Charan Ranganath
What this covers
Charan Ranganath builds an argument that memory operates not as a video recorder but as a prediction machine, constantly filtering experience to extract what is biologically useful for navigating the present and anticipating the future. In conversation, he walks through how the brain selectively encodes moments of surprise, significance, and high prediction error rather than treating all experience equally — which explains both ordinary forgetting and the malleability of what we think we remember. The hippocampus and its partner networks don't store faithful replays; they build internal models of events by mixing what actually happened with inferences used to fill gaps, much as a scientist uses theory to interpret data. Because imagining and remembering share the same neural machinery, the boundary between true and false memory is blurry, and a remembered experience can be subtly or substantially distorted simply through the act of recalling it.
Ranganath moves across several interconnected territories. He describes how event boundaries — like stepping from one room into another — partition continuous experience and correlate strongly with what we later retain. He traces social mechanisms through which misinformation spreads, showing how shared memories become "our memories" and how interrogation techniques can engineer false confessions by weakening the prefrontal cortex and feeding suspects imagined narratives. He covers why the testing effect beats restudying, how spaced repetition works through context mismatch, and why task-switching fractures memory formation. On technology, he argues the risk lies not in what AI will do but in how people adapt maladaptively to it — pursuing convenience over the productive discomfort that drives learning. Throughout, he frames forgetting not as failure but as design, since a brain that retained everything would be cognitively paralyzed.
Ranganath argues that memory is not a recording of the past but a reconstructive, prediction-driven system optimized to extract what is biologically useful for understanding the present and forecasting the future — which is why we forget most things, distort memories, and can be made to remember things that never happened.
- Memory is biased toward beginnings, ends, peaks, and points of high prediction error/surprise rather than faithful replay
- The act of remembering itself can change, strengthen, or distort a memory, enabling false memories and social contagion
- Imagination and memory share the same brain machinery (hippocampus and default mode network), so memories are constructions vulnerable to confusion with the imagined
Memory encodes biased samples—peaks, troughs, and boundaries—that drive decisions despite feeling accurate.
- The way we remember experiences is not a replay of the experience but something biased toward the beginning, end, and peak/best/worst moments; because we consult memory when making decisions and feel it is an accurate record, our decisions are driven by this biased but biologically relevant sample rather than what we actually experienced.
“the way we remember things is not the replay of the experience, it's something totally different. And it tends to be biased by the beginning and the end, and he talks about the peaks”
- A major discovery is that we break continuous experience into discrete events at event boundaries (like crossing from one room to another, explaining the 'why did I come in here' phenomenon), and there is a strong correlation between hippocampal activity at these singular event boundaries during a movie and a person's later ability to remember a story outside the scanner — even independent of age, though hippocampal boundary activity declines with age.
“you find this incredible correlation between the activity in the hippocampus at these singular points in time, these event boundaries. And your ability to just remember a story outside of the scanner later on.”
- Subjective time is compressed the farther back we look — one hour ago feels very different from two hours ago, but one year ago versus one year and one hour ago feels identical — and this compression is driven by memory and context change rather than physics; during the pandemic, people felt days dragged (monotonous within an event) while weeks flew by (no event boundaries to mark them), because eventful periods create more memories that make a span feel substantial in retrospect.
“Almost everyone in the class said that the days were moving by slower... the majority of them said that the weeks were passing by faster.”
- Rather than encoding episodic memories constantly (which would be massively redundant), the brain forms internal models of events and optimally encodes episodic memory at points of high prediction error, uncertainty, surprise, or motivational significance — these event boundaries show heightened hippocampal connectivity and a bump of activity associated with better memory.
“if you capture an episodic memory at the point of maximum uncertainty... or maximum surprise, you have the most useful point in your experience that you've grabbed.”
Learning happens through prediction error and surprise, detected at event boundaries and encoded via neuromodulators.
- Neuromodulators like norepinephrine, dopamine, and serotonin are released during biologically significant moments such as surprise, fear, and stress, and these chemicals promote lasting plasticity — a mechanism by which the brain prioritizes which information to carry forward; the brain achieves this efficiency on remarkably low power (estimates of 12-20 watts).
“neuromodulators, for instance, these chemicals in the brain like norepinephrine, dopamine, serotonin. These are chemicals that are released during moments that tend to be biologically significant, surprise, fear, stress, et cetera. These chemicals promote lasting plasticity”
- The testing effect — that testing yourself improves retention more than restudying — works because the error and mismatch from being tested exposes which connections are good and bad, conceptually similar to backpropagation in neural networks; you retain information better by constantly pushing yourself to your limit, and if you feel like you're coasting you are not learning.
“it's actually conceptually similar to what happens with backprop in AI or neural networks. So the idea is that you expose, "Here's the bad connections, and here's the good connections."”
- Memory is fundamentally about the present and future, not the past; the brain is optimized to learn the stuff from the past that is most useful for understanding the present and predicting the future (such as cause-effect relationships), and we constantly make memory-based predictions whose errors are where learning happens.
“really memory is all about the present and the future. The past is done. So biologically speaking, it's not important unless there's something from the past that's useful.”
- Even people who claim their memory is terrible are, in the moment, overconfident about how much they will retain later; self-testing reveals this gap painfully ('I thought I knew that, but I don't'), which can be demoralizing until you realize that this struggle is how you learn best.
“Even though everyone tells me, "Oh, my memory is terrible," in the moment they're overconfident about what they'll retain later on.”
Shared memories create tribal identity and social connection, enabling both cohesion and perspective-dependent narrative conflict.
- Going through difficult or near-death experiences with someone creates bonding, and the act of retelling the story repeatedly to others makes it a better and better story while shaping the memory — there is no point in suffering unless you get a story out of it.
“I like to say it's like there's no point in suffering unless you get a story out of it.”
- People's sense of collective identity is tied to shared memories — remembering the same things in the same way makes us feel socially connected and part of a tribe — and because we selectively retrieve memories that align with our existing beliefs and goals, narratives of the past (like the 'Vietnam War' versus 'American War') are perspective-dependent, which both enables tribalism and offers the opportunity to expand understanding by genuinely listening across perspectives.
“people's sense of collective identity is very much tied to shared memories. If we have a shared narrative of the past... we will feel more socially connected with each other”
- The foibles of individual memory are magnified socially through 'social contagion,' where misinformation spreads like a virus: once two people share a memory it becomes 'our memory,' and the more you trust or the more powerful a person is, the more they shape the shared narrative — a mechanism exploited at scale by propaganda machines and even by political push-polls that embed false claims into questions.
“There's a whole literature on something called social contagion, which is basically when misinformation spreads like a virus”
- Because the remembering self is what gives lasting satisfaction, you should deliberately do memorable things — such as taking vacations to interesting places — even when they are unpleasant in the moment, since the experiencing self suffers but the remembering self is glad you did it.
“do it even if it's a pain in the ass because the experiencing self will suffer for that but the remembering self will be like, "Yes, I'm so glad I did that."”
- Heartbreak is remembered intensely partly because memory captures biologically significant attachment, and the loss reveals all the good things in a new light; rather than just suffering, the moment of losing love is a good occasion to viscerally re-experience and appreciate memories more deeply, and therapy often helps by offering a different perspective to reframe the past rather than imposing changed memories.
“so much of the therapy process was guided towards reframing and getting people to look at the past in a different way, not by imposing changing people's memories... but just offering a different perspective”
- Nostalgia — originally coined by a Swiss physician as a disease afflicting traumatized soldiers — is double-edged: it can be a source of happiness and improved mood, but if it leads to selectively remembering 'the good old days' as gone forever and narrows your worldview, it becomes toxic at both personal and collective levels, fueled by how selective our remembering self is.
“nostalgia can be an enormous source of happiness... but it just depends on what they do with it because what you can sometimes see is nostalgia has the opposite effect of thinking those were the good old days, and those days are over.”
The brain organizes distributed memories by shared features like location and people present as reusable building blocks.
- Different memories produce distinct patterns of brain activity that can be decoded almost like a QR code — what was once dismissed as noise; by analyzing whether memories that occurred in the same place or with the same person share similar patterns, researchers can determine the brain's 'sorting scheme,' finding that some networks organize memory by who was present and others by location, providing reusable building blocks for constructing the present, imagining, and remembering.
“all that little stuff that we use to think of noise, we can now think of almost like a QR code for memory, so to speak. Where different memories have a different little pattern of bright pixels and dark pixels.”
- Unlike rats forming a fresh hippocampal map of each box, humans economically reuse past experience to learn new environments — once you learn one Ikea layout you can navigate any Ikea — and the brain represents space generatively as bits and pieces (intersections and turns) rather than as literal Euclidean coordinates, with hippocampal ripples increasing at navigationally important decision points and goals.
“if you learn a map of an IKEA... Once I learned my local IKEA, I can use that map everywhere. Why form a brand new one for a new place?”
- Developing expertise is fundamentally about training attention — learning what to look for and what is important rather than passively absorbing patterns; their study of expertise found that as people became experts in identifying novel objects, the biggest effects were massive increases in prefrontal cortex activity, not just tuning of the visual cortex, consistent with chess experts and athletes who look several steps ahead.
“as they developed this expertise of being able to identify what made them different from each other and what made them unique, we were actually seeing massive increases in activity in the prefrontal cortex.”
- During slow-wave sleep, sharp-wave ripples in the hippocampus replay compressed sequences of activity and open windows of communication with the neocortex; this helps form and stabilize memories but, importantly, also connects different events together so you can build internal models of how things behave, find the signal in the noise, and see connections across experiences.
“during those moments, there's a little window of communication between the hippocampus and these areas in the neocortex. And so that I think helps you form new memories, but it also helps you, I think, stabilize them, but also really connect different things together in memory.”
- Because memories compete with each other, processing information in a distinctive way (like a hot pink Post-it standing out among yellow ones) makes a memory easier to retrieve; combined with organization — the Memory Palace works by giving you a pre-existing structure to organize and link arbitrary information so you can find the cue you need.
“If it's hot pink, it's going to stand out and I find it really easily. That's one way in which if things are distinctive, if you've processed information in a very distinctive way, then you can have a memory that's going to last.”
Memory is construction using sensory detail and context to distinguish experience from imagination, a slow and imperfect process.
- Distinguishing memories of things that happened from things we merely imagined is a hard problem because both are mental constructions; we succeed (imperfectly) by using the prefrontal cortex to focus on sensory detail and the place-in-time context, since genuinely experienced events carry more vivid detail than imagined ones — but this process is slow and effortful and doesn't always work.
“the way that we often do it and succeed is by, again, using our prefrontal cortex and really focusing on the sensory information or the place in time and the things that put us back into when this information happened.”
- There is no clean distinction between true and false memories; every memory is a mix of bits of what happened plus our inferences and theories used to fill gaps, just as a scientist uses theory to interpret data. Minor distortions arise from filling blanks with meaning, while larger false memories form when imagined content or externally supplied misinformation gets incorporated because the act of remembering itself changes the memory.
“I like to say there's no such thing as true or false memories... we have these memories that reflect bits and pieces of what happened, as well as our inferences and theories.”
- Tyler Bonin's work found that the exact object-perception tasks where state-of-the-art computer vision models struggle (regardless of added parameters or layers) are the same ones where humans with damage to the perirhinal cortex struggle; humans only succeed if given enough time to move their eyes and sample, because unlike computer vision models that see all pixels at once, humans build perception through foveated eye movements over time.
“he found those were the exact ones where these humans with particular damage to this area called the perirhinal cortex, that was where they were struggling.”
- Imagination is the opposite operation of remembering — taking information from memory and recombining it in new ways — but both are largely constructions in the head using the same machinery; brain imaging shows that imagining possible future scenarios activates nearly the same maps (the default mode network closely interacting with the hippocampus) as remembering the past, which is why amnesiacs with hippocampal damage struggle to imagine vivid scenarios.
“if you look at brain images of people... imagining possible scenarios of things that could happen in the future... they look very similar. To the naked eye, they look almost the same as maps of brain activation when people remember the past.”
Episodic memory plus semantic knowledge solves stability-plasticity: surprising events form memories that supplement knowledge for novel reasoning.
- Human intelligence is powered by having both a 'dumb' episodic memory that records that something happened once at a point in time, and accumulated semantic knowledge; when a surprising situation violates expectations, an episodic memory is formed that can later supplement knowledge to reason about novel cases, letting us stop on a dime and change without erasing everything learned — a powerful solution to Grossberg's Stability-Plasticity Dilemma that lets us learn from far less data.
“the next time we're in a similar situation, boom. We can supplement our knowledge with this information from episodic memory and reason about what the right thing to do is. So it gives us this enormous amount of flexibility to stop on a dime and change, without having to erase everything we've already learned.”
- Adolescence and the teenage years are far more important for brain development than commonly credited; much mental illness like schizophrenia (now viewed as a neurodevelopmental disorder) emerges during this period of late adolescence and early adulthood when prefrontal cortex connections are still being reformed.
“the teenage years are just so important for the brain. And so that's where a lot of mental illness starts to emerge. Now we're thinking of things like schizophrenia as a neurodevelopmental disorder because it just emerges during that period of adolescence and early adulthood.”
- While mouse models are invaluable for studying targeted neural circuits and conserved learning mechanisms, humans differ fundamentally from rodents: rodents explore largely through smell with wide-field vision while humans use foveated eye movements for high-resolution sampling, humans have language and social structures, primate brains integrate information across far longer timescales, and human training data is vastly richer — so you cannot learn everything about human memory from a mouse.
“primate brains and human brains are especially good at integrating and holding on to information across longer and longer periods of time.”
- Contrary to the traditional view that working memory (short-term memory) is independent from long-term memory, much of executive function actually requires learning and synaptic change to occur, meaning the two systems are deeply interconnected rather than separate.
“a lot of my early work was basically saying that this working memory, which some memory researchers would call short-term memory is not at all independent from long-term memory. That is that a lot of executive function requires learning, and you have to have synaptic change for that to happen.”
Forgetting results from competition between overlapping memories, absence of retention signals, and retrieval failure.
- Memories are not stored as individual discrete items but are shared across distributed pools of neurons, so overlapping memories compete and can wipe each other out; we forget for three reasons — that competition, the absence of biological signals that promote long-term retention, and retrieval failure (not finding the right cue to reactivate the memory).
“in the brain, the way it works is you have this distributed pool of neurons and the memories are kind of shared across different pools of neurons. And so what you have is competition, where sometimes memories that overlap can be fighting against each other.”
- We don't remember the first few years of life because the hippocampus (critical for episodic memory) is still developing, and because the neocortex is so plastic and rapidly rebuilt during that period that the neural pathways to early memories get completely rerouted — analogous to bombarding a trained neural network with a year's worth of new data and then trying to retrieve the first patterns.
“the brain is so plastic, the cortex is so plastic during that time, and we think that memories for events are very distributed across the brain. Imagine you're trying to get back that pattern of activity that happened during this one moment, but the roads that you would take to get there have been completely rerouted.”
- Memory athletes who can memorize a deck of cards or an entire Ikea catalog were not born with extraordinary memory; they practice learnable strategies like the Memory Palace over and over again, demonstrating that elite memory performance comes from technique rather than innate ability.
“By all accounts, from people who become memory athletes, they weren't born with some extraordinary memory, but they practice strategies over and over and over again.”
- Spaced repetition improves long-term retention compared to massed study because each repetition occurs in a different context, and the mismatch between the recalled memory and the current context creates error that strengthens the content while detaching it from any single original context — the opposite of vivid episodic memories triggered by a specific smell or song; cramming, by contrast, helps short-term recall but leads to faster forgetting.
“if you space things out, you get a benefit for later on retention.”
- The goal of memory improvement should not be remembering more but remembering better — focusing on what is important — because the brain is designed to forget; Ebbinghaus showed people lose about 60% of arbitrary information within a day, and no one has violated this basic tendency, so expecting to remember everything is the wrong standard.
“you don't want to remember more, you want to remember better, which means focusing on the things that are important. That's what our brains are designed to do.”
Prefrontal cortex balances attention focus and flexibility; disruption in ADHD mirrors imbalance affecting decision-making.
- ADHD is associated with differences in prefrontal function such that attention is both more distractible (shifting too easily, harder to focus on what's relevant) yet capable of getting stuck (hyperfocus) on things of interest; healthy attention is a balance of focusing when needed and shifting when needed — the same flexibility-plus-stability balance that appears disrupted in ADHD.
“ADHD is associated with differences in prefrontal function, so that attention can be both more, you're more distractible... But then, once you get on something that you're interested in, you can get stuck.”
- False confessions are produced by a reliable recipe: an authority figure insists the person is guilty and claims to have witnesses, physical stress (sleep deprivation, cold, discomfort) down-regulates the prefrontal cortex impairing accuracy monitoring, then the interrogator turns nice and walks the weakened person through a plausible scenario, feeding them information to imagine — over time coaxing certain people into a memory for something that never happened, as confirmed by exoneration cases.
“stress starts to down-regulate the prefrontal cortex. You're not necessarily as good at monitoring the accuracy of stuff. Then they start to get nice to you... they feed you the information.”
- Task switching — checking email, social media, or phone between work — destroys productivity and memory because each switch creates artificial event boundaries and costs executive function while you spend seconds catching up mentally, leaving fragmentary memories; instead of meaningfully integrating and predicting in a rich model, you are perpetually catching up, which is why multitasking (per Uncapher and Wagner's research) is bad for memory and increasingly common, and why blocking dedicated time for tasks is an investment in attention.
“what happens is I don't remember anything from that day. It just kind of vanished, because what happens, I'm creating all these artificial event boundaries. I'm losing all this executive function every time I switch, I'm getting a few seconds slower and I'm catching up mentally”
Curiosity and dopamine drive information-seeking as a discomfort energizing striving rather than reward consumption.
- In their research on curiosity, the more curious people were about the answer to a trivia question, the more activity appeared in dopamine-related brain circuits — and crucially this activity was driven by the question (the drive to seek information) rather than by the answer itself, reinforcing that dopamine reflects the seeking drive rather than the reward.
“the more curious people were about the answer, the more activity in these dopamine-related circuits in the brain, we would see. And again, that was not driven by the answer per se, but by the question.”
- Dopamine is not a pleasure chemical but is about the discomfort that energizes you to seek a reward; reinforcement learning shows you mainly learn when outcomes deviate from expectation, so humans are wired to strive and exist in a state of wanting rather than to be satisfied, which may explain why life objectively improving (medicine, food, safety) does not straightforwardly translate into greater happiness.
“people talk about dopamine, for instance, being this pleasure chemical. And there's a lot of compelling research to suggest it's not about pleasure at all. It's about the discomfort that energizes you to get things, to seek a reward.”
Internal models invisibly connect sensory samples into vast meaning—the central mystery of how minds construct coherence.
- The most beautiful aspect of the human mind is the internal model — everything we see, hear, touch and taste is connected by an invisible 'dark energy' that holds the whole universe of the mind together; we sample only a tiny bit of the world in front of us yet construct enormous meaning from it, and figuring out where that comes from and how things connect is the central mystery, much like dark matter and energy in physics being the unmeasurable majority.
“there's everything you see, and there's everything you hear, and touch, and taste... but it's all connected by this dark energy that's holding that whole universe of your mind together. And without that, it's just a bunch of stuff.”
Brain development optimizes for social group coordination across lifespan, not individual fitness alone.
- The short window of peak prefrontal/episodic memory function in young adulthood, flanked by weaker function in childhood and old age, is optimal for the ecology of social species; like orca pods led by grandmothers who pass on traditions, brain development across the lifespan is optimized not just for the individual but for kin and the social group, enabling collective wisdom across generations.
“another species that has menopause is orcas. Orca pods are led by the grandmothers... they're the ones that pass on the traditions to I guess the younger generation of orcas.”
Deja vu arises from partial match to tuned expectations producing familiarity without full recollection.
- Deja vu — an intense, sometimes disturbing sense of familiarity without recollection, also producible by electrically stimulating the temporal lobes and occurring before some epileptic seizures — arises as a byproduct of error-driven learning: higher-order brain areas tune themselves with repeated similar inputs, and a partial match to existing tunings produces a strong familiarity signal that isn't enough to trigger full recollection of what actually happened.
“It's basically a by-product of our mechanism of just learning, doing this error-driven learning as we go through life to become better and better and better to process things more and more efficiently.”
Reconsolidation allows memory modification during recall, but requires full reactivation and shows mixed success treating trauma.
- Reconsolidation is the idea that when you recall a memory, the connections between neurons in that cell assembly become more modifiable; techniques have tried to exploit this to reduce the visceral fear component of memories while activated, though results are mixed, and a key obstacle is that you need the person to fully recall the traumatic memory in the first place to modify it.
“a phenomenon called reconsolidation, which is the idea that essentially when I recall a memory, what happens is that the connections between the neurons and that cell assembly that give you the memory are going to be more modifiable.”
Technology danger lies not in systems but in people adapting maladaptively, trading discomfort and novelty for convenience.
- Many problems attributed to technology come not from the technology itself but from people adapting to it in maladaptive ways — text messaging makes communication Spartan and telegraphic, and email autocomplete constrains language toward 'good enough' suggestions — so the danger of AI is less what AI will do than what people will do, potentially becoming less creative by avoiding discomfort, novelty, and cognitive dissonance instead of diversifying their training data.
“a lot of the problems that come up with technology aren't the technology itself, as much as the fact that people adapt to the technology in maladaptive ways. I mean, one of my fears about AI is not what AI will do, but what people will do.”
Humans and rodents share conserved learning circuits yet differ fundamentally in sensory processing, language, and timescale integration.
- Despite differences, humans and mice share conserved neural circuits and genetic learning mechanisms.
“Despite all of your rage at GIFs, you're still just a rat in a cage.”
Science's success is acknowledging vast uncertainty and paradigm shifts through changing assumptions, not certainty.
- A great scientific success is being aware of the vast amount we don't know and being able to orient toward it; science is built on assumptions and, following Kuhn's Structure of Scientific Revolutions, paradigm shifts come from changing those assumptions — and the public, especially during COVID, failed to grasp how much uncertainty scientists genuinely live in.
“he frames this idea of assumptions being core to the scientific process, and the paradigm shift comes from changing those assumptions, and this idea of finding out this whole zone of what you don't know to me is the exciting part.”
AI systems make different mistakes than humans; true replication requires matching characteristic human biases and heuristics.
- AI machines make different kinds of mistakes than people do, so claims that machines do things 'just like humans' are misleading; we should not consider human intelligence replicated until a system makes exactly the same characteristic mistakes, biases, and heuristics that people make — and current systems like ChatGPT show no such evidence.
“the machines make different kinds of mistakes than the people do, and I will never be convinced unless we've replicated human... I don't think we've replicated human intelligence, unless I know that the simulator is making exactly the same kinds of mistakes that people do”
Self is multi-faceted; mental time travel accessing different past compartments can change present identity.
- The self is not deterministically fixed by a single past; we have a multi-faceted past and can access different compartments of who we were, and the act of mental time travel can change who we are in the present — Schacter's lab showed that merely remembering or imagining doing something altruistic made people more likely to help others, undermining the illusion of a stable, determined self.
“Dan Schacter's lab did this really cool study where they would ask people to either remember doing something altruistic or imagine doing something altruistic, and that act made them more likely to want to do things for other people.”
fMRI measures slow integrative processes indirectly via blood oxygen, suited to story comprehension but not fast neural events.
- fMRI infers brain activity indirectly via the magnetic properties of deoxygenated versus oxygenated hemoglobin as blood flows to active regions, measuring responses about six seconds after neural activity and being sensitive to changes in neuronal excitability (possibly mediated by astrocytes) rather than to spikes directly — which is a limitation for fast events but a feature for capturing the slow integrative processes involved in understanding stories and movies.
“in fMRI, you can measure these magnetic field responses about six seconds after that burst of activity would take place.”
Psychosis involves learning connections from prediction errors wrongly and confusing internal signals with external reality.
- Psychosis can be understood partly as learning connections from prediction errors that you shouldn't learn, and partly as internal signals being confused with actual things in the outside world — the same mechanism that, when channeled differently, may explain the relationship between genius and psychosis.
“that's what psychosis is in some ways is first of all, you're just learning connections from prediction errors that you probably shouldn't learn. The other part of it is is that your internal signals are being confused with actual things in the outside world.”
Unconventional structure creates prediction error enabling creativity once habitual expectations are abandoned.
- Writing music in unconventional time signatures (like seven-four instead of four-four) creates prediction error for the listener because we are habituated to expect resolution in four — the riff resolves on the first beat of the next measure, making the music flow faster than expected — and learning to embrace this dissonance, like in jazz and math rock, opens up creative possibilities once you stop following the rules people tell you music should obey.
“the music flows a little bit faster than it's supposed to, and you're getting a little bit of prediction error every time this is happening.”
Brain-computer interfaces advancing rapidly across neurotechnology beyond Neuralink using decoding and AI.
- Brain-computer interface innovation is much bigger than Neuralink, which has impressive surgical-robot technology but shows no public scientific evidence of being near where the broader scientific community is; numerous startups and labs are making enormous rapid progress on speech and movement prosthetics by combining decoding techniques with AI and technologies like Neuropixels that harvest activity from many neurons per electrode.
“I think BCI is much, much bigger than Neuralink and there's just so much innovation happening.”