1x5 — Karl J. Friston: Human cognition proactiveness and the relationship with our beliefs system
The weak issue concerning what we call “mind” and “body” and their strong bond in experiencing the world, the endless questions about the several interpretations of what we are used to seeing as “reality” or “human being”, the hard analysis of people’s behaviour — which could explain: what does being alive actually means? How do human beings behave in modelling their lives? These questions, despite their apparent “philosophical” nature, could find an important wave of pragmatic understanding through the precious answers of Professor Karl Friston.
He is an authority within the theoretical neuroscience field and has engaged with some of the most important issues — from neurobiology to embodiment, via neuronal interactions. His unifying discoveries and experiments are building very important bridges among the cognitive sciences, medicine, and physics. This interview won’t be able to completely cover Professor Friston’s impressive research, but it will certainly be an important reference for all the designers who are interested in the way people think and act.
I would like really to thank again professor Friston for being available in disclosing the notions, concepts, and discoveries he is dedicating his life to. As always, his answers have been recorded and consequently summarized by me. Good talk!
We open this reflection from the way people behave, that is a particular set of actions whose understanding could be regarded in dualistic ways. Now, do you think we are more instructed by the environment or rather do we produce ourselves through our desires and capabilities?
“Well that is an ancient question and the answer from my point of view is both! But your question calls for some deep answers here. As a psychologist, I would try to understand perception in the brain. And the brain is in the game of providing the best explanation for the causes of its sensations. Here, the brain is insulated from the environment, in the sense that the environment provides sensory data and the brain’s job is to accumulate that sensory evidence and build models of how the environment works and how those sensory outcomes are generated. That is the whole point of active inference and its implicit circular causality. On this view, living is basically the process of gathering evidence for our beliefs. But we clearly change the environment, we act upon the environment. This is a fundamental notion of closing the action perception circle — the environment supplies us with sensations and then we use those sensations to infer states of the environment. In the other direction of fit, a simple example would be changing the direction of our eye gaze: literally palpating the world with our eyes and changing the sensory signals that we are sampling from that world. We can also construct niches of a physical sort and a cultural sort. In other words, not only we are in charge of the way that we gather data from the world, but we can also change the world that is generating those data. So, the brain certainly is constrained and informed and enriched by being immersed in an external world; in particular, an econiche that is adapting to, is enriched by, and inherits form and structure from the individuals that constructed it and populate it.”
Thus, given that embodiment, it could be considered a deeper human factor, psychological and emotional, such imagination is. How do we use our inventive abduction to interact with reality? Is it possible to sketch out a story of free-will hence?
“This question really highlights a very particular and special ability of the human mind, which may be specific to creatures like you and me. This is the ability to generate fictive predictions of the future, as in planning for example. Or a fictive world in terms of exploring multiple hypotheses. From my point of view, everything is inference. And it is really nice you use the word ‘abduction’ because it is the most accurate philosophical way to describe reasoning under uncertainty. In other words, making sense of our sensory exchanges with the world. So, you and I plan to do things, we can survive in the world because we can plan. If we plan, it means that we have beliefs about the consequences of our action. This means we have to entertain multiple counterfactual consequences. And to have that kind of generative model means you have to have an internal representation of the future — a future that is most consistent with your beliefs about the way the world will unfold.
Technically, you have a problem of hypotheses or model selection; and philosophically, you have a ‘choice’. Abduction is also interesting from an evolutionary point of view; where nature has to select which phenotype is most adapted to its econiche. Some people are using chaos theory to do exactly that kind of selection. In terms of free will and imagination, what is clear here is the importance of entertaining multiple counterfactual courses of action — fantasies about the future. We then choose or select (personally or sub-personally) those fantasies that fit our sensory data. On this view, you can call the brain a fantastic organ; it generates fantasy after fantasy, and we are just choosing the one that best matches our sensory data.”
Now, from a designing perspective, such inherent manipulation of sensory data gives action and meaning to the future. So, according to your experience, is that just for survival? Considering the consequences, what ought we pay attention to while we are inventing?
“That is an intriguing question, and I think the notions of circular causality and econiche construction are crucial for any designed environment. In my world, there are what we call the four “Es”: Enactivism, Embodiment, Extension, and Embeddedness. Those come as a bundle in the way that we engage with the world. Enactivism speaks to active perception and the way we make sense of things is inherently enactive; for example, what we see very much depends upon where we look. Embodiment is another take to the fact that we are physically embodied brains. The way we actively sample the world, in pursuing this active perception, active vision, active hearing, has to be enacted within a circular exchange with the world, that feels like an extension of our brain. A classic example here is memory, in the sense you put your mnemonic faculties ‘out there’ into the world. A nice example of this is keeping all your favourite phone numbers in your mobile phone. Finally, the embedded part is the notion closest to the situated cognition, so it is all action!
Now, in terms of econiche construction — that from a cultural viewpoint could include symbolic system construction — is the evolution of certain species in terms of how they change their environment. In short, our environment is indeed designed. And one of the highest forms of environmental econiche construction is cultural. This brings us back to your point, where the best example is language. We have constructed deontic cues, things that go beyond the individual; namely, signs and symbols and semiotics. We build those things as a species that enables us to survive and better model for our lived world. All of this depends upon on the way our embodied brain creates and designs its environment and offloads or extends cognition into that environment. A nice example here is affordance, and the way that environments learn about you, as a manifestation of action upon that environment.”
From such a designing perspective, that active and deliberated manipulation is in the way to trace worlds of meaning which lead us toward alternative possibilities. So, according to your experience when do we decide to stress our beliefs?
“Through this question one can see the central role of abductive processes, in terms of an optimal engagement with the world, especially in the way we interact with novelty. If you can engineer an opportunity to resolve uncertainty about what happens if I do that, doing ‘that’ becomes attractive because it has an epistemic affordance. For example, physical objects, pictures, or signs, invite certain orientating and exploratory responses. There are many affordances, but they are all examples of abductive inference at the end of the day. This can be described mathematically as resolving uncertainty. And that means certain deontic cues, or certain opportunities, would present themselves as affordances for resolving uncertainty. As a psychologist, I think 99% of the affordances we respond to are of this epistemic sort. We watch films, we tell each other stories, we read books, we watch the news, all of this is in the service of resolving uncertainty about the world around us. For example, we are addicted to the news, we all have a nice warm feeling when the BBC news at 10 starts, just because we know we will resolve some uncertainties about what happened today.
Sometimes people have overly precise beliefs that need to be dispelled by giving them alternative hypothesis, by broadening their space of hypotheses, such that they can explore and choose alternative ways of making sense of the world. And sometimes that needs somebody else to provide those alternatives: it could be a designer, it could be an architect, it could be a poet, it could be a novelist, it could be a teacher, or it could be a psychologist!”
Finally, we conclude right with culture and that powerful connection with our evolution. In your thought, how is our relationship with technology going? Do you see any upcoming human being reframe?
“This is a very broad-reaching question, and from my perspective, one of the key elements is connectivity. Today, I see connectivity put under pressure. Unbridled access to information and data puts pressure on our evolved and encultured models of interaction with others. Before the electronic age, there was only a limited number of people you can talk. But now you have a selection problem, a choice problem, because you can call on so many people and different sorts of evidence. In other words, the choices about how we gather our evidence for our world models are now vast. This requires us to harvest and accumulate many sources of information, or sensory evidence, in a way that we think will reliably inform our beliefs. If you give people a hundred buttons, a hundred channels, a hundred blog sites, then there is an irreducible uncertainty about what to do next; and that uncertainty is experienced as stress, which could become an illness.
I think we are already starting to see a pressing problem in the way we manage globalization, connectivity, and the exchange of money and information. Conversely, I see segregation and isolationism that is trying to return to an easier problem in terms of abductive inference. For example, ensuring you children have limited access to smartphone; just seeing BBC and not FOX news; reading only good books; only talking to people that you know… All of these things speak to the same issue, restricting the field of choice. Machine learning and artificial intelligence — that underwrites a lot of technical culture — is very good at assimilating a vast amount of data, but they are very bad about choosing which data to select: they don’t know which is the good data and which are the bad data. So, I think there may be a reversion to a simpler way of living, and this will entail less connectivity, and less globalization.”
