Skip to main content

Business and Markets

Andreessen Horowitz’s First Move in Biopharma

The well-known venture capital firm Andreessen Horowitz has recently set up a pool of money to invest in the biopharma space, which is a new area for them.You can read an interview with Vijay Pande here, who’s been brought in for this effort, and it talks about “computational biomedicine” and “cloud biology” as areas that they want to focus in. They’ve had quite a track record in technology startups; there are probably quite a few of us who wish that we’d helped bankroll the early days of Facebook and Airbnb, just to pick two. But there’s always a null-hypothesis question to ask about successful investors: was this expertise, or a run of luck? (To be sure, there were plenty of people who didn’t invest in these companies). If this was expertise, then, is it transferrable to another sort of industry?

And biopharma startups are another sort of industry, believe it. The folks running Airbnb or Über probably feel that they have a lot of regulatory headaches, since many areas that they’re trying to operate in are unhappy about their business models. Compare that to, say, mobile game development, where you code and test that sucker on your own and push it out onto the market. That’s the sort of thing that appeals to the classic Silicon Valley mindset – a fast, nimble, hard-working team of highly competent people, whose rate-limiting steps are how quickly they can think and how little they can sleep. The sharing-economy companies certainly came from that same greenhouse atmosphere, but dealing with the regulatory demands of many jurisdictions must have felt like running into a hip-deep steam of pancake syrup.

But that is nothing compared to what awaits you in drug development. In this business, you work for years before you can have the tiniest hope of ever selling anything to anyone. And before you can do that, you have to (by Silicon Valley standards) abjectly crawl before the regulatory agencies in the US and every other part of the world you want to sell in. Even to get the chance to abase yourself in this fashion, you have to generate a mountain of carefully gathered and curated data, in which every part of every step must be done just so or the whole thing’s invalid, go back and start again and do it right this time. The legal and regulatory pressure is, by Valley standards, otherworldly. If your new app doesn’t do what it’s supposed to do well enough, it will bomb on Google Play and you’ll have a lot of unhappy customers. If your new drug doesn’t do what it’s supposed to do, you will get off lightly by having it rejected by the FDA so that you never sell it at all. The alternative is to have something untoward show up after it hits the market, in which case you will be hit with so many lawsuits that you’ll think it’s snowing. Law firms will take out cable TV and radio ads, desperately searching to find more people who can sue you.

Arching over all these difficulties is that factor that I’ve spoken of many times as the “Andy Grove fallacy“. That link will take you to longer thoughts on this subject, but the short version is that software and hardware, for all their complexity, are human-built things, designed and annotated by humans. Terence was right: homo sum, humani nihil a me alienum puto (“I am human, thus nothing human is alien to me”). The chemistry and biology of a living organism, though, taken in detail, is as alien as anything we’ve ever dealt with. A few billion years of blind evolutionary tinkering gives you a mass of insane, irrational, tangled interwoven systems with no documentation in sight. Crazy new things keep getting uncovered all the time (siRNA! Who knew!), and the number and scope of the crucial things we have no clue about is just another one of the things that we have no clue about. If your worldview has been formed by using human tools to make human-designed applications for humans, you are in for quite a shock.

So with all these things in mind, I’ve been interested to see what sorts of companies Andresseen Horowitz is going to invest in, and how it’ll work out. The interview linked in the first paragraph suggests that they’re going to avoid actual drug discovery and development, probably for just the reasons described, and look for areas that are more specifically tied to the technology that they know more about. That’s probably a wise move, but there are other ways to sidestep those problems, unfortunately. Reports are that the firm is funding an outfit called Nootrobox (dang, that is an annoying name), which will sell nootropics to the public. Put crudely, they’re flogging smart pills in the nutritional supplement market. This isn’t actually the first investment in their new biotech pool, though – this one was in the works before that got organized, but that link above says that if the timing had been different, that’s where the money would have come from.

It is, unfortunately just the sort of thing I might have expected. The nutraceutical area is very lightly regulated indeed compared to actual drugs, as we all know, so that sidesteps one of the big problems mentioned above. And it’s not like these folks are going to do all that much actual CNS research to figure out mechanisms of action, either. Central nervous system work, as black boxes go in drug discovery, is matte, flat, stealth-bomber black, and many are the billions of dollars that have gone into it never to be seen again. So I don’t expect Nootrobox to do anything of the kind. No, they’re apparently selling “Rise” pills to improve your memory, and “Sprint” pills to give you more energy. This is the sort of thing I see advertised on the subway, next to the offers to learn how to be an HVAC technician – can “boost your immune system” be far behind?

I suspect that part of the pitch for this investment was a lot of stuff about biohacking, which would be an attempt to appeal to the Valley mindset. Take control, push your own buttons, don’t just take what Nature has given you, all that sort of thing. Applying the hacker metaphor to the brain, though, is wildly premature, since we have a pretty good idea of how computers work, since we built them, but we have very little idea about how the brain does what it does at all. So taken together, I have to say that I’m disappointed (but not all that surprised) to hear about this venture.

That, though, isn’t its point. Andreessen Horowitz is a venture firm, so the first question they’ve been asking is whether this idea will be able to deliver a return. From that perspective, I understand the call completely. I think that such a company has a good chance of being very profitable indeed. It’ll have cachet, given its background, and plenty of free publicity. I don’t doubt that its products will sell at a premium – that’s how you’ll know that this is really the good stuff, as opposed to those fly-by-nights who advertise on the subway. And there’s plenty of demand for its products, which don’t really even have to work. No, I think that Nootrobox could do quite well. It’s a pity.

Update: David Shaywitz has thoughts here on the firm’s entire approach to this sector, well worth reading.

25 comments on “Andreessen Horowitz’s First Move in Biopharma”

  1. Peter Kenny says:

    Perhaps an appropriate moment at which to consider the link between ligand efficiency and homeopathy?

    1. tally ho says:

      As in the asymptotic limit of ligand efficiency equals placebo – full circle back to Deepak Chopra, mind over matter. Pass the cannabis.

  2. anonymous says:

    You’re kidding me, right? They promised us flying cars, and all we got was 140 characters…and now a bunch of computer nerds and aspiring MD/PhD students selling us expensive herbs? Unbelievable.

    1. Kent Kemmish says:

      It sounds like you need to see my slide deck 😉

  3. johnnyboy says:

    Remember Neurochem, the biotech with the failed anti-amyloid drug for Alzheimer’s ? After the failure they marketed the drug as a nutritional supplement, “Vivimind”, which was supposed to aid memory. The ad campaign was massive, and so was the failure of the product, which they later offloaded to a health supplement company. Of course this is in Canada, where people are generally a bit more reasonable than Silicon Valley types – but it could still be a cautionary tale. Bullshit only goes so far, before it gets overtaken by other bullshit.

  4. Anonymous says:

    It should be evident to all that biotech and Pharma are not nearly as prone to disruption as the advertising business (disrupted by Google and Facebook) and the taxi business (disrupted by Uber), and travel agency business. The hotel business is likely to hold its own against Airbnb and others, and airlines have held their own. The media empires are transforming as well, and many will suffer at the hands of disruption.
    Pharma and biotech are not likely to be disrupted anytime soon, for all the reasons Derek mentions. The process of discovering and gaining approval for drugs is an incredibly heterogenous and regulated affair. It’s likely to be one of the last to fall, and more likely to be incrementally improved the disrupted.
    Academic computational science has done very little to help with the real problems encountered when discovering a new compound or biologic, and the experimentalists are continuously working to improve both their efficiency and transltability of testing. Let’s hope for a steady and incremental increase in the probability of success of drug discovery and development, because the standard of care bar is steadily rising and the opportunity spce is therefor squeezing. We have to be better than we were in the 90’s and 2000’s and 2010’s for the business model to remain viable. I believe we will be, but it’s not likely to be such a big jump that we herald it as Silicon Valley style disruption.

    1. Kevin Parker says:

      I don’t know…If IBM, Enlitic, Numenta, Transcriptic, Emerald, Organovo, Emulate, and ResearchKit turn out to be effective – if only a little – disruption could be knocking.

  5. Mark Thorson says:

    Who knew a protein from jellyfish could improve your memory?

  6. Dr. Manhattan says:

    “cloud biology”
    Excellent! That way they can be well prepared when the FDA rains on their compounds.

    Once again, as Derek points out, the Andy Grove fallacy. All of your disease problems will be solved by the smartest guys in the room. If we just aggregate enough data, the answer will drop out.

    We built computers, we know the fundamentals of who they work. Biology? We have a very long way to go at even the cell level, never mind the organ and whole being level.

  7. bad wolf says:

    I can think of at least one reason they would target the “5-hour energy drink” market–the billions made by the founder.

  8. Hap says:

    It sounds a lot like buzzword salad to me, and you only serve that when you lack anything more substantial on the menu, or when your customers don’t care or won’t notice.

    The other problem with a lot of the disruption is that it relies on the concept that if you don’t have to pay for the costs of the infrastructure you depend on (taxes, regulation, etc.), you can sell stuff a lot cheaper, or at least take a larger proportion of the money for other people’s goods and services for yourself. That doesn’t seem like a sustainable business model, and even less so in a business where the costs of a lack of regulation can get pretty high.

  9. pete says:

    Forget the fancy mission statement this is simple business — same as it ever was for many “world-changing” SV ventures.

    I’m sure they’ll make some dough, too, what with all the well paid health conscious young’uns doing long hours and worrying about encroaching mental & physical bald spots.

  10. cancer_man says:

    Andy Grove will be shown to be correct – the question is when. (I think April 3, 2019 but of course I could be off a day or two.)

  11. swattie91 says:

    In Silicon Valley the mantra is “Fail fast, fail often”. The second part is easy in drug discovery. It’s the first part that’s the challenge.

  12. angrygecko says:

    This would be the same Andreessen Horowitz that put money into uBeam? For those not following the Si world, uBeam is planning to get rich by outfitting Starbucks with multiple 140dB ultrasonic transmitters that will be used to wirelessly charge your cell phone. Never mind little things like inverse square law, freaked out animals and really bad end to end efficiency (Watts go in, milliWatts come out).

    Their buzzwords-on-parade press releases pegged my bullshit detector, but judge for your self: and the contrary

  13. Argon says:

    The Forbes article describes ‘cloud biology’ as highly automated labs made available for startups so that they can access experimental hardware without spending as much on capex. I can perhaps see this with sequencing projects but I’m having trouble imagining how this would work for many enzyme and cell assays. Is that really different from using a CRO? Do the words ‘cloud’ and ‘automation’ make it sexy?

  14. Biotechinvestor says:

    There is a bubble in the tech sector. Nowadays, there are more venture capitalists prospecting around Stanford University than there are entrepreneurs. This, among other reasons, is how we get unicorns: start-ups and pre-IPO companies with billion $ valuations.

    I’ve seen this before in the 1999-2001 bubble. One of the collateral effects is that busines concepts in the tech sector push themselves into pharma and biotech, and business development people from the tech sector start dabbling in biotech.

    I’m seeing this again now. Theranos is an example. I am glad this blog presented its views, which reflect my own skepticism. When I see a diagnostic company like this being reported in the pages of Fortune magazine and Forbes with more hype than the life science industry news, I knew something was wrong. I am seeing business development people from the tech sector dabbling in biotech again too. These are business development people that weren’t cutting it in tech. Now we have to deal with them in our sector, which is no fun either. It’s also leading to unrealistic valuations coming from investors who have no expertise or background in this industry.

    It’s not going to end well. The sooner this happens, the better. We need a correction to flush out the hype and encourage focus on real, good, solid science and its application.

  15. Patrick says:

    My brain broke and I had to stop reading at the phrase “cloud biology”. I mean… It sounds like they’re studying how a cumulonimbus poops.

  16. Norwood says:

    Proteins that fold during short simulations, do so in a manner that poorly represents an equilibrated reality (citation Fersht’s criticism around 2002 in PNAS of Pande’s work). Probably a valuable analogy to draw in the investment world.

  17. David Cockburn says:

    By way of contrast to this naivety some of the ideas coming out of Google Life Science sound promising.

  18. James says:

    There’s a slide deck from A16Z going around called, “Software is Eating Biology”. I kid you not.

    It’s difficult to tell from the slides how exactly that meal is going to take place. From what I gather they believe it’s 1) “Quantified Self” inspired sensors and preventative medicine, which makes sense. 2) Mining genomic data for potential cancer therapies (low bar for FDA approval) and then there’s the mysterious 3) Cloud Biology.

    I don’t see how any of these 3 cases justifies the hubris.

Comments are closed.