Curiosity: a new kind of organization to drive real human enhancement

All my life I dreamed about using biotech for life extension and true, significant human enhancement. Why? I’m curious about what happens next. What do we find in the vastness of space and the depths of our own consciousness? Plus I am a transhumanist and believe the idea of re-engineering ourselves to have more intelligence, happiness, energy, health and whatever else we want is just fucking cool.

This whole direction is not being worked on nearly as much as I expected it to be. Sci-fi promised us eternal life, and we can’t even get a metformin clinical trial properly funded.

To help make this dream a reality, we are announcing Curiosity — a venture builder which builds (1) companies that gather biomedical data direct from consumers at large scale (2) companies that use this data to develop new biotech IP (3) the tools, culture and incentives that helps them exchange this data with each other.

The backstory

When I was 18, I met and had a call with Elon Musk at Cornell when in 2004 he came to recruit engineers for SpaceX.

I felt “wow, so this is how you do this — first learn to build companies and make money; then obsessively pursue your dream, no matter how audacious, by identifying the core bottleneck for what you want and just doing it.”

The core bottleneck to improving our bodies on a myriad of vectors is a lack of insights, a lack of first-principles understanding of how biology works.

Systems biology, and human biology in particular, is notoriously complex. And the key bottleneck to understanding how human biology works is gathering a lot of the right kind of data.

OK, what is “the right kind of data?” And what exactly do we mean by “insights” anyway?

To answer these questions, we have to start with “why?” — i.e. “what do we want?” For this we do have a clear and simple answer — we want to understand how a particular phenotype (i.e. feature of an organism) comes about and then engineer desirable phenotypes with the right technologies. Health, intelligence, high energy, longevity, athletic ability, libido, good mood, aesthetics — take your pick.

Phenotypes are an emergent property of the complex system of our bodies — primarily of the interaction between our genetics and our environment and the resulting impact on our biochemistry and cellular systems. So the data we want is ultimately:

1. detailed phenotypes of interest — e.g. medical records or wearables data on how active someone is or how well they sleep
2. detailed info on genotype/biochemistry/cell biology which is concurrent with the phenotype

Then we should be able to explain (1) from (2) using modern ML, as long as we have enough data. And then we can attempt to perturb our biochemistry to generate desired phenotypes, and get on a rapid iteration flywheel of improvements. Which there will be a ridiculous amount of capital available for once the first results are shown.

And that flywheel is the holy grail of biotechnology, and the dream of science fiction. It will lead us to life extension, health and happiness — the Homo Deus that Harari wrote about.

This will happen. We have a plan to do it. And I cannot think of anything I could do with the next few decades of my life that would be of greater impact and importance.

An important note on privacy and ethics

The beauty of this kind of data is that the most valuable insights — how to extend life, cure disease, enhance capabilities — can be worked on without any knowledge of who the person in question actually is. We want insights about humans and biology in general.

That being said, I realized in the last year that there are many potentially exploitative business models driven by very personal biomedical data out there. We want nothing to do with them, because that goes towards the dystopian end of science fiction where our bodies are constantly invaded by governments and corporations. But the challenge is that if there are incentives to lean that way (and there are), these applications will happen. Incentives are destiny.

So how do we get the many significant benefits of this kind of data without falling down the incentive trap? I do not know. This requires a lot of careful thought and I will dig into it in it’s own dedicated future post.

Bridging the gap between consumer health and biotech: the time is now

What is happening right now is that all of the necessary pieces required to gather and interpret vast amounts of biological data directly from consumers, and using them for deep biotech research to improve health are finally in place.

Specifically:

• the US government has mandated medical records data interoperability by 2023. What this means is that any consumer can, in a simple web interface, say “I consent to give service X all my medical records” and all medical providers are legally obligated to give this data through an API.
• over the last decade, social/mobile companies have habituated consumers to easily provide consents for any kind of data, if consumers get some kind of a service they value (e.g. personalized insights or cost savings or just social competitions) in return.
• easy-to-use wearables from sleep trackers to CGMs to cellphones are generating streams of phenotypic data which consumers find easy to collect and give away, because it is a matter of one checkbox.
• various assays (tests) can turn blood into an incredible amount of genomic, transcriptomic, proteomic, metabolomic data at an ever-lower cost. One of our companies, Multiomic Health aims to generate >500gb of biochemical data from a single biobanked sample.
• various AI/ML technologies and associated infrastructure are getting ever more powerful and ever cheaper and ever more capable of finding insights given the right kind of data, which companies like Recursion and Insitro are starting to exploit.

Our view is that it is obvious what must be built to substantially accelerate progress in the space:

1. viral, engaging consumer services that deliver value and gather biomedical (phenotypes + samples) data cost-effectively and at vast scale and without intermediary gatekeepers
2. biotech companies that turn this data into extremely valuable insights about our biology and convert these insights into diagnostic and therapeutic products
3. a platform with aligned incentives and a culture where all participants will share data with each other — this is very complex and requires real organizational innovation btw

What we are building at Curiosity and why

The Curiosity Venture Builder builds (1) companies that gather large-scale biomedical data directly from consumers (2) companies that use such data to develop biotech IP (3) a data platform, financial structure, legal framework and organizational culture that incentivizes companies to share this data with each other.

To have the level of impact we want, it is necessary to build a number of companies that are both highly independent and coordinated around a single vision and set of cultural values.

There have to be many companies because that way many hypotheses can be tested through an evolutionary process where the ones that do not work are killed off as fast as possible. Noubar Afeyan and Flagship Pioneering have built such a process and created a >$100bn portfolio through it — here is a great video where he talks about it and their process has been one of our primary inspirations. It should generally be noted that the venture builder model has been extremely successful in biotech.

The companies must be independent because building engaging consumer apps and building biotech research programs are very different endeavors and need very different cultures and management. This is one of our core insights — it is just not possible to merge these in a single organization or conglomerate.

The companies have to be coordinated because purely free-market-arms-length data sharing agreements in biology tend not to work. We have done a lot of research on data sharing agreement attempts. Usually they run up against people being possessive of their data & excessively scared of competition (even though this whole space is a vast blue ocean of opportunity); or the fact that company business objectives are just too different; or that the data is too heterogeneous; or that the revenue timelines between data-generators and biotech-IP-generators are too different.

We are fixing this by building in cultural and financial incentives around data sharing from the very beginning, and just making it an expectation required of anyone we work with from the get-go.

How we are doing this

We build companies that are independent, with their own dedicated CEOs, their own third-party investors, and venture-scale (i.e. can build to >$5bn in market cap as an independent business). Coordination and data deals happen via aligned cultural and financial incentives.

We believe that entrepreneurship is not magic or a lottery, it is a profession. There are specific things that need to be done to ensure a high degree of success and we know how to do them. They come down to:

• a small founding team that gets along well, has a complementary skill set, and a shared big vision which drives energy and passion
• initial hypotheses on the path to the vision that are iterated and killed off as rapidly as possible through deliberate evolutionary bottlenecks, most of all early market testing and customer development
• a network of contacts which can easily be polled to recruit teams, test hypotheses and raise capital — all as fast as possible
• discipline to constantly challenge assumptions and pivot as needed until product-market fit is found and the business shifts to exploiting it (which is a much more established professional discipline than the creation of companies up to this stage)
• a cost-effective, smart approach to capital — spend as little as possible before PMF is found, then deploy a lot of capital to scale

We know how to find, inspire and coach high-potential entrepreneurs at scale

The operator CEOs we co-found companies with are absolutely key to achieving our vision, and we think of them as our valued customers. We look for people who either have a science/medical background and want to become entrepreneurs; or a mainstream consumer tech background who want to go into bio (there’s a lot of those now and it is hard to make that transition).

Every entrepreneur initially comes on-board as an entrepreneur-in-residence (EIR). We look for people who:

• are mid-career scientists/doctors who want to become entrepreneurs; or mid-career consumer tech experts who want to shift into biotech
• are underrepresented talent without access to all the right networks and to whom we can bring enough value to justify a co-founder stake for us
• are not married to a specific idea and are ready to rigorously assess ideas, pivot and search until the right business model is discovered
• are deeply committed to the mission of gathering and exchanging biomedical data to advance biotech research, no matter what form that process takes

Star Silicon Valley entrepreneurs and investors think there are very few people like that, and they all can succeed without much help. This is survivorship bias. Of course the person who meets a top VC doesn’t need help meeting a top VC.

One of our core insights is that there are very, very many people who can be great CEOs in the world, if they get support from a group like Curiosity, and who will not do it otherwise. And our team has the global reach, network and vision to get people like that to join.

For those who aren’t yet superstar entrepreneurs, being an EIR with Curiosity is a dream job. You get to find a business you can lead together with a group of amazingly talented people. You have both a founder-level stake in the business and a salary and backstop that if the initial idea does not work out, we will find a better one together. You also become a partner in Curiosity and an equity owner in all our portfolio present and future, and can advise other amazing CEOs with your own expertise and earn even more equity that way.

I wish I had a job like this when I was starting as an entrepreneur.

We’re already building traction.

In the past 18 months we co-founded four companies:

• Multiomic Health is developing personalized medicine for metabolic disease using multiomic data. It is led by life sciences veteran Robert Thong and Forbes 30-under-30 entrepreneur and popular author on health Slaven Stekovic, and funded by many great investors including Fifty Years and Formic Ventures.
• Novami Health is a digital medicine service that will bring the stellar concierge medicine experience I had with Peter Attia and other top doctors to the mass market. It is led by top B2C growth exec Julia Seregina and a star NYC concierge doctor who will remain nameless for a little longer, and is funded by HOF Ventures.
• Pheno is working on a way to use deep medical data to analyze risk in the life insurance space where a lot of biology-based risk is being underwritten based on poor proxy metrics like BMI. It is led by experienced R&D/data exec Vadim Pinskiy and funded by Tim Draper, Buckley Ventures, Metaplanet and others.
• Nexus is a clinical trial platform that enables small physician practices, and their patients, participate in clinical trials. It is led by tech entrepreneur and my former Chief of Staff Peter Boyko.

We also shut one company down — which is an important discipline. And our view is generally that at early stages, pivoting, shutting down, and exploring are what entrepreneurs must do. Curiosity itself arose from a pivot of a bio/nano-tech company I was building with Eric Drexler; and got funded by Jaan Tallinn, Naval Ravikant, Luke Nosek, Mike Maples/Floodgate, Village Global, Formic Ventures and others. I am really grateful to these awesome investors who bet on my passion.

Aside from myself, the venture builder is led by the very talented Corrina Kane who heads operations, as well as a brilliant data science exec from one of the biggest pharma companies. Three of our closest advisors are: the accomplished venture capital investor and former Bayer Leaps exec Alasdair Thong; my close friend and mentor Ed Baker, one of the world’s leading thinkers on consumer growth; and successful biotech entrepreneur Alok Tayi who is building ‘every cure for every community’ at Vibe Bio.

In the next several years we plan to launch 15-20 companies of the kind I describe above. We have created a detailed framework (which I will write about in a future post at length) to launch them extremely cost-effectively.

We want all of them to be venture-scale (i.e. >$5bn in market cap potential), and where the business model is centered around multi-sided network effects in biomedical data of the kind A16Z talks about here. And build the tools for them to begin exchanging data with each other, powering new insights and huge new industries in the coming decades.

The coolness and creativity and impact and sheer audacity of what we are building inspires me more than anything I have ever done. And we are all so committed to this that everyone’s vesting in Curiosity including mine is 10 years long.

And that long timeframe is important. I’ve spoken with enough people from Third Rock and Flagship to understand that what we are trying to do is very difficult. Still, attacking a difficult problem for a very long time usually leads to interesting unpredictable results. And I choose to focus on this particular difficult problem for the next few decades because well… what the fuck could be more important than making us all healthy, happy, and living as long as we like?

The rest… is the futurе.

If you like what we’re doing, get in touch.