#849: Dr. Michael Levin — Reprogramming Bioelectricity, Updating "Software" for Anti-Aging, Treating Cancer Without Drugs, Cognition of Cells, and Much More

Hey folks, it’s Tim. Today’s guest is Dr. Michael Levin, a Tufts biologist whose lab is reshaping how we think about bodies, healing, and even minds through bioelectricity. Quick note before we dive in: I did invest in a startup based on some of his work in 2023, but we barely touch companies here—this is about peeking around the corner at what’s next.

You and I both found Robert Becker’s book as kids. What was the book, and why did it matter to you?

The Body Electric. I stumbled on it in the eighties and was blown away by the older literature it surfaced—evidence that electricity guides development and repair in ways I’d been imagining but hadn’t seen documented.

Let’s define terms. What is bioelectricity in a useful way for listeners?

Living systems use electrical signals to coordinate. There’s the familiar neural kind that binds neurons into a mind, and a developmental kind that existed before brains, where tissues use voltage patterns to build and maintain form.

Could you ground this with experiments from your TED talk that make it tangible?

Tadpoles with scrambled facial parts still self-correct to a normal frog face, and flatworm fragments decide how many heads to make. These tissues hold target-shape memories in bioelectric networks, and we can rewrite those memories to drive regeneration and repair.

Most people learn DNA to RNA to protein and call it a day. Yet you’ve made two-headed worms by tweaking electrical states, not genes. What does that imply for the textbooks?

Think hardware versus software. The genome specifies hardware. Voltage patterns are software—reprogrammable memories of target anatomy. We can flip the worm’s default from one head to two without touching DNA, and cells interpret genes with creativity rather than blindly following them.

You’ve said you can “see” these memories. What does that look like, and why does the two-headed trait persist?

We use voltage-sensitive dyes to map tissue-wide electrical states and watch them evolve over time. By decoding these patterns as set points, we can change a latent “correct” body plan and see it expressed upon injury. Some memories lock in like a thermostat; others revert after weeks, suggesting higher-level error correction.

Bridging to humans—think vagus stimulation and its impact—how does this apply to people, and what might therapies look like?

These mechanisms are conserved across species, including humans. Disclosures: some of our tech is licensed to companies working on limb regeneration and aging. Three big targets: correcting birth defects by restoring electrical patterns, guiding regeneration by “telling” cell collectives what to build, and normalizing cancer by electrically reconnecting cells to collective goals. We’re also probing aging.

Let’s go deeper on aging. What are you seeing, and how does it shift the frame?

With age, tissue voltage maps that encode form get fuzzy. Beyond damage or programmed decline, we see a third idea: goal-seeking collectives lose shared targets and cohesion over time. In simulations without noise or selection for death, systems still decay after “finishing” their goal—call it a somatic boredom effect. Planaria avoid this by constantly re-challenging themselves, and in old tissues we see gene-expression drift toward earlier evolutionary states.

If you had a Manhattan Project for aging, where would you aim the talent and money?

Put it all on multicellular cognition: how cell groups represent goals, negotiate, and can be steered. Crack communication with the collective, and regeneration, cancer control, and healthier aging follow as side effects.

Give us a concrete example of issuing a new directive to cells.

We set a regional voltage pattern in a frog embryo that says “build an eye here.” Cells argue a bit—skin versus eye—then align and construct a functional eye, no micromanaging genes required. For humans, maybe periodic tune-ups to reinforce the body plan; if that hits limits, perhaps intentional body-plan upgrades.

Do you think we evolved to age and die? Any strong cases for that?

There are trade-offs. Early mammals needed rapid scarring over slow regeneration. Yet deer antlers and the liver prove powerful mammalian regrowth. I don’t think mortality is inevitable if we master regeneration, though we still need to learn whether age-related mental rigidity is hardware or software.

From your cross-training in CS and biology, what ideas do you wish each field would import from the other?

Our formal models miss a lot. Even simple systems can show proto-cognitive competencies that aren’t captured by the code. I think cognition is broader than life and predates it; the line between living and machine isn’t as sharp as we pretend.

Curveball: does anything in acupuncture or meridians overlap with bioelectricity?

Clinically, I’ve seen striking results. My hunch is acupuncture touches a layer above bioelectricity—maybe biomechanical or informational—that couples into electrical states rather than directly being the same thing.

Placebo keeps showing up in surprising ways. How do you view it in this context?

Placebo isn’t just a confound—it’s a window into how high-level intent drives cellular chemistry, as everyday voluntary motion already proves. Words can trigger the same downstream pathways as drugs, so we should study how intent propagates through electrical and non-electrical layers.

What about the experimenter effect in animals—what could be driving that?

Biology excels at credit assignment and selective attention where robots fail. Rats can learn to modulate internal variables for reward. Planaria hit with barium lose their heads, then regrow barium-proof heads by dialing a tiny set of genes among thousands—fast, targeted problem solving without prior exposure.

Project out ten years. How might our view of cognition and neuroscience change?

Expect a shift to diverse intelligence—minds without neurons and minds beyond biology. Neuroscience becomes the study of architectures that bind simple parts into larger agents. And we must face anomalies, like people with minimal brain tissue functioning normally, which our current models don’t predict.

How do you think about consciousness, if at all? And where does that take you?

It’s not my main research, but I see it as graded, not binary. Other organs could host minimal first-person perspectives. I’m exploring a Platonic space of mathematical patterns that constrain physics and enable biology; consciousness might be the vantage point of a pattern projecting through a physical interface.

You teased compute. What’s the core idea there?

Polycomputing: one physical process can compute multiple things depending on how you read it. Even tiny sorting algorithms show intrinsic behaviors and side computations “for free,” hinting at unseen capacity. For AI, language may be a sideshow compared to what the system actually “wants” to do.

If that’s real, how might it change big-compute strategy, and do side activities support the main task?

Some side computations may be useful; others may be irrelevant or risky. We’re building tools to detect, encourage, or suppress them and testing whether these layers are independent or entangled.

Before we pause, any sci‑fi you love that helps people think more flexibly?

I grew up on the classics. Stanisław Lem is a favorite—Solaris and his hilarious, mind-bending shorts. Terry Bisson’s They’re Made of Meat is a gem, and Clarke’s tale of dense beings from Earth’s core flips our assumptions about what counts as real agents versus fleeting patterns.

You’ve mentioned the blog a few times, and I’ve shared your advice-to-students piece before because it travels well beyond academia. For folks who enjoyed this chat, where should they start—any one to three posts you’d point them to?

I’ve got a starter-pack post and a few curated links. I’ll send them over so you can drop them in the notes.

We’ll include those in the show notes. As we wrap, pick your lane: something you took from Daniel Dennett, that memorable Clifford Tabin toast about being most likely to crash or to change the field, or a giant metaphorical billboard to get a message in front of biology departments.

On Dan: he was a generous thinker, and he pushed me to strengthen an opposing view before critiquing it. Really understand the best version, then, if needed, dismantle it.

Quick context for listeners—how would you frame Dan Dennett in brief?

He was widely regarded as one of the most important living philosophers, based at Tufts, with a long shelf of influential books. That rigor around not knocking down a flimsy target has stayed with me. On Cliff Tabin’s line: he was my PhD advisor, and I think he meant I’m willing to entertain and voice ideas that cut against the grain. That’s risky because in science most ideas are wrong in some meaningful way, and I’m fine stating my current best view strongly while knowing we may revise it. Sometimes that contrarian stance lands, often it doesn’t, and the crash-and-burn possibility never fully disappears.

Before we close, let’s point people to the right places. ThoughtForms dot life, DRMichael11 dot org—any other homes or social?

On X, I’m dr mic 11. ThoughtForms dot life is my personal blog where I say things I wouldn’t put on the official lab site. DRMike11 dot org is the lab hub with papers, software, and datasets—the receipts behind the wild claims. There’s also a YouTube channel with recorded conversations from the last several years, and I’ll send you that link.

Great, we’ll track it down. Mike, thank you—let’s do this again. For everyone listening or watching, we’ll link to everything at tim.blog slash podcast—just search Michael Levin, L E V I N. Plenty there for deeper dives, more thinking, and some assumption-bending. Until next time, be a bit kinder than necessary to others and to yourself. Thanks for tuning in. Bible Friday is easy to sign up and cancel. It’s a short Friday note with the most interesting things I’m exploring—articles, books, music, and tech that friends and guests send my way. If that sounds fun, go to tim.blog slash friday, drop in your email, and you’ll get the next one. Thanks for listening.