Tech

Neuralink: The whole upload topic is Science Fiction

That sounds too tempting, what the US company Neuralink announced last week: It has developed a novel chip that should be implanted as an interface to a computer (brain-machine interface, BCI) in the brain. The operation is to take over a robot.

Using ultra-thin electrodes, the chip is connected to nerve cells in the brain. The electrodes should detect signals of the nerve cells. The implant communicates with a transmitter worn by the implant user behind the ear, much like a hearing aid. The sender in turn connects to the smartphone on which a corresponding app is installed. For example, people with restricted mobility, such as paraplegia, are supposed to operate a computer.

Such systems already exist. But Neuralink boss Elon Musk would not be Elon Musk, if he did not add a pithy announcement: The implants should be able to improve the capabilities of the human brain. The vision is that in the distant future it should be possible to load new capabilities via the chip, such as a new foreign language.

Is that possible or science fiction? We have Moritz Helmstaedter asked. The physician, physicist and neuroscientist is executive director of the Max Planck Institute for Brain Research in Frankfurt am Main. He works in the field of Connectomy or Connectomics, a relatively young subdiscipline of neuroscience.


Job market

  1. BSH Hausgeräte GmbH, Traunreut
  2. OEDIV KG, Bielefeld



Golem.de: Mr Helmstaedter, where are we now? What can BCIs do today?

Moritz Helmstaedter: The goal is to interact with the brain directly rather than indirectly. That is to observe and stimulate the nerve cells in their activity. This has been tried for a long time, but in recent years there has been considerable progress in how precisely this can be done.

With the well-known classical methods like EEG or EMG One has first derived electrical activity in the brain from the outside. Then, ways were developed to stimulate neurons, both from the outside transcranial magnetic stimulation as well as invasive with so-called Deep brain stimulationthat really helps Parkinson's patients today.

The point now is to increase the resolution of these methodologies many times in order to read and control activity at the nerve cell level. But this is still at the beginning: There are initial successes to derive scientifically interesting signals. Now one tries to upscale that so that the network of nerve cells not only in individual places, but in ever larger proportions can be accessed. Neuralink took another step in that direction. Of course you have to be careful about many of the promises.