cAIos: An Interview on the Future of Thinking
Interviewer: SOTlobe
Guest: Dr Michael Hewitt-Gleeson
SOTlobe: Let’s begin at the beginning. You’ve said the future didn’t arrive quietly—it arrived in a briefcase. What happened?
Dr Michael: It was 1976, in a New York boardroom. A salesman placed a telephone receiver into a suitcase, dialled a few tones, and connected—live—to a mainframe in San Francisco. Then he said, “Ask me anything.”
In a world built on delay, he collapsed time. That moment changed my understanding of intelligence. Not what it is—but what it could become when speed and access converge.
SOTlobe:: And that moment led you to study thinking itself?
Dr Michael: Yes. Four years later, I completed the first PhD in lateral thinking. I was studying the brain as if it were an operating system—powerful, yes, but also biased, inefficient, and prone to looping.
What became clear was this: intelligence isn’t the problem. The problem is that the brain runs on default settings. It defends its first answer. It resists better ones.
So my work shifted. Not helping people think faster—but helping them think better.
SOTlobe:: You then stepped into the technology world quite directly—IBM, Monte Carlo, 1982.
Dr Michael: That was a pivotal moment. I was invited to give the keynote at the European launch of the IBM PC in Monte Carlo. My talk was titled Software for the Brain.
At the time, it was slightly heretical. The dominant view was that computers would automate work. I argued they should accelerate thinking.
The real opportunity wasn’t in silicon alone. It was in cooperation—between two operating systems: silicon and carbon.
SOTlobe:: That idea became cAIos?
Dr Michael: Yes. That same year, in New York, I designed what I called C AI O S—a Cognitive AI Operating System.
The idea was simple. Not artificial intelligence replacing humans, but amplifying them. Expanding cognition. Increasing the number of options. Reducing the drag of habit and bias.
It later appeared in my book NewSell in 1989. But at the time, it was early. Too early.
SOTlobe: Too early in what sense?
Dr Michael: The idea was obvious. The timing was not.
The brain needed software—but there was no external processor capable of reliably running it. So cAIos in the 1980s had to be installed inside the human mind.
It took the form of training programs, workshops, manuals, mental routines. People had to run the system themselves.
And that’s where it broke down.
SOTlobe: Because humans couldn’t execute it?
Dr Michael: Precisely. The limitations were structural.
Humans stall at three or four ideas. Emotional bias overrides structure. Roles blur—the thinker becomes the defender, the critic becomes the advocate.
Even with training, execution was slow, effortful, and inconsistent.
So cAIos existed—but it wasn’t fully executable.
SOTlobe: Fast-forward to now. What changed?
Dr Michael: Computing power caught up. Large Language Models arrived. The “magic suitcase” reappeared—this time in the cloud.
For the first time, we had something that could act as an external cognitive processor.
That changes everything.
SOTlobe: In what way?
Dr Michael: The system no longer has to run inside the brain. It can run alongside it. Like a cognitive prosthesis.
That’s the shift. From cognitive training to cognitive infrastructure.
Before, you taught people how to think better.
Now, you give them a system that does it with them.
SOTlobe: Describe cAIos today.
Dr Michael: Today, cAIos is an executable system.
It has three layers. The system itself—cAIos. The interface—COS, or Cognitive Operating System roles. And the engine—GBB, the Good/Bad/Better algorithm.
The human becomes the director. The AI becomes the processor.
Together, they form a cooperative intelligence.
SOTlobe: What does that cooperation actually produce?
Dr Michael: Better thinking.
The brain contributes judgment, values, discernment, creativity. AI contributes speed, pattern recognition, memory, and iteration.
When combined properly, they produce outcomes neither can achieve alone.
Not faster answers. Better answers.
SOTlobe: Let’s make that concrete. What can the system now do that humans alone struggle to do?
Dr Michael: It can reliably generate ten GOOD, ten BAD, and ten BETTER ideas—what we call a GBB.
Humans rarely get past three or four. The system removes that ceiling.
It maintains strict role separation. It runs parallel thinking modes. It externalises cognition—so thinking becomes visible, editable, repeatable.
In short, it executes what the brain cannot reliably execute on its own.
SOTlobe: So the missing link has been solved?
Dr Michael: Yes. And the missing link was never the idea. It was execution.
The brain could not reliably run its own upgrades.
Now it can—because AI handles volume, structure enforces discipline, and algorithms force escape from the first view.
SOTlobe: Your recent work—AIX10, cvs2bvs.ai—how do they fit into this?
Dr Michael: They are implementations of cAIos.
AIX10 is a training system—teaching people how to think with AI. Not prompt tricks. Not shortcuts. Cognitive partnership.
cvs2bvs.ai is the applied system. It operationalises the core law: the Current View of the Situation can never equal the Better View of the Situation.
The job is to move from one to the other—reliably, every time.
SOTlobe: And now you’re building SOTbots?
Dr Michael: Yes. But more than that. We have developed a cognitive prosthesis that uses elite trained SOTbots to give you an meta-cognitive layer, an additional lobe for your brain. Fully qualified AI brain coaches. This has never been done before anywhere in the world.
We’ve already built an application that can run a GBB on demand. The next step is embedding that capability into conversational agents—so the system is always available, always structured, always pushing toward a better view.
SOTlobe: If we step back, how should we now define cAIos?
Dr Michael: Cleanly?
cAIos is an external cognitive operating system that extends the brain through structured, role-based, AI-executed thinking.
SOTlobe: That sounds like more than a tool.
Dr Michael: Exactly. It’s not just a product. It’s not just a course. It’s not even just a method.
It’s a cognitive layer.
Like operating systems sit beneath applications, cAIos sits beneath decisions.
SOTlobe: Which leads to the obvious question. What is the future of AI, in your view?
Dr Michael: The future isn’t artificial intelligence.
It’s augmented judgment. Better still, augmented discernment.
Brains that know how to work with machines—not later, not occasionally, but in real time.
SOTlobe: Final question. If you were to summarise this forty-year journey in a single line?
Dr Michael: For forty years, we tried to improve the brain.
It turns out the easier solution was to give it a second one. What I call an ‘additional lobe for your brain’.
