Huberman predicts Meta, OpenAI, Anthropic and Elon Musk's companies will become biotech firms focused on reading and writing the brain, calling it the real AI arms race.
Huberman says he needs about seven hours of sleep, and that nicotine's focus benefits require avoiding chronic use because it is addictive and raises blood pressure.
Huberman says sleep-tech such as a sleep eye mask from an MIT-linked group is about seven to twelve months from release, aimed at Patrick O'Shaughnessy's struggle to fall asleep after exciting days.
Huberman sees AI's real race in reading and writing brains
The next frontier of AI is the brain itself, and Huberman argues the big players will become biotech companies built around reading and writing the nervous system. Current tools see only the cortex surface or act bluntly, and neuroscience still lacks a full model of thought. Measuring waking brain states directly is the hurdle that matters most.
Takeaways
AI labs will become brain reading and writing companies. Huberman predicts Meta, OpenAI, Anthropic and Elon Musk's companies will become biotech companies focused on reading and writing the brain. He calls that the real arms race of AI, and names non-invasive writing of the nervous system the biggest problem.
Today's brain tools are shallow or blunt. EEG and fMRI only capture signals at the cortex surface, and transcranial magnetic stimulation lacks spatial precision. Adderall, caffeine and nicotine raise neuron excitability broadly, while implanted vagus nerve stimulators treat depression, so precision remains the gap.
The brain model is still incomplete. Periphery sensing and the hypothalamus are well understood, and GLP drugs bind appetite circuits. Mark Mayford's mouse work shows reactivating neurons in the wrong order still reproduces behavior, challenging firing order, and what a thought is remains unclear.
Wearable heart data misreads the nervous system. Huberman argues heart rate and HRV are not direct readouts of the autonomic nervous system. Malise Olmes is developing a cuff that measures it directly. Huberman calls measuring waking brain states the most important problem to resolve.
Sleep needs honest self-assessment and a downshift routine. Huberman feels great on seven hours and groggy on eight. Falling asleep requires turning thoughts off, slowing the heart rate with long exhales that engage the vagus nerve, and forgetting limb position. He frames this as shifting attention to sensation, not meditation.
Health discourse will get noisier before it improves. Huberman says public talk about health and longevity is a mess of theatrics with a very low entry bar. He predicts a downturn in public health education before more PhD- and MD-trained educators arrive, and praises Tim Ferriss for staying true to himself.
Success follows genuine curiosity, not borrowed goals. His graduate advisor Barbara Chapman told him success depends on how badly he wants the answers, and he published eight first-author papers in her lab. He says following others' ideas of what to do is a slow death.
What comes next
Sleep technologies such as an MIT-developed eye mask that generates eye movements are seven to twelve months from release.
Tap a name to go there · tap a line to see why they are linked
On this episode 2
Show 1
Within a year of that release, Huberman expects someone could build eyeglasses that stimulate part of the autonomic nervous system.
He predicts combined stimulation and pharmacology for depression and PTSD, building on SSRIs, psilocybin and MDMA, which raise plasticity.
He expects the health space to be noisy in 2027 and 2028 before useful technologies arrive.