For decades, the enlightened leaders have come to know thinking as a two-speed transmission. System 1: fast, instinctive, your gut reaction. System 2: slow, deliberate, deep-dive analysis. It's a useful model, but in the age of AI's ubiquitous expertise it will fundamentally change how we ideate and make decisions. System 3 thinking may be the only truly decisive, game-changing way that difficult decisions can be made under pressure.
In this proposed new form of thinking, System 3 is where decades of experience are compressed into a single moment of clarity. It's not just a gut feeling, and it's not a 50-page deck.
It's expert judgment moving at the speed of intuition. It's the ability to be prescient, not just reactive.
This isn't some abstract theory. I've lived it. Let me give you an old example.
The Cylance Story: From Reactive to Predictive
Back when I was the global CTO for a major antivirus company, my job was a masterclass in System 2 failure. The entire industry operated on a reactive loop: wait for an attack, get a sample, analyze it, write a signature, and push an update. We were always, by definition, one step behind. I spent more time on planes flying to apologize to customers for our product failing than I did innovating — and as such called myself the "Chief Apology Officer." The data was always telling us what had already happened.
When we started Cylance, we threw that entire model out the window. The System 3 question wasn't, "How can we react faster?" It was:
"Can we predict an attack before it ever executes?"
We decided to use mathematics and AI to identify the fundamental attributes — the DNA — of malicious code, regardless of whether it had been seen before. We were betting on foresight over hindsight. It was a contrarian move; the industry said prevention was impossible. But my experience told me the reactive model was a dead end.
That shift from a reactive to a predictive model was pure System 3 thinking. We weren't just faster; we were operating on a different plane.
A Modern Example: The Remote Work Dilemma
Now for a more recent example, outside of cybersecurity. Think about the leadership decisions around remote work post-pandemic.
- System 1 reaction: Panic. "Everyone go home, now!"
- System 2 response: An endless analysis of productivity metrics, badge swipes, and office occupancy rates — trying to find the perfect, data-driven answer to "how many days a week should people be in the office?" They're drowning in data but have no real clarity.
- System 3 leadership: Integrating the data with a deeper, more human understanding of the situation.
A System 3 leader asks different questions:
"What kind of culture do we need to build to attract and retain the best talent in this new world? How does collaboration actually happen for us? What's the long-term impact on innovation and mentorship if we get this wrong?"
They might make a call — like a flexible hybrid model or a fully remote-first culture — that isn't perfectly supported by today's productivity data but is based on a forward-looking judgment about what will make the company thrive in three to five years. They are playing the long game, based on a synthesis of data, experience, and a deep feel for their people and market. That's not in a spreadsheet.
What System 3 Actually Is
In both cases — Cylance and the remote work dilemma — the breakthrough didn't come from faster reactions (System 1) or more analysis (System 2). It came from reframing the problem based on earned wisdom, something AI alone struggles with impotently.
That's what System 3 is. It's the one thing AI can't replicate.
AI can run the System 2 analysis for you, but it can't make the System 3 leap of judgment. It can process the past, but only we can perceive and shape the future.
The Leader's Mandate
The leaders who will win this next decade will be those who use AI to automate the analysis but reserve the final, critical judgment for themselves. They'll know when to trust the model and, more importantly, when to override it with their own hard-won foresight.
The question for us in this room isn't "how much data can we process?" It's:
"How quickly can we turn it into wisdom?"
Because the future doesn't belong to the fastest thinkers. It belongs to the System 3 thinkers.