In the opening keynote of the Future Intelligence Forum, cognitive neuroscientist Dr. Heather Collins argued that the economy has entered a new phase. Information became free with the internet; with AI, knowledge has become free as well. What remains scarce, and what will separate the organisations that succeed from those that fail, is human judgment.
Speaker: Dr. Heather Collins, Cognitive Neuroscientist & Brain Scientist
In the opening keynote of the Future Intelligence Forum, cognitive neuroscientist Dr. Heather Collins argued that the economy has entered a new phase. Information became free with the internet; with AI, knowledge has become free as well. What remains scarce, and what will separate the organisations that succeed from those that fail, is human judgment. Drawing on neuroscience and a series of live experiments with the audience, Collins explained how the brain makes decisions, where it differs from AI, and how leaders can use AI to strengthen their thinking rather than replace it.
The brain as a prediction machine
Collins began with a basic principle of neuroscience: the brain is a prediction machine. Its primary function is to anticipate what will happen next in order to keep us alive. The prefrontal cortex extends this capacity to professional life, allowing us to predict how a meeting will unfold, whether a deal will close and which move is most likely to succeed.
This predictive capacity, she explained, is the foundation of human decision-making. Understanding how it works, and where it can mislead us, is the first step toward using AI well.
Context is a feature, not a bug
To illustrate how the brain differs from AI, Collins ran a simple experiment. Half the audience was shown an image of a dinner plate, the other half an image of a flower. Everyone was then asked to unscramble the same set of letters. The room divided almost exactly between two answers, "plate" and "petal", each shaped by what that group had just seen.
Both answers were correct. The point was that the human brain is highly sensitive to context, whereas AI processes data in a supposedly neutral way. Collins described this sensitivity as "a feature, not a bug": to survive and succeed, we must pay attention to our surroundings. "We have a room to be in. We have an environment to pay attention to. We have context," she said. That context can change the quality of a decision entirely.
AI produces the signal; humans own the decision
Collins used the example of radar in the Second World War to describe the relationship between people and technology. Radar was excellent at producing a signal, indicating that something was in the airspace. But the operator still had to decide whether that signal was an enemy aircraft or a flock of birds.
That decision carries two possible errors. A false alarm wastes resources and puts people at risk unnecessarily. A miss allows a real threat to pass unnoticed. In the same way, AI produces signals, and different people will interpret the same output differently depending on their tolerance for risk. Neither the cautious nor the bold interpretation is automatically right. What matters is that responsibility for the decision remains human. In her words:
AI can make the call, but only humans can live with the consequences.
Fluency bias and the case for slowing down
A second experiment, a version of the Stroop test, asked the audience to name the colour in which each word was printed while the word itself named a different colour. The task is deliberately uncomfortable, because reading is automatic and must be actively suppressed.
Collins used it to explain fluency bias: "Anything that comes fast, your brain believes is true." Because AI generates information faster than we can absorb it, it can create an illusion of accuracy. Her recommendation is counterintuitive: when the stakes are high, slow down. A deliberate pause engages the anterior cingulate cortex, which she described as the brain's router. It detects a conflict and directs the information to the prefrontal cortex for careful evaluation. Given the trade-off between speed and accuracy in human decision-making, "if we slow down a little bit, the decision-making quality will actually increase."
From brain drain to brain gain
Collins was clear that AI is extremely useful, and that she uses it every day. The question is how it is used. In what she called "brain drain", people offload their thinking to the tool, fall into fluency bias and develop miscalibrated confidence in answers they have not verified, despite the known risk of hallucinations. In "brain gain", AI amplifies human strengths through knowledge building, critical thinking and metacognitive awareness.
She proposed a three-step framework to stay on the right side of that line:
- Construct. Define the problem before using AI, whether it concerns an investment, a partnership or an acquisition. Even a few minutes of independent thinking builds a foundation for what follows. "Your brain deserves a right to a first draft."
- Challenge. Keep prompting, examine the details, and ask the AI to challenge your assumptions, biases and blind spots. This turns a one-way flow of information into a genuine exchange that stimulates critical thinking.
- Calibrate. Develop metacognitive awareness, meaning an accurate sense of what you know and what you do not. Collins demonstrated this by asking the audience for the capitals of Chile and Australia. Many people answer Sydney for Australia with high confidence, although the correct answer is Canberra. Regularly asking oneself "how confident am I?" is, she said, the quickest way to build this skill.
Whether AI will change the brain itself remains an open question. Research has not yet caught up with the pace of the technology, Collins noted, although one finding is well established: attention spans have measurably declined over the past two decades. The outcome will depend on usage rather than on the technology alone. Some AI-first schools in the United States, where students spend several hours a day in personalised learning with AI, report that advanced students progress far faster than their peers. "It's not that AI is bad for the brain or good for the brain," she said. "It's all about how it's being used."
Meaning over capacity
In a final experiment, the audience was asked to memorise a sequence of twelve digits, and then a sequence of twelve letters. Most participants recalled only a few digits, but nearly all recalled the twelve letters, because they formed three words.
For an AI system, both sequences contain the same amount of information. For humans, meaning transforms twelve separate items into three. "There's no question, AI beats us on capacity," Collins said. "But for humans, meaning is important." Meaning is built by the hippocampus into the semantic memory network, the brain's store of everything we have learned about the world. It also allows us to infer, to fill gaps in incomplete information, and to read the room: to notice hesitation in a colleague's voice or a reaction during a meeting. "You are in the room. AI is not."
Keeping the human brain strong
If judgment depends on the brain, its health becomes a leadership concern. Collins pointed to three habits supported by research. The first is nutrition: omega-3 fatty acids and antioxidants give brain cells the infrastructure they need, and compounds found in foods such as blueberries are particularly beneficial for the hippocampus. The second is continuous learning, which physically strengthens the brain, including the corpus callosum that connects its two hemispheres. The third is social interaction. Contrary to the common narrative of cognitive decline, social information processing improves with age, and it underpins the skills leaders rely on most: reading the room, anticipating the course of a conversation and building relationships.
She also looked further ahead, to a field that blurs the line between human and artificial intelligence. Researchers can now grow small clusters of brain cells, known as "mini brains", from adult stem cells, and these can communicate with one another when connected by electrodes. Early data centres are already placing living brain cells on microchips. According to Collins, this "biological intelligence" has so far proved more stable than conventional AI infrastructure, with a much smaller carbon footprint and far lower cooling needs. "We're just at the tip of the iceberg," she said.
From a knowledge economy to a judgment economy
Collins concluded by placing these findings in a broader economic context. For most of history, information was scarce and valuable. The internet made information abundant, and advantage shifted to knowledge: the ability to make sense of that information. AI now makes knowledge abundant as well. "Our culture, society and economy are no longer knowledge-based. They're judgment-based."
The difference from the internet era is significant. Search engines made information abundant, but people still had to select, organise and integrate it themselves, a demanding process that built knowledge. AI now delivers that knowledge in a single, finished answer. Collins expects the consequences to become clear within about ten years, when the first generation of AI natives enters adulthood. Employers are already adjusting: many business owners she knows would not hire a graduate who cannot use AI fluently, including deploying agents, yet they also expect new hires to keep their own critical thinking. Universities, in her view, face a difficult period as they work out how to teach both.
In this environment, human judgment and critical thinking become the decisive differentiators, for businesses and for individuals. AI cannot read the room, form an emotional bond, or recall a conversation from weeks ago that changes today's decision. Nor can it imagine the future, because the brain networks that store memories are the same ones we use to imagine what comes next. Strengthening those networks, she argued, makes us better at prediction and better at judgment. Her closing words:
Better brains build better leaders, and better leaders build better worlds.
