Approaching Null Hypothesis

I started this blog with a specific, perhaps narrow, intention: to document my learning journey in adapting to AI tools. I thought the narrative was about the technology. I’ve realized since then that the technology is just the backdrop. The true story is the transition from being an institutional individual to an independent one.

Moving from the structured safety of an organization to the "wild jungle" of independent development changes the physics of how you work. It forces a confrontation with what I now call Problem 0.

The Three Capstones

To date, three projects have defined this transitional period. They have been my education.

1. The Infrastructure: My first website, integrated with a chatbot, Q&A, and forms. This was a lesson in spec-driven development and the regulatory hurdles of ICO.

2. The Game: A project that was genuinely fun to create, however, it currently sits dormant in my dev box. Here, I encountered the invisible barriers of saturated platforms/markets: Apple Developer Enrollment and the tedious internal testing requirements of the Android Play Store.

3. The High-Stakes Bridge: A third project that feels promising but I also understand the complexity and the risk it carries.

These projects moved me from the mindset of an individual contributor to an architect in a dynamic environment. They forced me to ask the questions that rarely appear when you are working for someone else: Where is the market? Who is the target user?

Iterating on Problems, Not Solutions

In a institution, the problem was usually a given. Success was measured by the elegance of the solution. Naturally, when I started this blog, I wanted to share my findings on tools and readings.

But sharing tools is an iteration on solutions. In the wild, the solution is secondary. If you solve the wrong problem, the quality of your code is irrelevant. I have pivoted my focus to iterating on the problem itself.

This is where the Null Hypothesis becomes my primary framework. In this context, the Null Hypothesis is the assumption that the problem I am trying to solve does not actually exist, or that no one cares enough to pay for a fix. My job is not to prove my app is good. My job is to fail to prove the Null Hypothesis.

We are entering an era where the marginal cost of software code is approaching zero. If the value proposition is just "the code," there is no moat. AI will eventually eat that feature or make it a native function of the OS, the focus must shift from the "brain" (the logic) to the "nerve" (the integration into messy real world).

The Shift

The world will soon be flooded with automated, even "junk", software, value lies in the curation of a specific outcome for a specific group of people. I don’t have all the answers yet, but I am beginning to sense the direction. 

Because of this shift, updates here may become less frequent. I am no longer interested in tracking every tool I find. I am looking for the "first real thing." My blog will continue as a hobby or a technical anchor, but my primary energy is now directed toward proving the Null Hypothesis wrong.