I recommend reading Richard Zeckhauser’s paper Investing in the Unknown and Unknowable to everyone working in venture.

Founders should read it too - many of the messages are highly relevant to them, especially given the rapid pace of progress in AI now.

For context, the paper focuses on discussing three types of investment scenarios that can be distinguished based on how much we know about both what can happen and how likely it is to happen.

The three scenarios are:

1. Known outcomes with known probabilities (risk) - e.g. rolling a fair die: You know exactly which outcomes can occur (1 through 6) and their exact probabilities (1/6 each). Everything is quantifiable, enabling precise calculation of expected values, risk measurement, and standard cost-benefit analyses.
2. Known outcomes with uncertain probabilities (uncertainty) - e.g. launching a product into a somewhat familiar market: You understand the possible outcomes, but the exact likelihoods are fuzzy or hard to pin down. You can still try to estimate probabilities, but you don’t know them precisely.
3. Unknown or Unknowable Outcomes (UU) - e.g. pioneering a new technology with no real precedents or when entering entirely new markets: You don’t even know the full range of possible outcomes - let alone the probabilities. Surprises (both good and bad) can appear out of nowhere. These “UU” situations often scare off typical investors, but they can offer massive rewards if you figure out unique insights or ways to reduce the fog.

Today, most venture investing and startup-building - outside perhaps 'true' deep tech - falls into category 2: known outcomes with uncertain probabilities. Very few venture investments are truly in startups with mostly unknown outcomes and possibilities.

At focal, most startups we see these days (we invest in startups selling AI/software solutions to businesses, many of them fairly technical) go after known markets with new yet fairly well understood and accessible (AI) technology.

There’s a good reason for that: Given the platform shift, there’s an incredible window of opportunity to disrupt almost every market with AI right now (hello vertical AI). So why not take advantage of that?

At the same time, this means that we (and almost every other venture firm) see a lot of very similar AI startups going after the same opportunities.

For it to be truly a “UU” (unknown and unknowable) opportunity, the following needs to be true:

• You can’t fully characterize the set of possible outcomes. There may be surprises you cannot even conceive of right now - entirely new breakthroughs.
• No useful “comps” or historical analogs exist (or they’re extremely weak). This often applies when you’re creating a brand-new product category or pioneering a radically novel technology.
• This means you cannot confidently list potential outcomes, rendering even rough probability estimates nearly impossible.

That is rarely the case for most startups we see.

Now, why should founders and investors go after scenarios that are unknown and unknowable given all the uncertainty that comes with it when there’s so much potential for disruption in known markets?

Because it’s where the opportunities are biggest due to:

• Fear and aversion: Most (investors and entrepreneurs) prefer situations where outcomes can at least be estimated or forecasted.. Almost all shy away from “we don’t even know what might happen” scenarios, meaning fewer competitors go after those markets or technologies.
• Information advantage: Since nobody possesses reliable data or high confidence, anyone who uncovers genuine insight - such as a new technique or hidden demand - can gain an outsized advantage.. If you develop knowledge faster than everyone else, you exploit the unknown, effectively “naming your own terms.”
• High variance, fat tails: In UU environments, the range of outcomes can include extreme positives. While some outcomes could be disastrous, others might surpass conventional boundaries. This high variance often yields disproportionate upside for the successful few.
• Mispriced potential: Uncharted, UU territory tends to be undervalued or overlooked because traditional valuation methods don’t apply well. People can’t easily run spreadsheets or point to comparable deals, so risk-averse decision-makers often pass.

On the contrary, if you don’t embrace uncertainty and make big, bold bets, you’re likely going after very fierce execution plays from almost the start - which can be won but come with a whole other set of challenges and $ requirements.

We try to stay away from those as a smaller firm that can’t give you almost unlimited capital. We strongly prefer UU scenarios and call it “opportunity in ambiguity” / strive to take on as much ambiguity as possible given the power law in venture.

Which - to be fair - is much easier said than done in the world of software investing we live in.

What does this all mean for founders?

As a founder, if you decide to take on venture funding, you’ve chosen the path of “go big or go home”.

Meaning that ideally, you operate under the scenario where the potential upside is biggest (unless you want to bet it all on execution): Unknown and unknowable.

This implies that you should seek out the opportunity where you have the most contrarian insight and / or a novel (technology) breakthrough. The more out there it feels and the more others hesitate, the more you may be onto something.

Hesitation includes investors as well - meaning investors 'not getting it' isn’t always a bad thing. A common mistake, especially early on, is interpreting investor rejections as a universal sign that 'no one wants this'. At the same time, blaming investors for 'not getting it' shouldn't become an excuse you use if your pitch isn’t resonating.

That said, beyond timing (e.g., being too early) and technology risks, the biggest danger in UU scenarios is that a hidden showstopper exists - something others recognize but you overlook due to your optimism (vs having a true insight).

However, if you have high confidence in your insight / breakthrough, then the best way to find out how big the opportunity in ambiguity is, is to give it all you got and embrace the uncertainty.

While execution plays seem attractive right now given the window of (AI) opportunity, the biggest asymmetric upside will continue to be found where no one else is looking.

It’s often a longer, harder, and less certain path - but the potential rewards should more than compensate for that.

So if you’re confident in your contrarian insight / technology breakthrough, then embrace the unknown and unknowable as much as you can. It will be worth it.

If you’re thinking about / have started building an AI native software startup, reach out.

We often get involved before you even write your first line of code, catalyze your first round with up to $750K and specialize in helping you get off the ground better and faster, alongside 200+ GTM leaders at some of the fastest growing software startups across North America.

There’s truly no “too early” for us at focal.