Neuroscientist Proposes Hybrid Intelligence Index as AI Doubles Engaged Students When It Adds Friction
Updated
Updated · Financial Times · Jun 8
Neuroscientist Proposes Hybrid Intelligence Index as AI Doubles Engaged Students When It Adds Friction
2 articles · Updated · Financial Times · Jun 8
Summary
A proposed Hybrid Intelligence Index would rate AI systems by whether human-machine collaboration leaves users cognitively sharper or duller over time, rather than by model scores alone.
EEG classroom tests found most students' high-frequency gamma activity collapsed within minutes of using the same AI agent, while a small minority who challenged its answers showed heightened engagement.
In a modified experiment, an AI prompted to answer with questions and context instead of direct solutions more than doubled the share of high-gamma, actively engaged students.
The argument is that current oversight and benchmarking miss this effect: Article 14 of the EU AI Act treats human oversight as a safety check, while labs in Britain, the US and EU still test models largely in isolation.
Over 60,000 students studied between 2012 and 2016 showed top performers were often wrong in discussion forums, underscoring the broader claim that economies optimized for AI fluency could erode human judgment.
If 'arguing' with AI is the key future skill, how do we prevent a massive cognitive divide in society?
As AI automates thinking, can we design it to be a cognitive workout partner instead of a mental couch?
Hybrid Intelligence in Education 2026: The Case for Productive Friction and the Hybrid Intelligence Index
Overview
Artificial Intelligence (AI) is rapidly transforming education worldwide, but its swift adoption raises important concerns. As students increasingly rely on AI tools, there is a risk of weakening critical thinking skills and producing work that lacks originality and unique value. This over-reliance, driven by a culture focused on speed and efficiency, can also create a 'moral distance,' reducing personal accountability when AI systems make decisions. The report highlights the need for deliberate challenges—'friction'—in learning to ensure students remain actively engaged, think critically, and develop the skills necessary for meaningful understanding in an AI-supported environment.