Meta AI researchers are developing a language model that mimics the way the human brain processes language

This Article Is Based On The Meta Research Article 'Studying the brain to build AI that processes language as people do'. All Credit For This Research Goes To The Researchers Of This Research 👏👏👏

Please Don't Forget To Join Our ML Subreddit

Meta AI is embarking on a long-term research initiative to better understand how the human brain interprets language. The research will be conducted in collaboration with NeuroSpin, a neuroimaging institute, and Inria, the French National Research Institute. Meta AI will compare how AI language models and human brains respond to identical spoken or written words.

Artificial neural networks of language are increasingly approaching human brain function, offering new insights into how thought can be implemented in neural tissue. The AI ​​models that most closely replicate human language now do so by methodically breaking down sentences by examining context and trying to anticipate the next word using machine learning.

While these technologies can give consumers a false sense of “humanness,” the models predict the next word based on large databases of the progress of previous conversations. On the other hand, the human brain anticipates words and thoughts in advance, considering whatever the utterance or concept may imply.


Giving an AI model the phrase “Once upon a time” and having it guess the next word is an example of a unique procedure for predicting the word “time”. When someone who was raised in fairy tales hears “Once upon a time,” their brain does more than just predict “time” as the next word. It also evokes all the magical designs that come with it, such as terrible witches, dragons, castles, heroes, and other culturally significant figures.


Brains create particular “brain states” when making these predictions, which can be visualized during brain imaging. Snapshots of brain activity were taken while volunteers read or listened to a story using functional magnetic resonance imaging and magnetoencephalography scanners. The researchers noticed something unusual when using machine learning on brain scans from public datasets paired with new fMRI and MEG images. According to the results, language processing in the human brain resembles ordered hierarchies, comparable to the functioning of AI language models.

There are parts of the brain comparable to visual processing algorithms that fire when words generate visual stimuli, areas similar to word-understanding algorithms, and entire networks that function similarly to AI language processors.

Specific brain areas are involved in vision and language processing, and their interactions create networks to generate narratives and representations for understanding. The results revealed that particular brain areas, such as the prefrontal (front of the brain) and parietal (middle of the brain) cortices, better represented language patterns with distant word predictions.

This suggests that the internal representations shared by brains and algorithms are useful for the algorithm to process language. This result was quickly validated after studying the brain activations of 200 participants during a simple reading test. Then, about a week later, a team from the Massachusetts Institute of Technology conducted an independent study investigation with remarkably identical results.

This research provides new insights into brain processes by using this research to draw quantitative parallels between the human brain and AI models. AI that can behave and react more in sync with the use of human language will naturally connect with humans. To find out more, see here.



James G. Williams