Print the page
Increase font size
Musk-Brand Brain Surgery

Posted November 09, 2023

Ray Blanco

By Ray Blanco

Musk-Brand Brain Surgery

“What was going on in that head of hers? It was exasperating, really, not knowing. It was an affront, a deprivation, to herself and to the world. She would bring this up at the earliest opportunity. They needed to talk about the thoughts she was thinking. Why shouldn't they know them? The world deserved nothing less and would not wait.”

From The Circle - by Dave Eggers

If there’s one thing that truly belongs to you, and nobody else, it’s the thoughts in your head.

They’re totally private and completely inaccessible to others. For better or worse.

As dystopian as it may sound, this is all starting to change.

There are a number of recent studies and experiments that have led to significant advancements in brain-computer interfaces, which would allow someone’s thoughts to be recorded and played to others.

While this may sound scary, it’s far from being a mind-reading machine and could offer a way for people suffering from ALS or experiencing complete paralysis to communicate with others.

Researchers at the University of California have made significant strides in brain-computer interfacing over recent years, with multiple ongoing studies. One of these studies used a very peculiar method to track and interpret brain activity.

At UC Berkeley, a song was played for patients with treatment-resistant epilepsy while they were undergoing surgery. Using electrodes, the patients’ brainwaves were recorded, interpreted using an algorithm, then replayed.

That song, Pink Floyd’s Another Brick In The Wall, was easily recognizable when replayed from the brainwaves of the 29 patients, after just one listening.

The use of music in this experiment is meant to focus on enriching speech, so that interfaces don’t all sound like Siri.

Another study at the University of California, San Francisco has already made surprising progress in interpreting a patient’s brainwaves into fully formed words.

Using a brain implant, users are already able to “talk” at a rate of up to 80 words a minute, roughly half the normal speed of speech and over three times the previous record for a brain-computer interface.

The key puzzle to solve for this process was how to decode electric signals from the brain into text.

Of course, this was a job for an AI algorithm.

The algorithm was trained to not necessarily look for specific words, but the base parts of speech called phonemes.

Co-author of the study Alexander Silver explained the distinction like this…

“If you make a P sound or a B sound, it involves bringing the lips together. So that would activate a certain proportion of the electrodes that are involved in controlling the lips.”

This method currently lets patients “telepathically” communicate with an error rate of only 9%. This was using a limited vocabulary of only 50 words. The error rate rose to 24% when using a vocabulary of 125,000 words, so there’s still a long way to go for this technology.

Monkey Business

The most well-known example of brain-computer interfacing is likely the Elon Musk founded Neuralink.

Neuralink is best known for its experiments using monkeys, both successfully and unsuccessfully.

In 2021, Neuralink released a video of macaque playing Pong, the classic video game, using only its thoughts.

Reportedly, the bluetooth-enabled chip implanted into monkey’s head could feed neural information into a decoder that could predict its desired hand movements, which would then move the on-screen paddle without any physical movement from Pager, the experiment’s subject.

More recently, the company showed a monkey doing what Musk described as “telepathic typing”.

The test subject quickly and accurately would navigate an on-screen keyboard using only its mind to click highlighted keys. 

Since the implanted chip doesn’t grant its user super-intelligence, the monkey wasn’t actually spelling words out on its own. But the experiment demonstrated how a human could easily use this interface to communicate effectively.

And we’re closer than ever to seeing exactly that. After the FDA approved Neuralink for human trials earlier in the year, the company is now seeking its first volunteer to receive a neural implant.

This ideal candidate will be a young adult quadriplegic. This first human test subject will have a portion of their skull removed as to receive Neuralink’s implant within the “hand knob” area of their premotor cortex.

Musk has issued bold promises of what can be accomplished using this new technology. When Neuralink was first demoed in 2019, the billionaire said, “This is going to sound pretty weird, but ultimately we will achieve symbiosis with artificial intelligence”.

The companies Synchron and Onward have both successfully tested their brain-computer interfaces on humans within the last year.

With that, we’d like to hear your thoughts. Do you think recording someone’s thoughts creates any ethical issues? Is it worth it to help someone who is unable to communicate physically? Share your thoughts at feedback@technologyprofits.com

AI Gets Political

AI Gets Political

Posted January 17, 2024

By Ray Blanco

AI takes center stage in Davos. How this week will shape what AI looks like for decades to come.

Consumer Electronics Show Debrief!

Posted January 16, 2024

By Ray Blanco

A run down of what the Paradigm crew learned from CES in Vegas - plus the week’s top tech stories.

Filtering Out Fake News

Posted January 12, 2024

By Ray Blanco

AI offers a unique solution to the Fake News epidemic.

Should I Buy Bitcoin?

Posted January 11, 2024

By Ray Blanco

It’s official, the Bitcoin ETF has been approved. Its price keeps going up, but should you buy-in now?

Live From CES in Vegas!

Posted January 10, 2024

By Ray Blanco

Zach, Matt, Ari, and Bob are providing live updates from the Consumer Electronics Show in Las Vegas.

Bigger Than Bitcoin

Posted January 09, 2024

By James Altucher

James Altucher tells us why he’s not bothering with the next Bitcoin bull run, even though he thinks it’s legit.