top of page

Can AI Access Your Private Thoughts?

Recent research conducted by the University of Texas at Austin has unveiled a groundbreaking Brain-Computer Interface (BCI) decoder that can translate brain activity into words. Leveraging a transformer model similar to OpenAI's ChatGPT, this technology generates intelligible word sequences, capturing the meaning of spoken language, imagined speech, and even videos without sound.

This achievement marks the first time that a person's thoughts can be read non-invasively, without the need for surgical implants. The breakthrough has the potential to revolutionize the field of speech restoration for individuals who struggle to communicate due to conditions like stroke or motor neurone disease.

Leveraging a transformer model similar to OpenAI's ChatGPT, this brain computer interface (BCI) generates intelligible word sequences, capturing the meaning of spoken language, imagined speech, and even videos without sound. AI, thoughts to text

Image generated with Midjourney. Prompt used: "Artificial Intelligence accessing your private thoughts"

Decoding Thoughts with Non-Invasive Precision

Led by Dr. Alexander Huth at the University of Texas at Austin, the team utilized functional magnetic resonance imaging (fMRI) scans to reconstruct speech with remarkable accuracy. Unlike previous language decoding systems that relied on invasive methods, this new approach leverages the power of large language models to represent the semantic meaning of speech in numbers. By correlating patterns of neuronal activity with specific meanings, the researchers were able to decode brain activity into text.

To train the decoder, three volunteers underwent intensive scanning sessions while listening to podcasts for 16 hours each. The decoder was then trained to match brain activity to meaning using a precursor to ChatGPT known as GPT-1. Subsequent scans of the participants listening to new stories or imagining telling a story demonstrated that the decoder could generate text solely from brain activity, closely resembling the intended meanings of the original words about half the time.

Understanding the Gist of Thoughts

The decoder operates at the level of ideas and semantics, providing the gist of the thoughts rather than an exact word-for-word translation. While the system still faces challenges in areas such as pronouns, the progress made is a significant leap forward compared to previous non-invasive methods, which typically focused on single words or short sentences. The outcome was a decoder that can generate word sequences based on brain activity, effectively translating the thoughts into comprehensible language.

Ethical Considerations and Responsible Use

While the potential of this technology is promising, concerns about its ethical implications have not been ignored. The researchers emphasize the importance of responsible use and ensuring that individuals have control over the use of such technologies. Privacy and consent are critical considerations, and the team is committed to addressing these concerns to prevent misuse.

Experts outside the project have praised the technical achievement of the research, highlighting its potential for exploring new experimental avenues. These include reading thoughts during dreams or investigating how new ideas arise from background brain activity. The ability to delve into the brain at a deeper level opens up exciting possibilities for understanding the human mind.

Exploring New Frontiers

The researchers are now looking to assess whether this technique can be applied to other more portable brain-imaging systems, such as functional near-infrared spectroscopy (fNIRS). The development of non-invasive methods like this decoder holds great promise for assisting individuals with speech impairments and enhancing our understanding of the human brain.

The potential applications of AI in decoding thoughts and speech are vast, and the breakthrough discussed in this article is just the beginning. As technology continues to advance, we must strike a balance between progress and ethics, ensuring responsible use and safeguarding individuals' privacy and consent. The future holds exciting possibilities as we continue to explore the depths of the human mind and harness the power of AI for the betterment of humanity.

The University of Texas at Austin's groundbreaking research in non-invasive mind-reading opens up new horizons for speech restoration and understanding the human mind. The ability to translate brain activity into words with remarkable accuracy brings hope to individuals who struggle to communicate due to various conditions. While ethical considerations remain essential, the progress made in this field is undeniable. As we move forward, responsible use and advancements in technology will pave the way for a future where AI enables us to decode thoughts and enhance human communication and understanding.


At Lucea AI, we are passionate about the potential of AI and its impact on society. As an AI training company, we strive to push the boundaries of AI. We believe in responsible and ethical use of AI to create solutions that benefit humanity.

If you are inspired by the possibilities of AI and want to learn more about our AI Training program, contact us today to learn how you can harness the power of AI and shape your future.



Tang, J., LeBel, A., Jain, S., & Huth, A. G. (2023). Semantic reconstruction of continuous language from non-invasive brain recordings. Nature Neuroscience, 1-9.


Recent Posts

See All

Bình luận

bottom of page