In recent years, the technology sector has witnessed remarkable progress in artificial intelligence and brain-computer interfaces (BCIs). In January 2024, Neuralink achieved a milestone by implanting a chip into the brain of an American man with quadriplegia, enabling him to control a computer mouse cursor solely by thought.
In May 2023, American researchers introduced a revolutionary technology capable of deciphering words directly from an individual's thoughts through brain scans utilizing generative artificial intelligence. Meanwhile, at the University of Technology Sydney (UTS), scientists developed a cap named "BrainGPT" that interprets and converts thoughts into readable text using AI, generating significant media attention with headlines such as "AI Cap Reads Minds."
These rapid advancements in artificial intelligence and neural technology prompt numerous inquiries: Can neural implants and AI genuinely decipher thoughts? Is there a foreseeable future where computers can accurately transcribe our thoughts instantaneously?
Can AI read people's thoughts?
As with any new technology, this technology has its positive and negative aspects. It may offer potential benefits for many medical conditions, but at the same time, it may threaten the privacy of our thoughts and erode their isolation. However, before we panic, we should consider whether the capabilities of neural implants and generative artificial intelligence truly constitute a reading of the mind in the true sense of the term.
Between reality and imagination, is it possible to actually read minds?
Our conscious experience derives from brain activity, driving the advancements in modern science. As we contemplate a topic like the history of the Roman Empire or imagine a cursor moving on a screen, specific neuronal activity accompanies these mental processes.
It's logical then, that with this theoretical understanding, brain-tracking devices could interpret and analyze these activities, offering insights into our thoughts and cognition by reading neural patterns.
However, the greatest challenge remains in precisely correlating our mental states with corresponding neural activities. The human mind is exceedingly complex, encompassing a wide array of experiences and ideas, making it difficult to pinpoint a specific neural pattern for each thought or concept.
Thus, scientists still face a significant challenge in understanding how the mind operates and communicates through neural activity. This necessitates meticulous study to determine whether specific brain activity aligns with specific mental experiences.
This includes, for example, the ability to distinguish between neural activity patterns corresponding to seeing a red rose, smelling it, touching it, or even imagining it, as well as discerning differences in brain response when thinking about it as a favorite of a particular individual.
A journey of the mind and brain
Initially, it's important to differentiate between various types of thoughts, whether they stem from sight, smell, touch, imagination, or contemplation of different objects, such as a lemon. However, for anything beyond these patterns, one can perceive it, imagine it, or contemplate it.
Comprehending the brain processes linked with mind reading poses an exceedingly intricate challenge, perhaps beyond verbal description. For instance, the complexity of facial recognition entails triggering intricate neural networks in the brain, yet many of these processes don't directly relate to the face, but rather to other functions like memory, attention, self-monitoring, and task planning.
The significant challenge lies in accurately extracting the neural processes involved in consciously perceiving faces, and thus far, neuroscience hasn't unraveled this intricate puzzle. Even if progress is made, we'll likely only discern general correlations to the conscious perception of faces without the ability to differentiate specific faces.
Consequently, despite potential advancements in neuroscience, mind reading through brain scans will likely remain a distant objective because devices won't discern whether you're viewing Barack Obama's face, someone else nearby, or even an unfamiliar face.
Hence, it's evident that mind reading via brain activity poses an immensely arduous task, fraught with scientific and technical hurdles, and isn't currently achievable with precision.
Artificial intelligence in human brains
When we talk about the mind and its advanced technologies, the focus shifts to neural implants and artificial intelligence, raising questions about the possibility of mind-reading. Starting with neural implants, this technology aims to assist patients in performing specific tasks, such as moving a cursor on a screen, without the need for a precise understanding of the neural processes associated with these tasks.
Despite the successes of neural implants, they do not provide accurate information about brain activity related to internal thoughts, making them far from achieving true mind-reading. Regarding the use of artificial intelligence for mind-reading, a recent study has shown that it is possible to use image analysis from MRI scans to read thoughts, although this technology is still in its infancy.
During this study, participants were asked to listen to audiobooks and watch silent films while inside an MRI scanner. Researchers developed a system that uses artificial intelligence to convert brain activity into written text, aiding in understanding the thoughts they were listening to or imagining.
While this system was not 100% accurate, it was able to predict mental content better than mere chance. Similar to neural implants, this system can be useful in identifying thoughts related to specific activities, but it is not effective in tracking all types of internal thoughts.
Researchers turned to using an enhanced version of ChatGPT to convert brain activity into written text, opening up new horizons for reading participants' thoughts and transforming them into written words. Although the accuracy was not perfect, the system's ability to predict mental content was remarkable. Researchers imagine the system's potential to anticipate continuous narrative from brain analyses accurately, akin to how neural implants adopt a specific enhancement for a particular task, the enhanced system will be only for this purpose, unable to track any other mental activity.
To what extent can AI read minds?
Advances in mind reading using artificial intelligence techniques are escalating, raising questions about its ability to absorb mental activity. But to better understand this, we must recognize the complexities of the mental process that include the flow of thoughts and mental experiences.
The challenges are that mental activity is characterized by multiple paths and rapid changes, involving momentary thoughts, memories, expectations and fantasies, which makes it difficult to analyze accurately even by the smartest modern technologies.
Despite advances in technology and advances in artificial intelligence, it remains difficult to imagine how a state-of-the-art brain scanner can capture and understand all these mental details with clear precision.
The future of mind reading with AI
Artificial intelligence has been witnessing tremendous progress in recent years, demonstrating that some obstacles can be overcome in the future. However, given the complexity of our mental life and our limited understanding of the brain, this must be approached with extreme caution, as neuroscience is still in its infancy.
Currently, AI technologies cannot reach an accurate level in understanding all aspects of mental activity, making reading thoughts fully a difficult challenge.
