Founder Mode: Wildfire
March 11, 2025
Keith Rabois drilled a key concept into me: “first, distill the root cause; then, establish founder-mode focus; else repeat.”
While he uses his own phrasing to describe this operational philosophy, it was not always easy to prepare for our board meetings. Even if you tried to anticipate what he’d narrow in at, he had better aperture and tenure.
As it relates to my interest with Red Sky Summit, I do think all ideas for wildfire distills the adoption rate, or inversely, penetration rate. No matter the solution or community proposal, the rate limiting step is how quickly the world adopts the solution partially, or universally.
This has led me to ask my version of a wildfire trolly problem.
Thought Experiment
When I speak to retired fire chiefs, I like to ask them the thought experiment: how would you protect your home at all costs? Here are the additional constraints, you are housing the painting of Washington Crossing the Delaware in your residence, and you have access to a Type 1 Fire Engine with a 4 person crew and its single water tank stationed for you 24/7 for an entire week, what would you do to protect the property at all costs?
I will later publish the results of my survey, but the general responses thus far theme are around home hardening, taking away sources of fuel, and using water effectively.
The challenge with prescribing status quo wildfire solutions is that it’s a bit like prescribing suggestions to improve demographic health. You can run a campaign to “eat an apple a day,” optimize daylight saving and school and work start times to promote 9 hours buffer of sleep for all, or sell Pelotons and Apple Watches promoting fitness, but at the end of the day, you have to get everyone enthusiastic about things like vaccination as well as do all the above to hopefully move the needle for epidemiologic health.
Similarly, you need everyone to participate in good wildfire hygiene--all the annoying, painful, and fun things--to be in a better spot collectively. “Everyone” includes governments needing to step up and manage wildfire fuel risk; it includes communities needing to make this part of education, awareness, or tagline (e.g. stop, drop, and roll); and it includes founders needing to provide desirable product solutions.
Principles Thinking
Before revealing my own answer to my Washington Delaware problem, I first want to distill fire suppression into two key principles:
Time to Response - no matter the situation, your time to respond to a fire is critical. Smaller fires are easier to put out than larger fires, and this should be measured in seconds, not only in minutes. It applies to urban fires as well as wildfires. The faster you deploy your solution, the easier it is to manage the technicalities of a growing fire.
Relative Humidity - the easiest way for me to think about this RH concept is to wonder why a paper cup filled with water is able to not combust immediately when placed above a campfire. It involves concepts of conduction vs. combustion and simple grade school physics. Assuming you can’t remove oxygen and you can’t control fuel supply, the next best thing to control for is the Relative Humidity of your target environment. Simply, it is hard to start a campfire when it’s pouring rain, though possible with lightning.
While these concepts are not complete, it’s a good starting point for me (h/t to Prof. Michael Gollner, Christopher Anthony, Dave Winnacker, and Mario Trevino for responding to my cold email and providing tutelage).
Solutions of the Future
No matter the flame retardant paint or material, metal box or metal covering, or oxygen+fuel controlled environment, we mostly have a pile of wood that we call home. Even if you were to build your home entirely of steel, you need to eventually handle the thermodynamics of keeping Washington Delaware in a happy state. Water cooling a steel home is the same principle as water cooling a nuclear reactor.
My answer: efficient use of water with instantaneous response time to targeted threats with the ability to maintain.
- ckl
Thanks to o3-mini-high and Grok 3 DeepSearch for providing critique. Here is a version of the post with editorial efforts by AI.