Observation Becomes Intelligence

Most people assume intelligence begins with analysis.

I used to think that too.

The conventional wisdom is that better decisions come from better software, better reports, better dashboards, or more sophisticated algorithms. Entire industries have been built around helping organizations analyze information more effectively.

Yet after spending more than two decades building products, teams, and technology companies, I've come to a different conclusion.

Long before a report is generated, a prediction is made, or a decision is reached, something has to be observed. The quality of everything that follows is constrained by the quality of that observation.

If the observation is incomplete, the analysis will be incomplete.

If the observation is inconsistent, the conclusions will be inconsistent.

If the observation is wrong, no amount of downstream intelligence can fully recover from it.

This realization didn't arrive all at once. It emerged gradually through a series of ventures that, at first glance, appear unrelated.

Years ago, while building Bumper, I became fascinated by customer behavior. The challenge wasn't a lack of data. Dealerships already had data. The challenge was understanding what customers were actually doing and identifying meaningful signals hidden within that behavior.

The more I looked, the more I realized that most organizations weren't suffering from a lack of information. They were suffering from a lack of meaningful observation.

The same pattern appeared again years later while building PAVE.

The automotive industry was filled with inspection processes, condition reports, and workflows designed to support downstream decisions. Most conversations focused on reporting, automation, or operational efficiency. What fascinated me was something much earlier in the process.

Everything depended on a simple assumption.

That the visual evidence entering the system could be trusted.

I remember reviewing inspection results and realizing that two people could photograph the same vehicle and produce dramatically different outcomes. Different images. Different observations. Different conclusions.

Everyone around me was focused on improving the intelligence layer.

I couldn't stop thinking about the observation layer.

How could we trust the decision if we couldn't trust what was being observed?

That question stayed with me.

It eventually led me into computer vision and machine learning long before those technologies became mainstream topics in the automotive industry. Not because I was chasing artificial intelligence as a trend, but because I became obsessed with a different possibility.

What if observation itself could become a technology?

What if cameras evolved from passive recording devices into operational sensors capable of generating reliable intelligence about the physical world?

That question consumed years of my attention.

It eventually became PAVE.

Later, it became Halo.

But the more important realization was broader than any individual company.

Observation is one of the most undervalued assets in modern business.

Organizations invest heavily in analytics, automation, dashboards, and increasingly AI. Yet surprisingly little attention is paid to the quality, consistency, and reliability of the information entering those systems in the first place.

The result is predictable.

We build increasingly sophisticated intelligence layers on top of increasingly inconsistent observations.

Artificial intelligence has only amplified this challenge. Every AI system inherits the strengths and weaknesses of the information used to train and operate it. Despite all the attention placed on models and algorithms, the most important question remains remarkably simple:

What did the system actually observe?

The answer often determines everything that follows.

This principle extends far beyond technology.

The best entrepreneurs are exceptional observers.

The best product builders are exceptional observers.

The best leaders are exceptional observers.

They notice patterns others miss. They identify changes before they become obvious. They develop an ability to separate signal from noise while everyone else is overwhelmed by information.

In many ways, entrepreneurship itself is a form of observation.

A founder sees something that feels slightly out of place. A process that doesn't make sense. A behavior that nobody questions. A problem that has become so familiar that everyone accepts it as normal.

Then they become obsessed with understanding it.

Every company I've helped build can ultimately be traced back to that process.

Not analysis.

Observation.

The insight comes later.

The technology comes later.

The intelligence comes later.

It all begins with paying attention.

Because before intelligence can exist, something must first be observed.