Can AI read your mind?
Not in the way science fiction promises — but brain-computer interfaces are already turning brain activity into words for people who cannot speak. Here is what the technology actually does today, what Japan is building toward, and where the hype gets ahead of the evidence.
What if you could communicate without speaking?
You are sitting in a room, unable to move or say a word, but your brain is still working normally. You know exactly what you want to say. Now imagine a computer picking up signals from your brain and turning them into actual words.
That sounds unbelievable, but researchers are already working on versions of this technology.
Scientists around the world are developing brain-computer interfaces, or BCIs, that can pick up activity from the brain and use AI to understand what a person is trying to communicate. Japan is also making some seriously interesting progress in this field.
How a brain-computer interface works
Every time you speak, your brain is already doing a lot of work before a single word comes out. Different patterns of activity appear when you think about speaking, moving or producing certain sounds.
Researchers can record some of these signals using technologies such as EEG. More advanced systems can also use electrodes placed closer to the brain.
Then comes the really cool part.
AI can be trained using large amounts of brain signal data. It learns patterns and connects them with particular words, sounds or actions. Once it learns those patterns, it can try turning brain activity into text or artificial speech.
Turning brain signals into speech is already happening
And this is not just something researchers hope to achieve someday.
In 2024, scientists developed a brain-computer interface that helped a man with ALS communicate by converting his brain signals into words. With a vocabulary of 50 words, the system achieved more than 99 percent accuracy. Researchers later expanded it to work with a much larger vocabulary.
Another study demonstrated a system that could generate speech from brain activity in a person who had lost the ability to speak because of paralysis. The goal was to make communication faster and more natural.
Just think about that for a second.
Someone who physically cannot speak can have their intended speech turned into a voice
Japan’s Internet of Brains has an even bigger idea
Japan’s government-backed Moonshot Research and Development Programme has projects exploring something called the Internet of Brains.
The idea is fascinating. Can information inside our brains eventually be captured, understood and communicated through technology?
One Japanese research project is working on using AI to recognise unspoken speech from brain signals. Researchers are also exploring whether more complicated information, including abstract ideas, could eventually be communicated through brain-machine interfaces. The reported results are still experimental, with plenty of challenges left to solve.
If this technology develops further, the possibilities are huge.
Imagine someone who can think perfectly clearly but has lost the ability to speak or move. Instead of being completely dependent on physical movement to communicate, they could potentially use brain activity to interact with a computer.
That is honestly mind-blowing.
Can AI communicate with someone who is brain dead?
No. There is an important correction to some claims circulating online. AI cannot currently communicate with someone who is brain dead.
Brain death is not the same as being in a coma or being paralysed. It means there has been an irreversible loss of brain function, including the brainstem. There is no established evidence that a person properly diagnosed as brain dead can regain consciousness or communicate.
Some studies have detected electrical activity in brains after brain death was clinically determined. But electrical signals alone do not prove consciousness or meaningful communication.
So no, Japan has not found a way to communicate with the dead. But what is happening with living human brains is already fascinating enough.
Why this is a multi-disciplinary breakthrough
And there is something else worth noticing. Technology like this does not belong to just one field:
- Electronics captures the signals
- Embedded systems help process them
- Software and AI make sense of the data
- Neuroscience helps us understand what those signals mean
That is a good example of where modern technology is heading. The biggest breakthroughs often happen where different fields meet.
We spent decades teaching computers to understand human language. Now we are teaching them to understand the brain that creates it.
