The tech company described the development as a step forward in learning how our brains process and sculpt sensory signals.
“Using magnetoencephalography (MEG) we showcase an AI system capable of decoding the unfolding of visual representations in the brain with an unprecedented temporal resolution,” Meta said in a blog post on its website.
Many of the images, according to Meta, can be replicated with millisecond precision through what it describes as “a non-invasive neuroimaging technique in which thousands of brain activity measurements are taken per second.”
The images are not perfect replications, but incredibly close compared to what the subjects were shown nonetheless.
“While the generated images remain imperfect, the results suggest that the reconstructed image preserves a rich set of high-level features, such as object categories,” Meta said.
It added that the development could be a “stepping stone” towards helping people who suffer from neurological issues such as brain lesions, where certain individuals might have lost the ability to verbally communicate.
The company also discussed other potential uses for the technology, noting that it could “ultimately guide the development of AI systems designed to learn and reason like humans.”
In a related research study published on Meta website, the authors acknowledge the potential privacy concerns revolving around the development as well.
“While the decoding of brain activity promises to help a variety of brain-lesioned patients, the rapid advances of this technology raise several ethical considerations,” they wrote in the study titled Brain Decoding: Toward Real Time Reconstruction of Visual Perception.
“Most notably the necessity to preserve mental privacy.”
“The subjects’ consent is not only legal but also and primarily a technical requirement for brain decoding,” they added.
The authors of the paper went on to endorse open and peer-reviewed research standards.
Meta has increasingly poured resources into generative AI amid a Big Tech pivot to systems that can match or exceed human ability to generate text, video, images and other content.
Those resources and research have come to fruition in the form of an AI platform that can translate almost 100 languages, new AI models to make coding easier, and even AI stickers, to name a few initiatives and products.