In an excerpt from his book The Vanishing Earth, journalist James Crawford describes a Columbia University experiment that reads and manipulates the brains of mice with laser precision. The research, led by neuroscientist Rafael Yuste, demonstrates that neural activity can be decoded and artificially replayed — a capability Yuste says will extend to humans.
The Experiment: Reading a Mouse's Visual Cortex
According to the excerpt, Yuste works at Columbia University in a lab that studies how the cortex responds to vision. His mentor was the Swedish Nobel Prize-winning neuroscientist Torsten Wiesel, who discovered that high-contrast dark and light bars are the strongest stimuli for the visual cortex. Yuste's team trained mice to associate moving bars on a screen with drinking: bars moving up and down signaled a drink from a water tube, while side-to-side bars meant stop drinking.
Using a sophisticated laser system, the researchers monitored brain activity through the mouse's skull, identifying exactly which neurons fired when the animal viewed the images. "We can see the neurons that are encoding the visual stimulus," Yuste explains.
A second holographic laser system projected points inside the brain, activating the same neurons that represented vertical or horizontal moving bars. With the screen off, the mice still behaved as if the images were present.
"The killer experiment was to turn off the screen," Yuste says. "Just like when you are playing the piano, you use different fingers on particular keys. So, we are playing the images on the cortex. And when we play them, we make the mouse behave in the way we want it to."
The effect was indistinguishable from real vision. As Yuste describes it, the mouse's licking behavior — number of licks, duration, and delay — was identical whether the image was on the screen or implanted. "He cannot tell the difference. He thinks that these things are real in front of him."
Yuste summarizes the achievement: they could "manipulate the mouse like a puppet" by choosing which image to put into its brain.
From Mouse to Human: fMRI and Machine Learning
The excerpt quotes Yuste directly: "And what we can do in a mouse today we can do in a human tomorrow." Over the past two decades, researchers using functional magnetic resonance imaging (fMRI) — which tracks the iron in hemoglobin supplying oxygen to neurons — have built increasingly detailed maps of the mammalian cortex.
The article explains that machine-learning AI, algorithms that sort enormous amounts of information and use statistical methods to make classifications and predictions, has pushed this research forward. According to the excerpt, fMRI scans can now identify "everything from depressive thoughts to the nuanced feelings of envy and schadenfreude." Other algorithms have:
- Accurately pieced together reconstructions of movie clips watched by subjects, based only on brain scans.
- Detected, in probing the brain activity of swing voters in a U.S. presidential election, which presidential candidates provoked anxiety or disgust, and which elicited positive responses or feelings of empathy.
| Technique | What it does | Evidence from the excerpt |
|---|---|---|
| Laser monitoring system | Tracks neuron firing through the skull | Identified neurons encoding vertical/horizontal bars |
| Holographic laser stimulation | Activates specific neurons inside the brain | Made mice behave as if seeing projected bars with the screen off |
| fMRI scanning | Tracks iron in hemoglobin supplying oxygen to neurons | Builds cortex maps; identifies emotions like envy and schadenfreude |
| Machine-learning AI | Sorts large datasets, makes classifications and predictions | Reconstructs movie clips; reads swing voters' reactions to candidates |
The Big Tech Stakes
The article's headline points to a broader concern: "Big Tech Wants to Harvest Your Thoughts." While the excerpt itself focuses on scientific results, the research trajectory described raises questions about neural data collection. The same machine-learning techniques that can identify depressive thoughts and political anxieties are being applied to ever more detailed brain scans. Yuste's assertion that what works in a mouse will work in a human suggests that decoding and potentially influencing human perception is not a distant prospect but a laboratory reality.
The excerpt does not name any specific technology company or commercial product. Instead, it presents the underlying science as a foundation that could be built upon. For enterprise technology leaders, the direction is evident: as neural data becomes machine-readable, the same questions that have surrounded data collection, privacy, and ethics will apply to the most personal data of all.
Sources:
