Last year, Nassim Taleb was on a podcast with Joe Walker. It became “controversial” in some circles because of some things he said about negative probabilities and quantum mechanics — ha ha, look at Taleb, barging into the wrong lane again — but the highlight for me was what he said about the world of startups and venture capital:
TALEB: No, no, the mechanism. They don’t make their money, venture capitalists, they don’t make money by waiting for the company to really become successful. They make their money by hyping up an idea, by getting new investors and then cashing in as they’re bringing in new investors. I mean, it’s plain: look at how many extremely wealthy technology entrepreneurs are floating around while not having ever made a penny in net income. You see? So the income for venture capital comes from a greater fool approach.
WALKER: Okay, so a Ponzi kind of dynamic?
TALEB: Not necessarily Ponzi, because you’re selling hope, you package an idea, it looks good, so you sell it to someone, and then they have a second round, a third round. They keep [doing] rounds so you can progressively cash in.
WALKER: Got it.
TALEB: It’s not based on your real sales. Or your real cash flow. Particularly in an environment of low interest rates, where there was no penalty for playing that game.
WALKER: Do you think there’s any skill in VC?
TALEB: They have skills, but most of their skills are in packaging. Not in …
WALKER: Not for the things people think.
TALEB: Exactly. Packaging, because they’re trying to sell it to another person. It’s a beauty contest.
Of course, these dynamics continue to be in play beyond the private equity stage, reigned in somewhat by financial market rules or what is left of them. Assuming what Taleb said was correct — and if you disagree you may as well skip reading (or better yet, drop a comment below because I would love to see the arguments against Taleb’s thesis) — the VC-backed startup model is uniquely ill-suited to funding drup development, and I would wager that many of the issues lamented over substack articles and policy papers can be traced back directly to its rise.
Of course, this blog being what it is you will not get graphs showing the parallel increase in sheer number of biotech startups, the cost of clinical trials, and the millions of dollars spent per sucsessful drug (by whatever definition you chose), with a few thousand words on why this particular correlation does imply causation. If you have enough time to make such a graph, please share. Instead, here are three reasons I came about by the way of armchair research for why biotech is not tech:
- Larger freedom to bullshit (than tech)
- Harder to test against reality (than tech)
- Less end-user price sensitivity (than in tech)
The first two apply to many non-tech fields: social sciences, philosophy, interpretive dance, etc, and in those two ways it is tech, by which I mean both software and engineering (though software is somewhat more malleable) that is unique. What sets anything related to medicine apart is that sick people and their families are willing to spend unlimited amounts of money to get their health back, or even to just squeeze out a few more good months out of a dire prognosis. This is as true for America’s hyper-capitalist health care “system” as it is for any “socialized medicine” country whose citizens still find ways to collect money via online fundraiser or SMS services to send their loved ones abroad for life saving therapies and stem cell scams alike.
Freedom to bullshit is particularly large in cutting-edge fields, which are exactly the places where we wound now go to startups instead of established players to sow the seeds. The data is more likely to be confidential, not yet peer-reviewed or even presented at a conference. Uncertainty is always higher here, and this is why the professional sociopathic bullshitters will have an edge: earnestness may get you a miss congeniality prize but you will never get the crown.
Testing biotech against reality can only be done — for now — in clinical trials, and those are notoriously slow, expensive, and easy to explain away in case of negative results (see the case of Sarepta). A failed rocket launch is a failed launch: we will have learned more for the next one but can’t pretend we’ve reached orbit. It is also severely supply-constrained: there are only so many clinical trialists around, running clinical trials is not their day job, and those who find clinical research interesting are likely to jump ship and transition to industry further limiting supply. This is the area with most room for improvement and kudos to Ruxandra Teslo and Willy Chertman for making the case for clinical trial abundance, but there is a major headwind to their efforts:
Good health has no price or a variant thereof exists as a phrase in many, maybe most, world languages. Let’s not go into why that is (economist Robin Hanson has some interesting ideas), because my interest is in the outcome: the health care ouroboros. Not to be confused with the ouroboreos What is that, you ask? It is the curious phenomenon of economists explaining that Americans pay more for health care than anyone else in the world because they subsidize the world’s drug development, while at the same time pointing to high costs of health care as a reason for why clinical trials in America cost so much! Truthfully, this would not be a problem if it didn’t directly contribute to item number 2, leading to an unsustainable positive feedback loop when it comes to price, and a negative one when it comes to actual working drugs making their way to patients.
So let’s review the proposed interaction between the three factors that has led to our current drug development quagmire. Bullshitters are attracted to the American medical honey pot. If there was money to be made in interpretive dance they would have chosen that, but alas it is health care that both has large amounts of money sloshing around and gives them enough freedom to perform their hijinks. The particularly outworldly bullshitters in other fields can only do this up to a point — Elizabeth Holmes hit a well when it came time to produce actual lab results — but in drug development reality is slow to come by and can be explained away, as noted above. If Liz had chosen therapy instead of diagnostics she could have avoided the anklet!
But so what? How does this prevent working drugs from coming to market? Well, it is more than anything a beauty contest: the better story wins, unrestrained by reality which is difficulty to know anyway because clinical trials are what they are. So, an otherwise sucessful drug is out-competed for clinical trials sites which can only run a certain number of trials at a time, until it is eventually dropped in the absence of a bullshitting champion. Who knows how many good drugs now languish in order for the likes of Sarepta to prosper? And this not because of any failure of structure-based drug design compared to screening, but because its rise as a drug development method came about at the same time as the field became more and more financialized.
I would like to highlight the last paragraph as an indication of me changing my mind about something, which is a rare occurrence. Two years ago almost to the day I wrote that no one was hiding miracle cures, mostly as defence of the FDA requiring clinical trials instead of “real-world data” for commercial approval of drugs. I still hide behind the final sentiment, but not behind the statement that there are no drugs that exist now which may be truly world-changing whose development is being held back by clinical trial regulation. I worry now that this indeed may be the case. The next step in bringing them to patients should not, however, be to loosen the criteria for market approval, but for getting them into trials as quickly as possible. And good luck doing that without breaking the vicious cycle of health care-trial-health care costs.