Wednesday, June 26, 2024

Natural-born cyborgs - learnings new imperative with AI

A much more philosophical and practical book than Anil Seth’s Being You is The Experience Machine by Andy Clark. It takes deeper dives into conceptualising the predictive brain as it looks both inwards and outwards. The book explodes into life in Chapter 6, with an absolutely brilliant synthesis of predictive processing towards goals and the idea of the extended mind (he with Chalmers wrote the seminal paper in early 90s). He starts with Tabitha Godlstaudm, a successful entrepreneur and dyslexic, who uses SwiftKey and Grammarly, seeing speech to text as her saviour. This is an example, of the extended mind, where we naturally use tools and aids to get things done. She moved beyond the brain because she had to. But this is about all of us.

When I travel, I book online, get boarding pass for the airport, book my hotel and glide through an electronic gate at customs using face recognition, order an Uber, get list of things to see on my smartphone, check out restaurants and use Google maps to get places – this seamless weave of mind and technology puts maid to the neuro-chauvinism around the mind being unique. As the “weave between brain, body and external resources tightens” our lives become easier, faster, more frictionless and opportunities to act and lean expand. This looping by the brain through the digital world becomes second nature, biology and technology entwine. The same is true in learning.


These theorists recommend that we need some humility here. It is not the traditionalists that show humility, as they are addicted to biological-chauvinism. The brain is perhaps less deep than they or we care to admit. What goes on inside our heads is limited, leaky, full of bottlenecks, with a seriously narrow working memory and fallible long-term memory. It stutters along, improvising as it goes.

He makes the excellent point that we ALL have a form of dementia, in the sense of constantly failing and fallible memories. This also means we ALL have learning difficulties. If our benchmark is the average human, that is a poor benchmark compared to the digitally enhanced human, with just a smartphone, especially one with AI, which means them all. We use AI unwittingly when we search, use Google maps, translate and so on. Increasingly we are using it to enhance performance.

Our brains don’t really care if what we use is inside the skull or smartphone, if it gets the job done and one consequence is his recommendation that we embrace the idea of a more fluid use of tools and supportive environment to get tasks done and to improve our own performance. He asks us to imagine we had Alzheimer’s and needed labels, pictures and reminders and supportive environments to get through our day. This idea of linking task performance to performance support, makes sense. It chains learning to real world action and actual performance.

We need to move on and see the encouragement and use of AI as a core activity. If we get stuck in the mindset of ranting and railing against the extended mind, even worse seeing it as a threat, we will be bypassed. The extended mind has been made real and relevant by AI. That is our game.

Natural-Born Cyborgs

In Mollock’s 2023 paper on productivity with 758 consultants from the Boston Consulting Group across 18 tasks, 12.2% more tasks were completed, with 25.1% faster completion and an astonishing 40% higher quality. The more fascinating finding from the data was the emergence of a group of superusers (cyborgs), who integrated AI fully into their workflow. Low latency, multimodal support has made this even more potent. Compare these to the Centaurs who benefit, but less, as they saw it as an add-on, adjunct technology, not as the extended mind.

We need to be looking at AI in terms of what Andy Clark calls the “natural-born cyborg’. We now collaborate with technology as much as people to get things done. It is a rolling, extended process of collaboration, where we increasingly don’t see it as separate from thinking. We have to free ourselves from the tyranny of time, location, language limits, and embrace the bigger opportunities that technology now offers through cognitive extension.

The naked brain is no longer enough and our job should be to weave that brain into the web of resources at the right time to get our jobs done, not drag people off into windowless rooms for courses or subject them to over-long, over-linear and impoverished e-learning courses, where they feel as though they have no agency.


Another consequence of the predictive, computational model of the brain is its ability to explain autism, PTSD, dreams, mental illnesses such as schizophrenia and hallucinogenic drug experiences. In The Experience Machine (2023) Clark goes into detail, with real case studies, on how predictive processing models suggest that psychosis can arise from disruptions in the brain's ability to accurately predict sensory input, leading to hallucinations or delusions. Schizophrenia is one example, similarly, the effects of drugs on perception and cognition can be understood as alterations in the brain's predictive models, changing the way it interprets sensory information or the level of confidence it has in its predictions. Dreams can also be seen as a state where the brain generates internal predictions absent of external sensory input. Autism is predictive activity with less regulation. This could also explain the often bizarre and illogical nature of dreams, as the brain's predictive models operate without the grounding of real-world sensory data

Hacking the Prediction Machine

He finally encourages us to ‘hack our predictive minds’. These hacks include the use of meditation, technologies such as VR and mixed reality along with AI. Therapies also fall into this category, encouraging us to break the cycle of negative. Cycles of predictive behaviour in conditions such as depression and anxiety.


There is much to be gained from theorists who push the boundaries of the mind into its iterations with the world, others and technology. The Extended Mind is an incredibly useful idea as it explains both why and hoe we should be implementing technology for learning. This book gives us a cognitive bedrock with a computational theory of the mind but is also fruitful in pointing us in the right direction on implementation through task support and performance support.

No comments: