A group of academics and scientists has unveiled a significant breakthrough in seamless telepathic communication. For the first time, artificial intelligence has successfully decoded continuous imagined speech directly from the brain, marking a new era in human-AI interaction through an innovative model known as MindSpeech.
In a pioneering trial conducted by the startup MindPortal in June, researchers demonstrated that AI could interpret continuous imagined speech from the brain without the need for surgery. This development paves the way for future advancements in human-technology communication.
Read our latest News : Announcing the Full Agenda for the 2024 ISC2 Security Congress
Unlike earlier attempts that relied on simple or memorized verbal cues or required invasive surgery, MindSpeech has been developed to process more complex, free-form thoughts under specific test scenarios, all non-invasively. This marks a significant leap forward in the technology’s capacity to understand and translate human thoughts dynamically and intuitively.
In a groundbreaking study published on arXiv and ResearchGate, currently under peer review, MindPortal’s researchers employed a novel ‘word cloud’ task. Participants were prompted to imagine sentences based on topics covering over 90 percent of the most frequently used words in the English language. The data was collected using a portable high-density Functional near Infrared Spectroscopy (fNIRS) headset, capturing neural activity across the entire head.
Participants generated between 433 to 827 sentences, with an average length of 9.34 words per sentence, providing a robust dataset for training the AI model. The innovative approach used the Llama2 Large Language Model (LLM) for text generation, guided by brain signal-generated embeddings. These embeddings, created by combining context input text with brain signals, enabled the AI to generate coherent text from imagined speech.
The results were statistically significant for three out of four participants, showcasing the model’s ability to decode continuous imagined speech from thought.
“Our findings are a promising indication that we’re moving closer to achieving more natural and intuitive ways for machines to understand human thoughts,” said Dr. Suyi Zhang, PhD, AI Research Engineer at MindPortal and lead AI researcher for the study. “While the technology is not yet ready for widespread use and requires further refinement, our study’s results point to a future where this kind of telepathic communication could become a reality.”
Though more research is needed in the field of human-AI communication, MindSpeech has, for the first time, demonstrated that AI can interpret continuous, imagined speech directly from the brain without invasive surgery, hinting at how we may communicate with AI in the future.
“MindSpeech is the first research to show that AI can decipher our free-flowing imagined thoughts from the brain. Our team at MindPortal has proven that AI can interpret thoughts into text without requiring surgery or relying on a limited set of sentences or verbal cues,” said Ekram Alam, Co-Founder & CEO of MindPortal. “This study offers a glimpse into how humans will communicate with AI in the very near future.”
Signatory Statement:
“The launch of MindSpeech, a result of MindPortal’s groundbreaking AI research, marks the translation of free-form thought into text. This is a major step toward a future where seamless communication through brain-computer interfaces, utilizing non-invasive physiological measures and AI algorithms, becomes a reality.
Previous efforts were limited to decoding memorized words or phrases, but the new approach implemented through MindSpeech represents a significant advancement in human-AI interaction, capturing the fluid nature of internal speech. MindPortal’s AI research provides, for the first time, preliminary evidence that AI can decode continuous imagined speech from the brain.” — Professor Maysam Chamanzar, Former Neuralink/DARPA Scientist; Professor Gabriele Gratton, Former DARPA Scientist, University of Illinois; Professor Monica Fabiani, University of Illinois; and Professor Hamid Dehghani, University of Birmingham
To share your insights with CyberTech Newsroom, please write to us at news@intentamplify.com
