Researchers at the University of Texas at Austin have developed an artificial intelligence (AI) system capable of interpreting and reconstructing human thoughts.
Scientists recently published a paper in the journal Nature Neuroscience that explores the use of artificial intelligence to non-invasively translate human thoughts into words in real time.
According to the researchers, current methods of decoding thoughts into words are either invasive — meaning they require surgical implantation — or limited in that they “can only identify stimuli from among a small set of words or phrases.”
Austin’s team circumvented these limitations by training a neural network to decode functional magnetic resonance imaging (fMRI) signals from multiple regions of the human brain simultaneously.
While conducting this experiment, the researchers had several test subjects listen to hours of audio broadcasts while an fMRI machine non-invasively recorded their brain activity. The resulting data was then used to train the system on a specific user’s thinking patterns.
After training, the test subjects’ brain activity was monitored again while listening to podcasts, watching short films, and silently imagining they were telling a story. During this part of the experiment, the AI system was fed fMRI data and decoded the signals into simple language in real time.
According to a press release from the University of Texas at Austin, the AI was able to get things right about 50% of the time. However, the results aren’t exact – the researchers designed the AI to convey general ideas being thought of, not exact words being thought of.
Fortunately for anyone concerned about AI hacking into their thoughts against their will, scientists are very clear that this is not currently a possibility.
The system only works if it is trained on the brainwaves of a specific user. This makes it useless for scanning individuals who have not spent hours submitting fMRI data. Even if this data was generated without the user’s permission, the team ultimately concludes that decoding the data and the device’s ability to monitor thoughts in real time require active participation on the part of the person being scanned.
However, the researchers note that this may not always be the case:
(o) Your privacy analysis indicates that subject collaboration is currently required to train and use the decoder. However, future developments may enable decoders to bypass these requirements. Furthermore, even if the decoder’s predictions are inaccurate without collaboration subject matter, they may be deliberately misinterpreted for malicious purposes.”
In related news, a team of researchers in Saudi Arabia recently developed a method to improve accuracy in diagnosing brain tumors by processing MRI scans through a blockchain-based neural network.
In their paper, the Saudi researchers explain how processing cancer research on a secure and decentralized blockchain can improve accuracy and reduce human error.
Related: Explain what is fixed
While the two experiments mentioned above are cited as early work in their respective research papers, it should be noted that the technology used in each is widely available.
The AI underlining the experiments conducted by the team at the University of Texas at Austin is a generative pre-trained transformer (GPT), the same technology that ChatGPT, Bard, and similar large language paradigms are built on.
The Saudi Arabia team conducted the cancer research using AI trained on Nvidia GTX 1080s, GPUs that have been available since 2016.
Realistically, there is nothing stopping an intelligent developer (with access to an fMRI machine) from combining the two ideas in order to develop an AI system that can read a person’s thoughts and record them on the blockchain.
This could lead to a “proof of thought” paradigm, where people can make immutable tokens (NFTs) of their thoughts or immutable ledgers of their feelings and thoughts for posterity, for legal purposes, or just for bragging rights.
For example, the impact of NFT coinage from ideation to blockchain could have implications for copywriting and patent applications as the blockchain serves as proof of when exactly an idea or idea was registered. It could also allow famous thinkers such as Nobel laureates or contemporary philosophers to write down their ideas in an immutable record – a record that can be commodified and used as a collectible digital asset.