Theranos: Blitzscaling, Blitzfailing, or Blitzfrauding?
linkedin.com
linkedin.com
That doesn't even begin to describe the extent of malfeasance. John Carreyrou specifically mentions one patient who based on Theranos testing ordered thousands of dollars worth of further medical tests. Their prothrombin time was completely worthless, and people had their anticoagulants adjusted on random numbers. People might have died because of that.
In any case, would love to see the folks who made the Enron documentary ("smartest guys in the room") make one about Theranos.
For example, Reid suggests Theranos could have have started by offering traditional blood drawing services transparently, instead of taking the pin pricks. Why would CVS have permitted access to their stores for such an incremental test? Would investors have financed such a mundane trial?
Theranos' access to patients and capital seems to have depended explicitly on falsifying how close they were to a breakthrough. That's not a hustle that happens to be dishonest, it's a business model that can't progress without dishonesty.
Strip away the potential for dishonesty, and "blitzscaling" is just "grow as fast as you can by honestly engaging your market"... Duh?
I realize there are reasons for some people to question medical decisions, like if your family was victim to something like the Tuskegee Syphilis trialing, but people need to know their thirty minutes of online browsing doesn't replace a decade of medical education.
(note I'm not saying we shouldn't be able to do some of our own testing like blood sugar level, cholesterol, some others, but more like things where just because you had an elevated level of X doesn't mean you have Y and you should self treat immediately. And in probably jaded on this because she deals with a lot of people that think either vitamin C or the power of prayer will cure them or that she's lying about cancer because a chiropractor's thermagram doesn't show it)
My wife teaches a class at Stanford called "The Religion of Stanford and Silicon Valley". It has been interesting to see how the course has morphed over the last few years, with Theranos, Juicero, Hampton Creek, and other notable failures (both financial and moral). The above quote from Elizabeth Holmes definitely made it into a Powerpoint slide.
How would such "explicit risk mitigation" look like? Passing legislation that indemnifies the manufacturers?
How are car risks currently expressed and addressed? How many risks are currently accepted instead of mitigated?
If two autonomous wehicles (by the same manufacturer to make things interesting) crash into each other what mitigation can there be?
This is what the (literal, in the literal sense) firewall between the engine and the cabin mitigates - the fire stays in the engine compartment, so the passengers have more time to escape the (usually stopped) vehicle.
Back to my question; if two autonomous wehicles by the same manufacturer crash into each other what mitigation can there be?
It's like building an airplane. If you follow all best practices the plane will be pretty safe but from time to time some still will crash. There is almost no way around that.
Similarly, you could have a late stage cancer drug that kills half the people who take it and saves the other half and it provides a massive benefit if 100% if all those people were going to die otherwise. The fact that a lot of people die is not the only factor to be considered. What matters is how the ultimate outcomes are changed. It would be quite irresponsible to say that the cancer drug shouldn't be allowed solely because it kills half the people who take it.
This leaves some sort of bulk protection for manufacturers ?
And blitz.
No, it's "pursuing rapid growth by prioritizing speed over efficiency in the face of uncertainty", which is what the author wrote in the very first sentence of the article.