary goes over how Theranos pivoted unsuccessfully through several approaches - microfluidics, a gluebot, and then miniaturization, which was perhaps as expected, given that nearly all her early scientific advisors knew better about the limitations of these technologies from the beginning. This was not to say that it was absolutely impossible for an undoubtedly conventionally-brilliant nineteen year-old with some introductory courses under her belt to essentially push the scientific cutting-edge forward, of course - just that it was vanishingly unlikely.
Adding to the overall murkiness, Theranos published next to nothing about their development work, which while kinda understandable given that this was essentially revealing trade secrets, wouldn't do much to foster confidence. Holmes had set up an intelligence agency of sorts, where employees from different departments were disallowed from discussing their work, such that only she (and possibly some few trusted underlings) was aware of the whole (not-so-nice) picture. This, together with some good P.R. and the backing of more than a few retired warmongers, turned out to be sufficient to secure US$700 million in funding, and a Walgreens rollout.
Behind the scenes, however, it was a total clusterf**k of cherry-picked tests, secret (and improper) usage of existing Siemens analyzers, and outright cheating. To be clear, it never had to get this far - Holmes could have bowed out, probably at any time before about 2010, after admitting that she had done her best and it didn't quite work. That would be it. It happens, particularly with hardware (e.g. the Kreyos scandal), but also in software (e.g. the initial release of No Man's Sky, all manner of other vaporware). She could have walked away, reputation unblemished, or even enhanced, given Silicon Valley's abiding taste for comeback stories.
We know, of course, that she didn't, and instead doubled down on the outright lying... with predictable results.
Tech Hype, or the Unicorn is a White Horse with 3D-Printed Horn
A large part of how Theranos got as far as it did, was how much investors wanted to believe - and frankly, this is hardly limited to biotech. The latest cautionary tale would be that of co-working space provider WeWork, which started off with an expected IPO valuation of US$47 billion in January, only to have it cut to US$10 billion (possibly already generous, given IWG's valuation of US$3.7 billion) upon cursory inspection, with the CEO stepping down and possibly a third of its workforce to be laid off. This might go to show, once again, that it may be the timing that matters the most - hop onto the crest of a bubble wave, and it don't matter what you're pushing, be it blood tests, dog walks or subletting by another name, as long as you can brand it as "tech", "A.I." or old-fashioned "big growth!".
Quite interestingly, there's a Theranos connection too, with SoftBank also having loaned them US$100 million - which may just go to show how easily Saudi, UAE and Our Most Successful Investment Firm (and perhaps even Malaysian) easy money flows, in this low-returns environment. Given how nonsensical WeWork's valuation was, it's small wonder that there're rumblings on whether SoftBank's simply out to launder money here. Personally, I'm not exactly sanguine about the long-term future of brick-and-mortar, but I suppose it could still make sense in certain specialized applications, such as labs-for-rent - and, most of all, at the right price, in an age where cars, shoes and sweaters are increasingly getting placed on payment plans.
On the other hand, I remain unconvinced that the opposite - tech being underappreciated - doesn't hold true here, all the sudden attention on A.I. considered. It'll probably have somewhat more staying power than biotech, though.
Healthcare Costs & Implications
It's difficult to disentangle Theranos' motivations from medical costs, and while we have discussed it through the example of the EpiPen injection device for insulin in 2016, the scourge of rising prices has shown few signs of slowing. Over in Great America, TRUMP remains battling both his own party and colluding drug firms towards having the government negotiate drug prices, but to little avail thus far, with nary a peep of support from the bought-and-paid-for FAKE NEWS. Not only that, his representative might have been suicided - and with billions on the table, why wouldn't that be plausible? Faced with this, the only responsible thing to do was to make sure that prospective immigrants could pay their own way.
Singapore isn't entirely in the clear either, with a number of overcharging cases hitting the media recently, atop the use of debt collectors by one of our general hospitals hitting the Chinese papers. This perhaps reflects the special status of healthcare in the public consciousness, as being an unavoidable necessity that should be guaranteed, no matter what. It's one thing to sic the bailiffs on a guy who's behind on his Ferrari, and another to hound him about a medical bill, so it seems.
Disregarding bad-faith treatment by the black sheep of the medical community, as with wanton prescription of opioids and false billing claims, it remains that the overall economics of healthcare is exceedingly dysfunctional by its nature. There's a huge principal-agent problem to begin with, in which the medical professionals are supposed to be much more knowledgeable of the patient's needs, than the patient himself. To compound this, the doctor-patient relationship tends to be tighter than, say, with car workshops. It might be expected to shop around for a mechanic, but less done to ignore the family dentist's call for fillings... and this is without going into emergency work that the patient may have effectively no say in.
From the hospital's perspective, however, they have wages and bills like everyone else, and in countries without universal healthcare and privately-run practices, this money has to come from somewhere. Thus, the abomination of US$200 slings, near US$50000 to set a broken leg, and US$350 to hold one's own baby after birth. Then, since next to nobody actually pays list price, what usually happens next is prolonged haggling, pleading and threats between the patient, hospital admins and insurance lawyers, that's worthy of a Turkish bazaar.
Given all these bugbears, it seems more and more to me that the clubby atmosphere of the traditional medical fraternity might not indefinitely escape co-option by technology, such as the A.I. medical booths covered in February. For the right price, I'd gather many would be prepared to take certain risks, because they can't afford a US$200 sling elsewise anyway. From the ongoing debate about telemedicine in the broadsheets, local doctors are likely recognizing how the wind's blowing.
Today & Tomorrow
Finally, to conclude the Holmes story thus far, she's reportedly happily married to a hotel heir, with some wags uncharitably foretelling a baby bump right on cue for her July 2020 trial. Still, with in-supermarket and fingerstick tests remaining very much in the works, it's easy to dream about what could have been. Could she have simply bought over a better-run competitor, with a half-billion dollar warchest? Or could she have instead tamped down her promises to something more realistic? Hey, she had Kissinger on her side...
If there's one lesson to be learnt from the whole mess, it's this: you can't, and frankly don't want to, get away with fudging results in medicine. It's not worth it.
Next: Short Diwali Breather
Copyright © 2006-2020 GLYS. All Rights Reserved.