Reading 4 min Published Updated
Researchers at the University of Texas at Austin have developed an artificial intelligence (AI) system that can interpret and reconstruct human thoughts.
Scientists recently published a paper in Nature Neuroscience investigating the use of AI to non-invasively translate human thoughts into words in real time.
According to the researchers, current thought-to-word decoding methods are either invasive, requiring surgical implantation, or limited in that they “can only identify stimuli among a small set of words or phrases.”
The Austin team circumvented these limitations by training a neural network to decode fMRI signals simultaneously from multiple regions of the human brain.
In this experiment, the researchers asked several subjects to listen to podcasts for hours while an fMRI machine non-invasively recorded their brain activity. The resulting data was then used to train the system in the thinking patterns of a particular user.
After training, the subjects were again monitored for brain activity as they listened to podcasts, watched short films, and silently imagined telling the story. During this part of the experiment, the AI system received the subjects’ fMRI data and decoded the signals in plain language in real time.
According to a press release from the University of Austin in Texas, the AI could do everything right about 50% of the time. The results, however, are not precise – researchers have designed AI to convey the general ideas that are thought about, rather than the exact words that are thought about.
Fortunately for those who worry about having AI infiltrate their thoughts against their will, scientists are quite clear that this is currently not possible.
The system only works if it is trained in the brainwaves of a particular user. This makes it useless for scanning people who have not spent hours providing fMRI data. And even if such data was generated without the user’s permission, the team ultimately concludes that both the decryption of the data and the machine’s ability to track thoughts in real time require active participation on the part of the person being scanned.
However, the researchers noted that this may not always be the case:
“[Наш] privacy analysis shows that subject cooperation is currently required for both learning and using the decoder. However, future developments may allow decoders to circumvent these requirements. Moreover, even if the decoder’s predictions are inaccurate without the participation of the subject, they can be deliberately misinterpreted for malicious purposes.
In related news, a Saudi Arabian research team recently developed a method to improve the accuracy of diagnosing brain tumors by processing MRI scans through a blockchain-based neural network.
In their article, Saudi researchers demonstrate how processing cancer research on a secure decentralized blockchain can improve accuracy and reduce human error.
Related: what is immutable, explanation
While both of the above experiments are cited as early work in their respective research papers, it is worth noting that the technology used in each is widely available.
The AI at the heart of the experiments conducted by the team at the University of Austin in Texas is a generative pre-trained transformer (GPT), the same technology that ChatGPT, Bard and similar large language models are built on.
And the cancer research by a team in Saudi Arabia was conducted using AI trained on Nvidia GTX 1080 GPUs available since 2016.
There really is nothing stopping a smart developer (with access to an fMRI machine) from combining the two ideas to develop an AI system that can read your thoughts and write them to the blockchain.
This could lead to a “proof of thought” paradigm, in which perhaps we could mint non-fungible tokens (NFTs) of our thoughts, or write down the immutable books of our feelings and ideas for posterity, legal purposes, or just to show off.
The influence of, for example, minting NFTs from thought to chain can have implications for copywriting and patent applications, where the block chain serves as proof of exactly when a thought or idea was written down. It could also allow famous thinkers such as Nobel Prize winners or contemporary philosophers to organize their ideas into an immutable record that could be commodified and used as collectible digital assets.