/
Navigation
Chronicles
Browse all articles
Explore
Semantic exploration
Research
Entity momentum
Nexus
Correlations & relationships
Story Arc
Topic evolution
Drift Map
Semantic trajectory animation
Posts
Analysis & commentary
Pulse API
Tech news intelligence API
Browse
Entities
Companies, people, products, technologies
Domains
Browse by publication source
Handles
Browse by social media handle
Detection
Concept Search
Semantic similarity search
High Impact Stories
Top coverage by position
Sentiment Analysis
Positive/negative coverage
Anomaly Detection
Unusual coverage patterns
Analysis
Rivalry Report
Compare two entities head-to-head
Semantic Pivots
Narrative discontinuities
Crisis Response
Event recovery patterns
Connected
Search: /
Command: ⌘K
Embeddings: large
TEXXR

Chronicles

The story behind the story

days · browse · Enter similar · o open

How researchers, including at Meta's AI lab, use AI to study EEG readings, decoding how neurons in the brain communicate and exploring the nature of cognition

Wall Street Journal Christopher Mims

Context & Ripple Effects

This WSJ piece sits at the start of a measurable arc in neuro-AI: labs using large AI models to turn brain recordings into decoded signals. Weeks later, researchers showed a GPT model paired with fMRI readings could decode continuous language from human subjects non-invasively — a step beyond the EEG work described here.

By 2025, Meta itself had moved from passive reading to action decoding, sharing work on [[a:882267|a magnetic scanner plus deep neural network that identified which keys subjects pressed while typing]]. The thread connecting these is sensor cost: EEG is cheap and wearable, which is why Meta's interest in it matters more than any single paper.

First-order effects

  • Researchers gain a non-invasive, low-cost complement to scanner-based decoding: EEG plus AI models lets labs study neural communication without fMRI-grade equipment.
  • For Meta's AI lab, the work functions as flagship fundamental research that supports the aggressive researcher recruitment and retention effort sources have detailed, including Zuckerberg personally writing to DeepMind researchers to recruit them.

Second-order effects

  • Rival labs face pressure to publish comparable brain-decoding results or concede that frontier AI companies own the intersection of large models and neuroscience — a field previously dominated by academic groups.
  • Sensor makers and clinical-EEG vendors become strategic inputs: as decoding accuracy improves, demand shifts toward higher-resolution portable recordings rather than clinical-grade machinery.

Third-order effects

  • The trajectory from EEG cognition studies to fMRI language decoding to keystroke-level decoding suggests brain-signal interfaces are moving from lab instruments toward consumer-adjacent technology — raising privacy questions regulators have not yet addressed for decoded neural data.
  • If the pattern holds, frontier AI labs' research agendas increasingly set the direction of cognitive science itself, concentrating both the compute and the talent needed for the field inside a handful of companies.

The trend: AI models are converting progressively richer brain recordings into decoded language and intent, pulling neuroscience's frontier into the hands of big-tech AI labs.

Discussion

  • @tunguz Bojan Tunguz on x
    Very interesting article by @mims. I've long suspected that ML/AI could revolutionize psychometrics. It turns out that some people are already working on it. 'Do androids dream of electric sheep? We don't know yet, but we may soon be able to determine if humans are think... https…
  • @lynneafarr Lynne A. Farr PhD on x
    This is hugely exciting and somewhat frightening at the same time. Very interesting read. AI Is Teaching Us New Things About the Human Mind https://www.wsj.com/...
  • @hellohypercube @hellohypercube on x
    Neural nets can reprogram other neural nets if there is a communication link between them. That includes neural nets in servers reprogramming humans. Traditional computers using Ethernet etc can't do it. But any neural net using KVM link can. https://hellosemi.com/... https://twi…
  • @wsj @wsj on x
    Tools from cutting-edge AI systems are unlocking things about human beings which were difficult or impossible to figure out, simulate, or convincingly demonstrate before. https://www.wsj.com/...
  • @hkanji Hussein Kanji on x
    A little less than half of the time, Meta's AI algorithm was able to correctly guess what words a person had heard, based on the activity generated in their brains https://www.wsj.com/...
  • @kevinokeefe Kevin O'Keefe on x
    We all saw this. > AI research could one day lead to humans connecting with computers merely by thinking-as opposed to typing or voice commands. But there is a long way to go before such visions become reality. https://www.wsj.com/... - @WSJ's Christopher @Mims https://twitter.c…
  • @singularityumia @singularityumia on x
    Thought is ever-changing electrical patterns unconnected to individual neurons. Meta is working on a system to read your mind. 🧠 https://www.wsj.com/... via @WSJ @WSJTech @mims