Published in the San Diego Union-Tribune, August 25, 2025
by Neil Senturia
We have all heard this mantra — we are a data-driven company, show me the data, it is all about the data.
There is another school of thought. This from American billionaire businessman Barry Diller: “If it’s a good idea — get on with it.” Diller was famously known for and is the undisputed poster child for “trust your gut.” And he has the billions to back it up. So the puzzle for today is intuition versus data.
Laura Huang, professor at Northeastern University, has written a book, “You Already Know, the Science of Mastering Your Intuition.” For all the entrepreneurs out there, wrap your head around this one.
Huang interviewed a large number of venture capital investors asking them how they decided to invest or pass on the various companies they were considering. She says 100% “of the people would talk about data, financials, market, but they also talked about, ‘yeah, I just had this feeling that I had to invest in this person.’”
This next quote will explain why this startup game qualifies to be in the insanity hall of fame. “One investor said, I rub my tummy, I just rub my belly.”
Sure, I guess all that “product market fit crap” that you worked on for the past six months should simply be reduced to a nutraceutical roll-on lotion with triple lipid anti-wrinkle cream, and you can name your own pre-money.
To be fair, intuition is a major element of any decision. Of course, the more experience you have, the better your intuition should be. Then you have to factor in biases. Biases are stubborn. They do not yield easily to data.
Case in point. My little artificial intelligence company was looking for an expert in marketing. The first candidate, no relationships to any of our team, came from LinkedIn, thrice removed. Fabulous. Perfect. Why the hesitation.
I say, we should look more, maybe we can find better, we need to do a more thorough search. Classic bias. My partner hired him the next day. So far, we look like geniuses. “I knew it the minute I saw him.” I will remind him of this when we are bankrupt.
Now, let’s fold some machine learning into the intuition bucket. The LLMs (large language models) were trained on a trillion words and patterns, and in essence when you ask it a question, the answer they give you is a guess, but a very contextual guess derived from massive previous ingested data and correct answers.
OpenAI CEO Sam Altman has called AI a magic trick, but after all, it is really just data. And the process of human intuition is iterative. Pattern recognition distilled.
Here is the problem again. We learn in a linear way. We like to see truth and rational answers in a sequence (think Fibonacci), but the core power of gut decision making is that it is not linear. We assimilate experiences from childhood onward, and then in a particular moment, we grab them and put them all together in an instance and we decide, we act. It is the human instinct to do something.
Dan Kahneman’s book “Thinking Fast and Slow” is the seminal work on when to trust your gut. What appears to be all gut emotional is actually a “combination of priors.” We each access our own personal LLM.
Instinctive versus emotional. The subtle nuance is that once you make a gut decision, even if it is wrong, you work really hard to prove that you made the right one. A quick caution — “all in” can blind you to sunk costs.
How we remember priors is also subject to confirmation bias. This explains why we do the same stupid thing over and over again. Co-founder, marriage partner, call the divorce lawyer. But she reminded me of my grandmother.
Your logo is not the problem. Those decisions are simple. It is the Billy Beane Moneyball data-driven decisions that can give you AI hallucinations.
Let me bring it home for all startups, courtesy of Mr. Billy Beane. “The problem we’re trying to solve is that there are rich teams, and there are poor teams, then there’s 50 feet of crap — and then there’s us. It’s an unfair game.”
I tell my team those famous words: “You get on base, we win.”
Rule No. 794: Seemed like a good idea at the time.