Tech and biotech just don’t mix. At least that’s the conventional wisdom—which is, in part, why tech and biotech broke up in the early 2000s. Venture capital firms that housed both split their firm into two practices, or wound down one. And there was good reason for this; or rather, there really wasn’t any compelling reason for them to be together in the first place. Tech and biotech investors didn’t speak the same language and they didn’t need to. Biotechs had little use for debuggers; tech certainly didn’t need to know about DNA. More importantly, the nature of the investments themselves—capital required, risk/return profiles, modes of company-building and scaling—were all so different that knowledge and experience didn’t readily translate across tech and biotech.
But the world has since changed. Today’s life science companies are awash in data and program biology. Tech companies offer data pipelines, core infrastructure, and connective tissue across the entire healthcare value chain, from life sciences to care delivery. The well-worn playbooks are out of date, as the intersection of the worlds of biology, computer science, and engineering blurs lines across the entire healthcare industry. We call this new hybrid of tech + biotech simply “bio”. World, meet bio. Bio, eat world.
It’s still met with skepticism: Why are Silicon Valley VCs investing in biotech? But that’s actually the wrong question to ask. The real question should be, are old-guard tech or biotech firms the right investors to evaluate and support the next generation of companies emerging in this new bio era?
Why are Silicon Valley VCs investing in biotech? That’s actually the wrong question to ask. The real question should be, are old-guard tech or biotech firms the right investors to evaluate and support the next generation of companies?
Below are 16 enduring myths and misconceptions that still persist in traditional tech and biotech circles when it comes to bio—and why they’re outdated mindsets for this new world.
Yes, an entire generation of biotech companies has been built by VCs; and they have created real value for patients. But a new breed of bio entrepreneur has emerged at the intersections of biology, engineering, and computation. These founders are building their companies themselves, instead of handing over the reins to VCs to do so. Which means the role of the bio VC has to change too—to support rather than dictate, to catalyze rather than create.
Conventional biotech wisdom says that you need experienced drug hunters to run biotech companies. Yes, experience matters, but as technology increasingly transforms how we discover and develop drugs (including even the nature of what a medicine is), a deep understanding of how to wield powerful new technologies matters too. Leading faculty and recent PhDs in emerging, multidisciplinary fields like AI or biological engineering are, by definition, already the world’s experts in those areas. And in many cases, these scientific founders are actually better positioned to drive their own innovations forward than the industry veterans of previous wars.
Not usually. “Platform” is way overused in biotech, typically to describe what is actually deep expertise in a specific area (e.g., cancer or neurodegeneration). This is a departure from the standard definition of “platform” in tech, i.e., a base of technologies upon which other applications are developed. The new generation of bio companies are built on computational and engineering platforms that use proprietary technology to see something others can’t see, or do something others can’t do, like create detailed gene circuitry maps of cells, or engineer immune cells to attack cancer. In other words, a bio platform gives you a technology-enabled knowledge or capability edge, ideally both, and a foundation on which many future medicines can be built. This is a huge leap beyond most traditional bespoke biotech platforms that result in few, if any, new medicines (because discovery is really hard!).
Speaking of platforms, there are few that have drawn as much skepticism from traditional biotech as artificial intelligence. AI, like any tool, is certainly not a silver bullet. But it is uniquely well suited to help us connect dots in biology’s unimaginably complex network that we humans could not do otherwise. Some of AI‘s most renowned experts are focusing their efforts in healthcare; a swarm of new AI startups have emerged; and even established pharmaceutical companies consider themselves data science companies and are staffing up accordingly. We’re still in the early days, and while anticipation does sometimes crossover into hype, AI is already taking some of the guesswork out of clinical diagnosis and drug discovery.
Advances in biotech are subject to breathless headlines years before they have any meaningful impact on patients—where hope is plentiful but time is not. But technology innovation cycles in bio are accelerating, so that promise->potential->practice is now happening in record time. The cost of DNA sequencing dropped over 1000x in less than a decade, far faster than Moore’s Law (a phenomenon now known as “Flatley’s Law”)—and it’s now a standard tool for R&D and diagnostics. CRISPR for editing human cells, which emerged a scant five years ago, is already transforming how we discover, diagnose and even treat disease: 2019 saw the first human clinical trials in the US, with many more in the pipeline.
Traditional biotech drug discovery follows a well-worn route. Once a novel disease target is validated, it’s a “race to the molecule.” If you don’t get there first, then you’d better be best-in-class when you do. If you fail (and sadly many do), there’s not much that translates over from one effort to the next. But with engineered biology, we create generations of technology, like versions of software, that build on top of both the successes and limitations of previous iterations. A new generation of CAR T is being built on the foundations of the first. And the modular aspect of these medicines—i.e., ability to reuse and repurpose common components—means that new applications will be easier to build. The original CRISPR-Cas9 construct has already been remixed and reimagined (like CRISPR-CasX), giving us new flavors and functionality. So not only are innovation cycles accelerating, they’re compounding too.
Some argue that we’re not in a new era, and what we’re seeing is in fact just the continued upward progress of biotech. But when sufficiently dramatic, evolutions themselves evolve into revolutions, and steady progress gives way to sea changes. In the last few years we have entered the era of programmable medicines, where we can engineer living things to serve as gene and cell therapies. We’ve programmed digital therapeutics too. Things are moving quickly and the FDA is committed to modernization and reform in order to keep pace—even offering guidelines for medical AI that learns and evolves over time. Everything is on the table: entrepreneurs weary of the status quo are re-imagining the industry and scrambling time-honored business models. And bio has implications beyond human health. Engineered biology is transforming a broad range of industries from food to fashion—it has the potential to change everything. If this is “just” evolution, it’s a Cambrian explosion.
OK, this one’s partly true. Bio companies are a hybrid between traditional tech and biotech. They have wetlabs for the slippery, squishy stuff; drylabs for data and design. Building a bio company requires the ability to merge disciplines that have traditionally been siloed—engineering a multidisciplinary culture, which retains a tech mindset of relentless iteration that drives continuous improvement. As my partner Marc Andreessen says, “Moore’s Law was a goal, not a prediction.” The same way it did for the tech industry, infrastructure like shared lab spaces and AWS have already dramatically reduced the cost and time needed to get a bio startup off the ground.
When dealing with biology, we’re at a distinct disadvantage since we’re working in a system we didn’t design. But key advances have changed the game: platforms to see biology at unprecedented breadth and resolution (the DNA sequencer is an engineering marvel); sophisticated AI to systematically unravel biological complexity and industrialize discovery; and perhaps most importantly, biological engineering to program biological systems directly. All of this takes us into a world where scientific discovery can and should be engineered—and brings us even closer to purposeful design of biology.
It’s of paramount importance to make sure that any new medicine is safe and effective—all bets are off as soon as you put something inside the human body—and the 90% clinical trial failure rate is well known. But what the industry refers to as regulatory risk—that the FDA won’t approve your drug—is really a misnomer; it’s actually a stand-in for science risk—that your drug doesn’t work. In a world where it’s unclear if your target modulates disease and/or if your molecule sufficiently and exclusively hits that target, this kind of risk can feel like flipping a coin—you either get it right, or you don’t. But if the cause of disease is well known, as with genetic disease, an engineered solution is less binary. You can iterate your way to success, component by component—once we know how to deliver one gene to one cell type for a given disease, it’s easier to deliver a different gene to a different cell for the next one.
In healthcare, reimbursement is absolutely central to the go-to-market plan, in large part because the ultimate end-user (patient) is different than the payor (insurance companies, employers, governments). But “reimbursement risk” is a misnomer too; it confuses and conflates the rigorous process for securing reimbursement with the real risk, which is not being able to demonstrate a compelling value proposition in the first place. It’s extremely hard to justify a premium for a new therapy that’s not a meaningful improvement over existing standards of care. But as programmable medicines become more sophisticated, they offer the tantalizing potential for long-term cures for our most devastating diseases, saving lives and avoiding downstream costs—making reimbursement less a risk and more an obligation to fairly distribute the costs and benefits to all stakeholders.
It’s certainly true that it’s more expensive to develop a novel drug than to launch a software product. In traditional biotech, value-inflection is very much a step-function: you won’t know if your molecule works until it’s been tested in many patients, many years and millions of dollars from now. An engineering approach has the potential for smoother, more linear value creation. Take genome editing. You can tinker with a CRISPR system until it performs and dramatically reduce science risk by repairing disease-causing errors and starting with ex vivo (outside of the body) applications. Certain big effects may require smaller trials—a few dramatic cures go a long way. It’ll still take time and money, of course, but these bio companies have the potential to iterate and generate more proof (and value) along the way.
This is a strawman. While there’s a time and a place for speed and risk-taking agility, Silicon Valley has a long and enviable track record of doing the hard things too. It builds satellites and launches rockets. It uses technology to radically transform previously staid industries, from automobiles to music to movie-making. Some of these innovations serve as diversions, but others serve our national defense. Suggesting that Silicon Valley is incapable of the seriousness that healthcare deserves is an ill-informed argument at best, and disingenuous at worst.
Except that it wasn’t. The fact is that there weren’t any venture firms significantly invested in the company; those that took a look were unable to diligence the company and its technology. Outside of personal relationships, the company didn’t have a deep bench of directors or advisors from the tech or biotech worlds either. What made Theranos a “Silicon Valley company” was its zip code. Good biotech VCs—anywhere—do deep diligence, including site visits, background checks, data analysis, intellectual property audits, and even blinded sample-testing challenges. Questions and issues often arise in the diligence process, but the inability to conduct diligence in the first place is a bright red flag.
This conveniently ignores the fact that one of the first and arguably most successful biotech companies in history, Genentech, was born here. Genentech’s legacy is such that today South San Francisco and the surrounding Bay Area houses a thriving ecosystem of established biotech companies—and next-generation bio companies, too. And it’s not just startups. Along with a16z, a rich ecosystem of seed, early- and late-stage venture investors has emerged, focused on and specialized in bio. Like the bio founders they support, these investors are increasingly multidisciplinary, with computer scientists, biological engineers, MDs, former biopharma executives, and seasoned healthcare entrepreneurs all working side by side.
Should tech just stick to tech, and let biotech deal with biotech? Maybe, but it’s hard to stay in your lane when the roads are merging. The good news is that it’s a long winding road, traversing a wide world of opportunities to have a transformative impact on biology, technology and, most importantly, patients. Yes, some paths will be best suited for old-school biotech or traditional tech. Others will sit squarely within bio’s sweet spot. And then there will be those where tech, biotech, and bio come together to bridge from the old world… and build the new.