Elon Musk’s Neuralink received approval last week from the US Food and Drug Administration for human clinical trials, which a former FDA official called “really a big deal.” I don’t disagree, but I am skeptical that this technology will “change everything”. Not all profound technological advances have broad social and economic implications.

With Neuralink’s device, a robot surgically inserts a device into the brain that can then decode some brain activity and connect the brain signals to computers and other machines. A person paralyzed from the neck down, for example, could use the interface to manipulate her physical environment, as well as to write and communicate.

In fact, this would be a huge advance for people with paralysis or traumatic brain injuries. For others, I’m not so sure. For the sake of argument, since there are many companies working in this space, assume that this technology works as advertised. Who exactly will want to use it?

One fear is that brain-machine connections are expensive and that only the rich can afford them. These people will become a new class of “super thinkers” who will rule over us with their superior intellects.

I don’t think this scenario is likely. If I were offered $100 million for a permanent brain-computer connection, I wouldn’t take it, if only for fear of side effects and possible neurological damage. And I’d like to know for sure that the control nexus goes from me to the computer, not the other way around.

Also, there are other ways to increase my intelligence with computers, especially the recent innovations in AI. It’s true that I can think faster than I can talk or write, but I’m not in such a rush. I would rather learn to type on my phone as fast as a teenager does.

A related view of the direct brain-computer interface is that computers will be able to rapidly inject useful knowledge into our brains. Imagine going to bed, turning on your brain device, and waking up knowing Chinese. It sounds incredible, but if that were possible, so would all sorts of scenarios, not all of them benign, in which a computer can alter or control our brains.

I also see this scenario as remote – unlike using your brain to manipulate objects, it sounds like true science fiction. Current technologies read brain signals but do not control them.

Another view of this technology is that computer owners will want to “rent” the powers of human brains, much as companies rent space in the cloud today. Software programs are not good at some skills, such as identifying objectionable images or speech. In this scenario, connected brains largely come from low-wage workers, just as social media companies and OpenAI have used low-wage labor in Kenya to grade production quality or help make quality decisions. content.

Those investments can be good to increase the salaries of those people. Many observers may object, however, that a new and more insidious class distinction will have been created: between those who have to plug into machines for a living and those who don’t.

Could there be scenarios where higher paid workers want to be connected to the machine? Wouldn’t it be useful for a spy or corporate negotiator to receive real-time computer intelligence while making decisions? Would professional sports allow such brain-computer interfaces? They can be useful in telling a baseball player when to swing and when not to.

The more I reflect on these options, the more skeptical I become about large-scale uses of brain-computer interfaces for people without disabilities. Artificial intelligence has been progressing at an astonishing rate and does not require any intrusion into our bodies, let alone our brains. There are always earplugs and some future version of Google Glass.

The main advantage of the direct brain-computer interface appears to be speed. But extreme speed is important only in a limited class of circumstances, many of them zero-sum competitions and efforts, such as sports and games.

Of course, companies like Neuralink can prove me wrong. But for now I’m still betting on artificial intelligence and big language models, which are just a few comfortable inches away from me as I write this.

© 2023 Bloomberg L.P.