on deeptech side-quests
distractions vs leverage-amplifiers
There are two ways people will tell you to think about deeptech side-quests.
One way is as a distraction. You set out to build a company with a grand vision of bringing into existence a new capability for humanity that solves a huge problem some OOM better than any current solution, radically transforming the market and improving people’s lives. If you raised money, your investors (ostensibly) gave it to you to “vanquish all adult mosquitoes” and not “make a 7% better Mortien”. You need to be laser-focused on this vision, and the specific customer persona this vision serves. As a founder, you only have 14-16hrs in a day and your time is your biggest asset. Spending time for a side-quest’s sales distracts you from talking to your scientist for the main-quest. Spreading yourself too thin increases risk for the company succeeding on its reason to exist.
The other way to think about side-quests is as a leverage-amplifier. The side-quest can be a forcing function for iterative refinement of your capability wielding, it can provide battle-tested points for your current technology chops, it can beam out strong signals that you mean business, and ultimately it can grow a tech tree that bakes-in powerful optionality.
A sign that your side-quests are leverage-amplifiers is when the moonshot capability you are launching towards has multiple step functions, nested mini capabilities, with each one unlocking a new market by solving a customer’s pain point. This new market may be much smaller or quite radically different from your true market, but as long as its key is on a branch of your tech-tree, the work of planting roots and budding up your stem has already been done, and generally means it can be pursued with leverage.
The leverage from a forcing function
A forcing function is intentionally induced pressure to produce a production-grade outcome. Forcing functions can come in many flavours but almost all force you to consolidate your newly unlocked mini-capability, by pushing you to meet market forces and consumer demands. This inevitably means iterating quicker, leading to continuous refinement on your technology and giving you a better shot at your Grand Capability.
The leverage from battle-tested points
Putting your technology in the arena means it garners battle-tested points, undoubtable evidence that you technology works as someone is willing to pay for it. Battle-tested points are generally key to moving up a tier for the next set of partners, contracts and grants, giving you a better compass and providing precious rocket fuel on your trajectory.
The leverage from the signalling of business intent
The ability of a deeptech founding team to sell is a crucial signal that separates the science projects from the generational successful companies. Many deeptech investors tend to be burned by science project founders who do fantastic research but never translate that to massively profitable companies. A capability only generates real-world value such if refined, packaged, and positioned in a way that it can be exercised to solve problems in the real world, otherwise it is pure science, wonderful and much-needed, but not investable.
The leverage from optionality-generation
A deeptech company that has cultivated a stacked tech tree where each branch can solve a valuable problem that customers are willing to pay for is a company that has tremendous optionality. Such a tech tree effectively de-risks the entire enterprise of shooting for the Grand Capability, the company is no longer beholden to the whims of investors and any one branch can fail for tech/market reasons with the company continuing to survive. Indeed, maybe the envisioned grand capability turns out to be lacklustre compared to the so-thought mini-capability that turns out to be a billion dollar value-add. Mining asteroids is cool but whole-earth imaging is pretty cool too.
Dognosis bounding on side-quests
Dognosis is working towards detecting all diseases from breath by letting dogs teach AI how to smell. To do this, we build wearable canine brain-computer interfaces (BCI) to measure dog brain activity non-invasively. While our intention to build the BCI is to train ML models to understand how dogs understand the scent of cancer, it has lots of other use-cases as well. Veterinarians can use it to localize epileptic foci and perform curative seizure treatment animal pharma can develop better calming semiochemicals, university labs can do more interesting experiments, and security groups can improve on dog selection, training, and deployment.
All of these side-quest applications force our BCI to be capturing canine brain data as accurately as possible, gives us more neuro data to build better models, allows us to hone and showcase our ability to sell, and creates multiple different pathways to survival in case it takes longer than expected to crack the olfactory code.
To us, it’s a no-brainer that these side quests are worth pursuing and that seriously doing so increases the odds of us winning. This is likely true for many other deeptech companies. Be wary of main-quest-maxxing.