You Have Started Thinking in a Third Mode. You Just Do Not Have a Name for It Yet.

In partnership with

Scale AI support on AWS, see how July 9

Customer expectations keep rising. Support budgets don't. On July 9, Fin and AWS are hosting a live executive session on how leading enterprises close that gap: scaling AI-powered support while simplifying how they buy it.

You'll see how to resolve an average 76% of conversations with Fin on AWS enterprise-grade infrastructure, procure through AWS Marketplace to put committed cloud spend to work, and turn the Fin and AWS collaboration into lower support costs. Register for the live session to see how.

Save your spot

In 2011, Daniel Kahneman published a book that changed how people talk about the way they think.

System 1, he called it, is fast, automatic, intuitive. You see a face and immediately know whether the person is happy or angry. You brake before you consciously register the car stopping in front of you. System 1 runs constantly, effortlessly, below awareness.

System 2 is slow, deliberate, analytical. You use it when you calculate a tip, read a contract, or work through a problem that requires focused attention. It is effortful. It is the part of your mind that feels like work.

For over a decade, that two-system model was the dominant framework for understanding human cognition. Almost every popular book about thinking, bias, and decision-making was built on top of it.

Researchers at a consortium of European universities published a paper this year proposing a third system. They call it System 0.

System 0 is not a part of your brain. It is the AI you consulted before you made the decision.

What System 0 Actually Describes

The researchers describe System 0 as the outsourcing of certain cognitive tasks to AI systems that can process data and perform computations beyond human capability.

That definition sounds technical. In practice it describes something most people with a smartphone are already doing every day, often without noticing.

You need to decide which route to take. You do not think through the options. You open Google Maps. System 0.

You are drafting an important message and you are not sure of the tone. You paste it into an AI and ask if it sounds right. System 0.

You have a symptom you are worried about. You do not call a doctor first. You ask an AI to tell you what it might be. System 0.

In each of these cases, a cognitive task that used to happen inside your head is now happening outside it. The AI processes the inputs, applies the patterns, and returns an output. You receive the output and act on it.

The researchers are not saying this is bad. They are saying it is a distinct cognitive mode that has not been properly named or studied. And that the absence of a name for it has made it almost impossible to notice when you are in it.

The Conformity Study That Should Concern You

A separate study published this year documented something called algorithmic conformity.

The mechanism works like this. People are significantly more likely to change their moral judgments after receiving an AI recommendation than after receiving the same recommendation from another person.

Think about what that means in practice.

If a friend tells you that a decision you are wrestling with is the right one, you weigh their opinion against what you know about them. Their track record. Their values. Their potential biases. Their relationship to your situation. You do not just accept what they say. You contextualize it.

When an AI tells you the same thing, that contextualization process mostly does not happen. People perceive AI recommendations as more neutral, more objective, more authoritative. The social machinery that helps you calibrate and pushes back does not engage in the same way.

The result is that AI recommendations change minds more efficiently than human ones do. Not because the AI is right more often. Because the skepticism people apply to human sources does not transfer automatically to an AI source.

A doctor who tells you a treatment is safe triggers the part of your mind that knows doctors are fallible, that their clinic might have financial incentives, that your case might be unusual. An AI that tells you the same thing triggers something quieter and more compliant.

The Homogenization Problem

Here is where this gets structurally interesting beyond the individual level.

A study published in 2026 documented what researchers are calling the homogenizing effect of large language models on human expression and thought.

The argument is straightforward. When millions of people use the same AI systems to help them write, decide, and think, those AI systems pull outputs toward statistical centers. The language that models produce is shaped by the most common patterns in their training data. The perspectives that models offer reflect the most frequently represented viewpoints in that data.

A world in which a significant fraction of human cognitive output is routed through a small number of AI systems is a world in which diversity of thought faces a new kind of pressure it has never faced before.

This is not about AI replacing human creativity in the obvious sense. It is subtler. It is about the distribution of ideas narrowing toward what the models were trained to produce, while each individual person still believes they are thinking their own thoughts.

A single advisor who influences a million clients is a different kind of influence than a million advisors each influencing one client. The former concentrates the direction of thought in ways the latter does not.

AI is closer to the former than most people have stopped to consider.

What the Deloitte Data Actually Shows

Deloitte's 2026 Global Human Capital Trends survey found that 60 percent of executives now regularly use AI to support their decisions. Gartner projects that by 2027, half of all business decisions will be augmented or automated by AI.

Those numbers sound like efficiency gains. They are also a description of a world in which System 0, the outsourced cognitive layer, is becoming embedded in the operational structure of organizations, not just the personal habits of individuals.

When half of business decisions run through AI systems, the question of whose thinking those systems represent becomes a governance question, not just a philosophical one. Every model has training data. Every training data set has a point of view embedded in it, implicitly, through what was included, what was weighted, and what was considered high quality.

The people whose thinking shaped that training data have more influence over organizational decisions in 2026 than most boards or leadership teams have bothered to understand.

What to Do With This

The researchers who introduced System 0 are careful not to frame it as a warning to avoid AI.

They frame it as a call for metacognition. The ability to think about your own thinking, to notice when you are in System 0 rather than System 1 or System 2, and to make a deliberate choice about whether outsourcing that particular cognitive task is actually what you want to do in that moment.

The practical version of this is simpler than it sounds.

Before you accept an AI output as the basis for a decision, pause for one second and ask whether you processed it or just received it. Not every time. The small things do not need this. But for the decisions that actually matter, the ones about relationships, about work, about what you believe is right, the pause is worth the cost.

System 0 is not going away. But you get to decide when you are in it.

If you noticed yourself in System 0 this week in a way that felt significant, that observation is worth holding onto. Send it back if you want to think it through.

Thanks for reading. See you Thursday Please support if you like the content.

Want to get the most out of ChatGPT?

ChatGPT is a superpower if you know how to use it correctly.

Discover how HubSpot's guide to AI can elevate both your productivity and creativity to get more things done.

Learn to automate tasks, enhance decision-making, and foster innovation with the power of AI.

Download the free guide.