Vision defined

Founders get confused about vision. I think it's because they read too much HackerNews and too many blog posts from people who saw stages but not zero to IPO, and somewhere along the way the word "vision" got tangled up with the word "dream." They're not the same thing. They're not even close to the same thing. And confusing them is one of the major ways founders end up lost.

The best way I've found to explain this is fog of war in a video game. AOEII is the perfect example. But you know how fog of war works: You start the game and you can see a little bit around you. Maybe a small radius. Everything else is black, or clouded, or just not there yet. You don't know what's out there. You have to walk into it. You gather some resources, you venture out, you discover something, and now that part of the map is revealed. You can see it. You know it's there. It's not a guess anymore. You've been there. You can go back. You can build on what you found. It might change while you're not there. But you went.

Vision is cleared map. Vision is what you can know because you've walked there. It's not what you think might be out there. It's not what you hope is out there. It's the territory you've personally explored and can now navigate with confidence. You spent money, time, resource to do something, and now you know something.

This is important because a lot of founders think vision is their idea. Or they think it's their idea plus some lines, some colour, like if they can explain it clearly enough then it becomes vision... But that's not it. Vision comes from what you know. And you can only know what you've touched, tested, tried, failed at, succeeded at, talked to people about, built prototypes of, gotten rejected on.

Vision is earned sight. You can't have vision of territory you haven't explored.

So what's a dream then? A dream is the reason you walk into the fog at all.

Here's where it gets interesting. The dream isn't just "I want to build something cool" or "I want to be a founder" or whatever vague aspiration. A good dream, the kind that actually leads somewhere, has a specific structure to it. It goes something like this: these new primitives exist, or they're emerging right now, and I can see an extrapolation that almost nobody else is making. I can see how these primitives might combine or apply in ways that aren't obvious yet. That's the secret. And it's not a fantasy. It's a bet. It's grounded in something real, something observable, but the conclusion you're drawing from it isn't obvious to everyone else.

The dream says there's something valuable in that direction, even though you can't see it yet. The dream picks the arena. It says: this is the bounded territory where my secret should pay off, this is where I'm going to plant my flag and start clearing fog.

And then vision is what you discover once you're actually in that arena, doing the work. Vision is what you now know is true because you've touched it.

This explains why both matter and why confusing them kills companies.

Dream without arena means you're a wandering philosopher. You have interesting thoughts about what might be possible but you never commit to a territory, so you never accumulate vision. You never actually learn anything. You just keep thinking.

Arena without dream means you're a grinder. You picked some market, you're working hard, but you have no reason to believe this territory is more valuable than any other. You're clearing fog but you don't know if there's anything worth finding.

Vision mistaken for dream means you stop exploring. You think you already know what's out there because you've cleared a little bit of map, so you stop walking. You get complacent. You miss the bigger thing that was just a bit further into the fog.

Dream mistaken for vision is maybe the most common and most dangerous one. This is when you make detailed execution plans for places you've never been. You have this beautiful dream about what's possible and you start acting like you can see it clearly, like it's real, like you just need to go execute. But you haven't actually been there. You're navigating by fantasy, not by cleared map.

So what does this look like in practice? What does a good pitch look like if you understand this distinction?

The dream part of a pitch should establish primitives as observable fact. Not "I think robots are coming" but "here are videos of robots doing things in factories in China right now, here's evidence of supply chains forming around them, here are products shipping today with sensors that didn't exist at this price point two years ago." You're showing the building blocks, the raw materials, the things that are actually true and observable today.

Then you show your extrapolation, which is the secret. You're connecting dots that other people aren't connecting. Most people see cheap lidar sensors and think autonomous vehicles. You're saying no, the spillover is the opportunity. The same supply chain that's making sensors for cars is going to make sensors for home devices, and nobody's paying attention to that yet. That's a non-obvious chain of reasoning. That's the dream.

Then you name the arena and the ambition. Not "revolutionize hardware" which is meaningless, but "take out Dyson" which is a clear, bounded, targeted war. Consumer home devices. Everyone knows what that means. Everyone can picture the territory.

And notice you're saying "I have a dream" not "I have a vision." Because you're being honest about what you actually know versus what you're betting on.

The next part of the pitch is showing the vision, the cleared fog, what you actually know because you've been walking around in this arena. This is where you say things like: I've spent eight months on this. I've talked to six factories in Shenzhen. Three of them are producing lidar units at twelve dollars that cost a hundred and eighty dollars two years ago. One of them is hungry for non-automotive customers because their EV order book is lumpy. They'll do small runs. I bought fourteen robot vacuums and tore them all down. The navigation is still mostly bump-and-guess or camera-based SLAM that fails in low light. Nobody's using the cheap lidar yet because the firmware integration is a mess, but I found a guy who worked on the DJI obstacle avoidance stack. I've tested three prototype housings. Turns out the form factor everyone uses is driven by old dustbin geometry assumptions that don't apply if you have real spatial awareness.

Every one of those statements is discovered knowledge. It's not "I believe" or "I think." It's "I went there and I found this." An investor hearing this can now evaluate something real. Does this founder see more of the map than others? Are they ahead in fog-clearing? Would my money buy them more territory faster than someone starting cold?

Then comes strategy, which is the next set of fog-clearing expeditions. And the key word here is limited. Strategy is not a sprawling plan. It's not a gantt chart. It's two or three missions that will clear the most valuable fog with this capital.

Given what I now see, which is vision, and where I believe this leads, which is the dream, here are the specific expeditions we're going to run. Lock in an exclusive prototype run with the Shenzhen factory, fifty units, proves we can ship real hardware. Hire two firmware engineers to nail the lidar integration, that's the moat, everyone else will hit the same wall we hit. Seed fifty units into homes with a logging stack so we get six months of real spatial data, that's the vision we don't have yet, that's the fog we need to clear to know what V2 looks like.

Each mission is bounded. Each mission clears fog. Each mission connects to the dream. Each mission is evaluable, you can tell whether it worked or not. The investor is now buying specific expeditions, not a vague hope.

And here's the part that really separates founders who understand this from founders who don't. If your missions are clear, you can make if-then statements. Not a full decision tree going out to infinity, but one or two levels deep. You can say: if the fifty-unit run comes back and the lidar integration holds but the form factor gets complaints, we know the stack works and V2 is an industrial design problem, we hire a product designer not more engineers. If the home data shows people actually want it to do more than vacuuming, they're asking it to patrol or monitor or map for other purposes, then we're not a vacuum company, we're a home spatial platform, different raise, different pitch. If the firmware proves harder than expected and we burn three months stuck, we acqui-hire from the DJI diaspora and extend runway by cutting the prototype run to twenty-five units.

This signals everything. It signals you know what you don't know. It signals you're not committed to a fantasy, you're committed to learning. It signals you've thought about failure modes without being paralyzed by them. It signals the money has a job and you know how to evaluate whether it did that job.

Compare this to the founder who pitches a fixed plan and then when you ask about contingencies just says "we'll iterate." That's admitting they don't know what fog they're clearing or what they'll do with what they find. They've confused their dream for vision. They think they can see the whole map. They can't. Nobody can. The good founders know that and structure everything around the act of earning sight.

The dream sets the bearing. The vision shows the next ten steps. The strategy is choosing which steps to take. And the whole thing is a loop, because every step you take clears more fog, and the new fog you clear updates your vision, and your updated vision might even update your dream, and you keep going until you've either found what you were looking for or discovered that it isn't there.

That's the game. That's the whole game.

much love!

j.