Summary of "Free Will Absolutely Does Exist with Dr. Kevin Mitchell - Factually! - 248"

Episode context

Host Adam Conover interviews neuroscientist Dr. Kevin Mitchell about his book Free Agents: How Evolution Gave Us Free Will.
The conversation is framed as a response to a prior episode with Robert Sapolsky, who argued (from neuroscience/physics) that free will is an illusion.

Mitchell’s central claim

Free will can be naturalized: agency and a meaningful kind of control emerge from evolved biological organization — no supernatural “ghost in the machine” is needed.

Biology and physics do not, by themselves, negate meaningful agency. Evolution produces systems that pursue goals; that goal-directedness grounds meaning, representation, and causal efficacy at higher levels.

How Mitchell builds the account (evolutionary / systems framework)

Start at the origin of life: living systems are organized, boundary-defined, anti-entropic processes that persist by doing work. Persistence (survival) supplies a primitive “goal.”
To meet environmental change, organisms evolved control systems that:
- sense information,
- evaluate it relative to goals,
- act to restore or maintain favorable states.
Information and meaning matter: living systems treat sensory input as meaningful (food, threat, mate), not as bare physics. That gives higher-level explanatory principles beyond low-level physics.

Examples and scaling up:

Bacteria: receptors detect chemicals and bias movement toward food / away from danger — an instance of adaptive control / primitive agency.
Worms: limited sensory horizons and simple repertoires show basic decision-making (e.g., wiggle forward/back to find food or mate).
Nervous systems and brains add layers of processing, creating representations decoupled from immediate stimuli. Vision and hearing enable distant sensing and planning, extending temporal horizons.
Neural activity instantiates patterns that represent objects, states, and goals; these representations have causal power at the organismal level even though they’re implemented in wet matter.
Metacognition and language: humans can reflect on and reason about their own thoughts, adopt policies/goals, and thereby exert higher-order control over future behavior.

Response to determinism and Sapolsky-style objections

Mitchell rejects two extremes:

Magical absolutist free will (complete freedom from all prior causes) — incoherent for evolved organisms and unnecessary.
Absolute denial of agency (all behavior is just deterministically produced by prior physical states) — ignores levels of organization and empirical indeterminacy.

Technical and conceptual points against strict physical determinism:

Multiple levels of organization: higher-level constraints and meanings can have causal influence when low-level laws underdetermine outcomes.
Indeterminacy and noise:
- Neurons and biological processes are noisy; variability is inevitable and functionally useful (exploration, adaptation).
- Chaotic nonlinear classical dynamics (sensitive dependence on initial conditions) and quantum indeterminacy mean the future isn’t a single pre-determined timeline in practical or conceptual terms.
The “God’s-eye” determinism assumption requires infinite precision in state specification (an unrealistic premise). The universe’s finite information capacity makes that picture implausible.
Practical point: even if some causal history shapes agents, that doesn’t remove the capacity for organisms to exercise rational control in many contexts.

Degrees of freedom, responsibility, and moral/legal consequences

Free will is not all-or-nothing; it exists in degrees.
Different agents and situations afford different degrees of rational control (e.g., children vs. adults, addicts, people with severe mental illness, people in poverty).
Acknowledging causal influences doesn’t force discarding responsibility; it refines how responsibility and mitigation should be considered. Legal systems already apply graded judgments (e.g., insanity pleas, diminished capacity).
Denying any agency at all removes useful distinctions we use socially and legally to judge and respond to behavior.

Everyday phenomenology and pragmatic consideration

Our lived experience of deliberation, reasons, planning, and reflection are not mere illusions; they’re higher-level processes that do explanatory and causal work.
For practical decisions (insanity pleas, punishment, rehabilitation), we already operate with a graded, pragmatic framework for responsibility. Theoretical metaphysics shouldn’t invalidate those practices.

Analogies and supporting illustrations used

Computer/electrons: electrons follow physical laws, but organization (hardware/software) imposes higher-level behavior; similarly for brains.
Weather forecasting and chaotic systems: good short-term predictability but increasing uncertainty farther out; neural computations are similarly complex, noisy, and sensitive to small differences.
AI language models: complex systems that are partly unpredictable; variability is part of how they function and mirrors biological unpredictability in some ways.
Quantum mechanics: wavefunction evolution is deterministic, but measurement (or interactions) yields probabilistic outcomes — leaving room for underdetermination of specific future states.

Practical upshots and methodological takeaways

When discussing free will, avoid a binary framing. Instead:

Recognize multiple explanatory levels (physics, biology, cognition) and that higher-level organization can have causal efficacy.
Consider degrees of control: context, cognitive resources, developmental stage, illness, addiction, stressors, etc.
Distinguish phenomenology (what people experience: deliberation, reasons) from metaphysical absolutism; take phenomenology seriously as explanatorily relevant.
Use evolutionary and systems-level analysis: start from persistence/goal-directedness and trace how information, sensing, control, and representation evolve and scale up.
Inform policy and moral decisions with nuanced scientific understanding rather than simplistic determinism, since mitigation and responsibility judgments require appreciation of degrees of agency.

Conclusion

Mitchell offers a middle-way, naturalistic, evolutionary account where free will is an emergent property of organized, information-processing, goal-directed biological systems.
He acknowledges open questions and the need for more detail, but argues this framework is consistent with neuroscience and physics (properly understood), and with our moral practices.
Adam Conover finds the view compelling and notes the topic remains contentious and intuitively difficult; pragmatically, society already operates on graded notions of control and responsibility.

Practical summary checklist (for evaluating claims about free will)

Ask what notion of free will is being used (magical absolutist vs. evolved, graded agency).
Identify the relevant level of explanation (physics, cellular, neural, cognitive, social).
Look for whether the system has:
- boundaries and mechanisms for persistence (life),
- sensors and control loops,
- representations decoupled from immediate stimuli,
- ability to plan over time and use language/metacognition.
Consider sources of variability/indeterminacy (neural noise, chaotic dynamics, quantum indeterminacy).
Assess degrees of rational control available in the context (developmental, pathological, situational constraints).
Translate conclusions into pragmatic implications for moral/legal responsibility rather than all-or-nothing metaphysical verdicts.

Speakers and sources featured

Adam Conover — host (Factually)
Dr. Kevin Mitchell — guest; professor of genetics and neuroscience, Trinity College Dublin; author of Free Agents: How Evolution Gave Us Free Will
Dr. Robert Sapolsky — referenced (previous guest with opposing view)
Carlo Rovelli — referenced (quantum physicist mentioned)
FastMail — sponsor mentioned
Princeton University Press — publisher of Mitchell’s book
Other mentions: producers and show staff named in credits (April Nicole, Sui Sue, Sam Rodman, Tony Wilson)