←
Sean Escola124 posts
Sean Escola avatar✓

Sean Escola ✓

@SeanEscola

Chief Neuroscientist & Venture Partner. PL Neuro, PLC Neurotech, @protocollabs.

◷ Joined August 2024▧ 6 Photos and videos
Sean Escola avatar
Sean Escola@SeanEscola·Apr 1

PL Neuro is officially live! plneuro.xyz We exist to break bottlenecks, accelerate progress in neurotech & NeuroAI, and to invest in innovations that benefit humanity. This project of @protocollabs will focus on: 1. Neural augmentation 2. Biologically-inspired intelligence 3. Whole brain emulation

Post attachment
Max Hodak avatar
⟳ Sean Escola retweeted
Max Hodak@maxhodak_·Nov 11

Anyone who thinks EEG is significantly underrated should be asked to bring a live demo playable by others

Jacob Robinson avatar
⟳ Sean Escola retweeted
Jacob Robinson@JTRobinsonLab·Nov 9

Ok y’all, EEG can’t measure the same brain signals as implanted BCIs. It doesn’t matter how much AI you throw at it. Spatial averaging filters out high-frequency neural activity by the time it reaches the scalp. Anything higher than ~70 Hz for research-grade systems or ~20 Hz for consumer systems is below the noise floor. Implanted electrodes are much closer to the brain and can resolve brain activity up to ~200 Hz (epidural), ~600 Hz (ECoG), and ~2 kHz (penetrating). Saying you can infer high-frequency brain signals from EEG is like saying you can infer the lyrics to a song you’ve never heard by listening to the bassline. Sure, you can guess some lyrics, but they certainly won't be accurate. That doesn’t mean that EEG can’t be useful. But we shouldn’t pretend that EEG can do the same thing as an implant. It can’t because: Physics.

Post attachment
François Chollet avatar
⟳ Sean Escola retweeted
François Chollet@fchollet·Nov 7

Data scaling and compute scaling are *multipliers* for the generalization power of an approach. Trillions of dollars in investment can make very inefficient, low-generalization approaches look adequate, but that doesn't make them the ultimate end-all-be-all of AI. You should work on methods that *can* scale, because if you don't benefit from this multiplier you simply can't compete. But you should keep in mind scaling is only a multiplier: a cleverer approach will scale *better*.

Richard Sutton avatar
⟳ Sean Escola retweeted
Richard Sutton@RichardSSutton·Nov 7

Brute force computation is not enough for intelligence. Clever new methods are needed. The science of AI is far from finished. IMHO

Tom Roseberry avatar
⟳ Sean Escola retweeted
Tom Roseberry@TomRoseberry789·Nov 7

Neurotech is having a moment, and EEG is the consumer-friendly frontrunner. But what can its signals actually do? I built an interactive site that walks through the physics, from a single neuron to robot arms to telling “focus” from “calm”, to show what EEG can and can’t do.

0:09
Leo Dobler avatar
⟳ Sean Escola retweeted
Leo Dobler@leo_dobler·Oct 31

I somehow always get asked about less invasive BCI approaches. Here is which physics limits they hit and won't get you to high bandwidth I/O fMRI - obv great for research. Never wearable because of B0 field. Also slow, hemodynamic. Ultrasound - has some decoding results mainly NHP, some human with windows. Skull attenuation + coupling makes intact-skull wearable hard. Hemodynamic too. EEG - volume conduction + skull LPF. Averages over huge populations, kills HF content. Cool for demos, not high bandwidth. ECoG - you're already invasive. Still mostly population LFPs. For spikes just go a layer deeper... Photogenetics - genetic modification won't get approved any time soon, still needs light delivery

Saturnin Pugnet avatar
⟳ Sean Escola retweeted
Saturnin Pugnet@satpugnet·Oct 27

Time is becoming a poor measure of how distant the future really is. Ten years used to be a manageable stretch of the future, but with AI, it does not tell you much anymore. After much thinking, I came up with a better framing to measure the future: tokens of thinking. Let me explain:

Precision Neuroscience avatar
⟳ Sean Escola retweeted
Precision Neuroscience@PrecisionNeuro_·Oct 29

A major milestone for @PrecisionNeuro_ ! We’re expanding our global clinical research network through a two-year partnership with Cleveland Clinic Abu Dhabi (@CCAD), with support from the Department of Health – Abu Dhabi (@DoHSocial) to evaluate Layer 7. Announced at the Abu Dhabi Future Health Summit, this partnership advances our mission to push the boundaries of neurotech for people worldwide. Read the full news release below. #BrainComputerInterface #DoH #AbuDhabi #BCI #NeuroTech globenewswire.com/news-relea…

Sean Escola avatar
Sean Escola@SeanEscola·Oct 27

Very excited for the future of neurotech with @plcapital and glad I get to play a role!

0:47
Matthew Botvinick avatar
⟳ Sean Escola retweeted
Matthew Botvinick@mattbotvinick·Oct 22

I'm at a meeting on democracy and AI -- the world's most important topic -- and there's something bugging me. It seems to me that the hyperspecialization of modern life, the fact that we all become hyperspecialists in some area over time, means that public discussions of AI and democracy end up being debates among people who are anchored on only one piece of the picture. Some focus on national security, others on child safety, others on economic impacts. No one has a vantage point that spans the entire complex picture. This is not a good setup, because policymaking is always about tradeoffs. If no one person can see all the dimensions involved in the tradeoff, all that results is stalemates. Perhaps when we get to an end of work, we can all look up from our hyperspecializations, and think comprehensively about the public good.

Precision Neuroscience avatar
⟳ Sean Escola retweeted
Precision Neuroscience@PrecisionNeuro_·Oct 14

Introducing the Precision Brain Atlas: high-resolution neural recordings from 70 patients across 12 academic medical centers. A growing resource for studying human neural function- and a step toward a Whole-Brain Interface.

Saturnin Pugnet avatar
⟳ Sean Escola retweeted
Saturnin Pugnet@satpugnet·Oct 14

For AI labs, BCI will play a role at every stage of a foundation model’s life: data collection, pre-training, post-training, and inference.

Post attachment
Saturnin Pugnet avatar
⟳ Sean Escola retweeted
Saturnin Pugnet@satpugnet·Oct 14

Phase Lock went a lot more viral than expected. Thank you for all the thoughtful messages. I am still working through them. My brother’s review was particularly insightful: “Yes, no chance I am reading all that bro, but I am sure it's great” So for him, and everyone else here who prefers something more succinct, here are the key takeaways through a few of my favorite visuals:

Post attachmentPost attachment