Merge Labs explores ultrasound and gene therapy to create a safe, noninvasive brain-computer interface that connects human thoughts with AI.
Using minimally invasive techniques, rather than methods like “implanting electrodes in the brain,” is crucial. Two major technological paths have emerged in particular: using ultrasound (sound waves) to communicate with brain cells (neurons). Gene therapy is the use of gene therapy to engineer neurons to respond to ultrasound or other signals (such as magnetic fields). This mission is ambitious: “I think something, and ChatGPT (or any computer) will immediately understand and respond Altman stated it very clearly.
Technological Background Why and How?
Brain-computer interfaces (BCIs) are necessary, carrying the risk of infection or neuron damage. Reading brain signals with external devices is difficult; the signals are weak, noisy, and have low accuracy. The complex network of neurons is heterogeneous in each individual. The signals associated with “thinking” are highly diverse. This is why BCIs to date have primarily been used for therapeutic purposes, not for “thought-to-control” purposes.
Merge Labs’ Technical Vision
The direction Merge Labs is moving in is Using ultrasound waves to transmit and receive signals to brain tissue without open-skull surgery. Magnetic Fields: Some sources suggest that Merge Labs is also looking into creating a larger “interface” using sound waves and magnetic fields. Being less invasive: Altman has clarified that he “wouldn’t want to implant something in his brain that would kill his neurons.”
How the technology could work a possible approach
Based on publicly available information, a possible sequence could be as follows: A person is fitted with a device (or wearable on the head/near the brain) that may contain an ultrasound transducer and/or a magnetic field generator. Neurons engineered through gene modification will respond to ultrasound waves or magnetic signal stimulation – that is, they will activate, changing the signal history. The device will read and analyze this responsive signal, converting it into digital commands – such as “think, move the mouse right.” As a result, computers, mobile phones, AI systems, etc. will understand and respond to that “thought command.” Moving forward, at the manufacturing or consumer level, this could be extended to “normal users think, and the machine responds instantly.
Why is this news so big?
Competition and Trends Merge Labs is directly competing with companies like Neuralink, which have worked in the field of implantable chips. Merge Labs’ “less invasive” approach makes it an exciting option for the future. The BCI market is growing rapidly, with investors showing significant interest. Merge Labs has already received initial funding. Additionally, the concept of a “merge” between AI and human-brain interfaces is gaining increasing attention. Altman previously wrote that “the merging of man and machine” is a future direction.
Potential Impact
If successful, this technology could give people the ability to control devices simply by thinking—reducing traditional interfaces like keyboards, mice, and touchscreens. This could lead to major advances in the medical field: new disability-relieving options could open up for paralyzed individuals and those with neurological disorders. The level of human-machine interface could change; communication through thoughts could become possible.
Concerns and Risks
Privacy: If your thoughts could directly control devices, the question arises—will “my thoughts remain my own?” Security: The risk of brain signals being hacked or misused; or the problem of psychological pressure/management.
Ethical and Legal: Who will decide whether this technology is safe for whom and who will it be applicable to? What will be the policy for human testing? Practical feasibility: This technology is still in its infancy; it will take time to progress from a “vision” to a “commercial product.” Widespread Use and Cost: Gene modification + ultrasound + magnetic interface = a very complex and expensive solution.
Current Status What is known from available information
Merge Labs has not yet publicly announced a major product or human trials; it is in the early stages. The company has added biomolecular engineer Mikhail Shapiro to the team—his previous work involves controlling neurons with ultrasound. According to sources, Merge Labs plans to raise around $250 million in funding next year, and the initial valuation could be around $850 million. Altman has clarified that he wants a “read-only” interface, meaning that feedback can be generated by mere thought, but not too invasive.
The Way Forward What can we wait for?
Short-term (1-2 years) Merge LabsOr other companies may initiate pre-clinical or preliminary human trials, for example, capabilities such as “controlling smartphones by thought” or “mind commands beyond voice commands.” A new chapter of human-machine merging may begin (5+ years). It is possible that the gap between human cognition and machine intelligence (AI) may narrow further. This could even lead to “human augmentation.” The concepts of “human understanding machines” and “machines understanding humans” may become commonplace. Ethics, laws, and socio-cultural structures will be transformed to accommodate this technology. For example, who will own “thought rights” and “brain data
About the Author
