The Unified Science of Mind and Brain: Bridging Psychology, Neuroscience, and AI with Matt Botvinick

Matt Botvinick, Director of Neuroscience Research at DeepMind, discusses the current state of understanding the human brain, highlighting a significant gap between high-level functional comprehension and detailed neuronal mechanisms. He advocates for a unified approach to psychology, cognitive science, and neuroscience, viewing the brain's primary purpose as producing adaptive behavior from perceptual inputs. Botvinick argues that understanding the mind fully necessitates mapping psychological phenomena onto neural events, rejecting the idea of understanding the brain without considering the behavior it generates, or studying neural activity without a clear explanatory target.

Botvinick acknowledges the value of psychology as a discipline, stating he has learned more from it than from neuroscience. He reflects on the historical use of metaphors in psychology, drawing a parallel to Mendelian genetics preceding the discovery of DNA's physical structure, suggesting that such abstractions are valuable for guiding the search for underlying physical mechanisms. He expresses a personal dissatisfaction with remaining solely at a metaphorical level of description, emphasizing the ultimate goal of understanding how mental activity arises from neural activity. He also critiques traditional psychology's limitations, such as small sample sizes and highly controlled lab settings, which can miss the richness of real-world human behavior, while expressing excitement for modern data sources like internet behavior.

He recounts his entry into science through connectionism (now deep learning), which he found compelling for its focus on the richness and complexity of human cognition, exemplified by debates on past tense formation. Botvinick notes that current deep learning models, despite their advancements, often lack the flexibility and adaptability characteristic of human behavior. He questions whether the key differences between artificial and biological neural networks lie at the lowest neuronal level or higher levels of abstraction, but ultimately views the brain as a computational device, albeit one whose full computational richness remains largely unexplained.

Botvinick is captivated by the paradox of the brain's intimate closeness yet profound mystery, as it is responsible for the 'full transparency of everyday life' and the entirety of human experience, including emotion and the 'magic' of language. He finds discussions of artificial intelligence often too narrow, failing to capture the full scope of human experience. He also touches upon the fascinating aspect of collective intelligence and group behavior, seeing it as fundamental to human cognition and intelligence, where large numbers of humans converge on ideas in a distributed manner.

I think we're at a weird moment in the history of neuroscience in the sense that there's a there I feel like we understand a lot about the brain at a very high level but a very very coarse level

to me the point of neuroscience is to study what the brain is for

I don't know what it means to understand the brain if there's no if part of the enterprise is not about understanding the behavior that's being produced

these psychological phenomena are can be explained through a very different kind of causal mechanism which has to do with neurotransmitter release

I have learned much more from psychology than I have from neuroscience

the cost of doing highly controlled experiments is that you by construction miss out on the richness and complexity of the real world

the paradox that lies in the fact that the brain is so mysterious and so it seems so distant but at the same time it's responsible for the the the the full transparency of everyday life

one seeming limitation of the systems that we're building now is that they lack the kind of flexibility the readiness to sort of turn on a dime when this when the context calls for it that is so characteristic of human behavior

I'm happy to think of it as just basic computation but mind you I won't be satisfied until somebody explains to me how what the basic computations are that are leading to the full richness of human cognition

the music of the language the wit the something that makes a rich experience something that would be required to pass the spirit of the Turing test is lost in these benchmarks