Null to zero

Launching a startup will take almost nothing, it nearly does now. With a few prompts you will have a working piece of software, a few more and you have a website to sell it from. Prompts to automate your outbound sales, your SEO, your ad buying, email marketing, LinkedIn self-promotion, even your VC outreach will be automated.

Your strategy isn’t really your strategy, it’s Claude’s. A prompt output you asked to remove the emdashes from and then directed it be put in your Notion workspace or Google Drive. You will skim it, and maybe reword the intro.

“Are you technical? Who’s going to build it?” I am, well a magic box will, but I will tell it to.

AI has made this possible. To go from idea to MVP takes so little effort now that thousands of software startups will be built and launched with very little technical skill demonstrated. Very few of them actually should be. To be clear this isn’t a post where an old man shakes fist at sky and tells you how it used to be better when it was harder. Lower the barrier to build and more people who otherwise wouldn’t have built startups will… this is undeniably a good thing for early stage companies and our economy.

Rather, this is asking if the valuable part is building, or deciding whether to build. Zero isn’t the starting point to a startup; when you’re on zero you know that one is possible. You understand the shape of the thing you’re trying to build when you’re on zero.

Before zero comes null. This is where real uncertainty lives. The problem may be undefined, the market likely isn’t proven, and no one knows what should be built. I call this null to zero; and the valuable work here isn’t building, it’s discerning.

This work is asking the hard questions of whether the problem actually exists, and if it does exist does it exist enough for someone to care that it is solved.

At this point your work looks more like anthropology than anything else. It is dozens of conversations where your goal is to refine your set of beliefs and then build your conviction about them. This stage’s output isn’t a product, it’s a decision. Go or no go. Your conviction might just be that spending the next few years building in this space is a fools errand. Not falling into quicksand is also a win.

I love this work, not because it is as exciting as going zero to one; but because it is more important.

As AI marginalizes the technical skill and reduces the cost in launching something you would think this work becomes less valuable. With less cost in time and money to create a product why not take ten shots and increase your chances of hitting on one? But it’s exactly that work of determining whether there is a problem worth solving, or a market worth serving, that often times illuminates the real value proposition worth going after.

If you knew nothing of anatomy and only shot birds with buckshot you might never figure out what the actual target was. Refining your taste for opportunity is a skill worth cultivating.