What is Consciousness? A Panel Discussion with Anil Seth and Colleagues at the Francis Crick Institute

In this Francis Crick Institute discussion, Prof Brian Cox hosts a panel of experts to explore what consciousness is, how it arises from brain activity, and whether artificial systems could possess it. The talk covers definitions, neural correlates, memory, hallucinations,动物 and plant consciousness, theories such as panpsychism and integrated information theory, and the question of free will and AI consciousness. It presents wide ranging perspectives from Anil Seth, Katarina Schmack, Steve Fleming and Alex O Connor, highlighting the idea that consciousness may be a brain system level process rather than a single localized state, and raising important scientific and ethical questions about the future of AI and neuroscience.

Introduction and Setting

The program opens with Professor Brian Cox introducing a panel from the Francis Crick Institute who are at the forefront of consciousness research. The aim is to tackle a topic often treated as a philosophical puzzle, yet one for which neuroscience and cognitive science offer concrete, testable questions. The panel includes Anil Seth from the University of Sussex, Katarina Schmack from the Crick Institute, Steve Fleming from University College London, and Alex O Connor, a philosopher and host of a philosophy podcast. Cox frames the discussion as an inquiry into how subjective experience arises from physical brains and what processes can explain the emergence of consciousness while acknowledging that the problem remains one of the most challenging in science and philosophy.

The moderators and panelists set out the agenda: to ask what consciousness is, where it might be located, how it arises, and what explanatory frameworks best account for it. The conversation intentionally begins with concise definitions before moving into empirical questions about brain activity, perception, memory, and cognitive processing as they relate to conscious experience.

Defining Consciousness: Minimal Yet Practical

The panel invokes a classic philosophical touchstone by Thomas Nagel, who defined consciousness in terms of subjective experience: there is something it is like to be a particular organism. Anil Seth emphasizes a practical, working definition: consciousness is the state of being aware and having subjective experiences. Steve Fleming adds the crucial distinction between conscious awareness of something and unconscious brain processing that can still influence perception. Alex O Connor highlights the challenge of remaining faithful to the subjective nature of consciousness while adhering to scientific objectivity. The exchange anchors the debate in a pragmatic stance: science can study correlates and mechanisms without necessarily delivering the full metaphysical account of what it is like to be a bat or a human, yet it can still provide predictive, actionable insights that matter for medicine, psychology and AI.

Can We See Consciousness in the Brain

The question raises the hermeneutic barrier between first person experience and third person observation. Anil Seth describes neural correlates of consciousness as footprints that betray the presence of consciousness but are not consciousness itself. The group discusses how brain regions, energy consumption, and neuronal activity relate to subjective experience, yet the same external stimulus can lead to different conscious perceptions depending on attention, expectation and whether it is noticed. Katarina Schmack discusses how predictions and top-down processes influence perception, and how hallucinations in psychosis can be informative about the brain mechanisms by which experience is assembled. The panelintensely clarifies that correlates are indispensable for scientific progress but do not exhaust what consciousness is, and that reality monitoring plays a key role in distinguishing real from imagined experiences.

Reality Monitoring and Imagination

The discussion delves into experiments showing that imagination and perception can be confounded. An example is the brain’s ability to blend imagined patterns with actual stimuli when asked to imagine and then be presented with real stimuli. The prefrontal cortex is implicated in monitoring the reliability and source of sensory signals, which contributes to the sense of reality. This reveals how the brain constructs a coherent experience by binding sensory data with higher order processes that assess reliability and source attribution. The dialogue also touches on how people sometimes misattribute actions to internal decisions, highlighting the complexity of the subjective sense of agency.

Memory, Time and Consciousness

Memory is presented as a fundamental, but not exclusive, partner of consciousness. The panel references William James and the specious present, the sense of a present that integrates recent past and a sliver of the future. Clive Wearing s case is discussed as a demonstration that consciousness can persist without normal memory formation, challenging simplistic links between memory and consciousness. The participants acknowledge that memory shapes self-identity and continuity through time, yet subjective experience can endure amid impaired memory. The role of memory in the evolution of consciousness is proposed as an adaptive feature enabling the simulation of others and the prediction of future events, a function that could have conferred evolutionary advantage by facilitating social learning and empathy.

Consciousness in Animals, Plants and Non-Human Systems

The panel addresses cross species and cross Kingdom questions. They describe studies in model organisms such as mice, where researchers quantify perceptual reports and confidence to infer hallucination-like experiences. Through brain imaging and optogenetic manipulation, researchers connect dopaminergic signaling and neural circuits to perception in the absence of actual external stimuli. The octopus and other animals are cited as possessing sophisticated cognitive capabilities that invite careful cross-species generalization about consciousness. The speculative question of plant consciousness is discussed with caution; while plants may lack sentience by traditional criteria, the broader question of consciousness as a continuum remains open to rigorous scientific inference. The panel emphasizes a continuum view rather than a sharp threshold between conscious and non-conscious systems.

The Theoretical Landscape: Panpsychism and IIT

The panel explicitly addresses popular theories of consciousness. Panpsychism is discussed as a controversial metaphysical stance that posits some form of mental property at the base of reality. Critics worry that such a view incurs explanatory problems and does not offer testable predictions. Integrated Information Theory IIT is examined as an empirical framework that aims to quantify the degree of consciousness a system possesses, depending on its information integration. The participants stress that IIT is a productive scientific hypothesis that motivates experiments, even if it remains debated. They stress the importance of theories that generate testable predictions and guide experimental design rather than purely philosophical speculation. The discussion acknowledges a spectrum of philosophical positions while keeping a clear-eyed focus on testable science that could eventually yield a robust account of consciousness.

Consciousness, Perception Disturbances and Psychosis

Ruth Goldsmith s question connects subjective experience phenomena such as hallucinations to broader theories of consciousness. The panel agrees that extreme experiences of psychosis illuminate how the brain constructs shared reality and how conscious experience can deviate from conventional reality. The concept of reality monitoring becomes central: distinguishing internally generated experiences from externally triggered stimuli is an essential component of conscious experience. The discussion links these ideas to therapeutic insights, including how manipulating brain activity can alter perceptual experiences and potentially inform new treatments for hallucinations.

Free Will and Decision Making

The Libet experiments are revisited in light of subsequent research that questions the interpretation of readiness potentials as precursors to conscious decisions. The panel suggests that there is a time window in which conscious experience and action align, while the sense of free will may involve retrospective interpretation rather than a strict preconscious veto. Split brain research illustrating the left hemisphere s role as an interpreter who constructs reasons after actions have occurred underscores the complexity of agency. The panel offers a nuanced take: conscious experience contributes to decision making and the sense of agency, but this does not necessitate a simple, linear causal chain from awareness to action.

Artificial Intelligence and Consciousness

Imran s question about AGI touches on a central contemporary issue. The panel considers whether artificial systems could ever possess subjective states or if consciousness is inherently tied to biological life. A spectrum of views emerges: some believe AI could eventually display genuine consciousness as systems achieve multimodal integration, autonomous agency and shared goals; others warn that current AI primarily simulates intelligent behavior without phenomenological experience. There is also a caution about social and ethical risks of imbuing AI with perceived consciousness, including misaligned priorities and the potential for harm. The discussion closes with a recognition that the science of consciousness will continue to intersect with AI as models grow more sophisticated and capable of integrated, time-bound, context-aware behavior.

Ethical and Philosophical Reflections

Across the dialogue, the panel stresses that the scientific study of consciousness should remain anchored in empirical methods while acknowledging the substantial philosophical questions that persist. They emphasize a practical objective: to explain, predict and control aspects of consciousness that have tangible implications for medicine, mental health, education and AI governance. The panelists advocate for cautious advancement, rigorous testing, and thoughtful ethical reflection as we push the boundaries of what we know about consciousness and what we may one day build in machines.

Concluding Thoughts

The discussion does not culminate in a definitive theory of consciousness but instead sketches a roadmap for an interdisciplinary effort that blends philosophy, neuroscience, psychology and AI. The takeaway is that consciousness may best be understood as a system-level process arising from interactions across wide networks in the brain, rather than a single brain region or an isolated cognitive function. The future of consciousness research lies in integrating experimental data with robust theories that generate testable predictions, while remaining mindful of the ethical questions that arise as AI systems grow more capable and the nature of experience remains a central mystery of life.

In summary, the program exemplifies how scientists pursue a deeply mysterious question by combining minimal philosophical definitions with rich empirical investigation, showing that the path to understanding consciousness is as important as the destination itself.