Monsters, Megabytes, and the Limits of Human Consciousness
I’m staring at a split image. Both sides of the image have alternating pictures of celebrities: George Clooney, Tom Cruise, David Duchovney, Brad Pitt, Jennifer Aniston, Drew Barrymore, Cindy Crawford, Gwenyth Paltrow, etc.
The images are swapping out quickly at about 3 per second.
I’ve been instructed to stare at a small crosshair smack-dab in the middle of the split image. As the pictures fly past, these otherwise beautiful celebrities morph into monsters with misshapen heads, enormous eyes, and elongated lips.
Go ahead, try it for yourself:
We are daily bombarded with sensory input. Our eyes, ears, nose, mouth, and skin send a stream of 11 million bits per second of information to the brain for processing. However, the conscious part of the human brain can only process 41 bits per second (bps) of that deluge of sensory input.
Most of those 11 million bps streaming into your brain are filtered out. The 41bps that gets through to your consciousness is what creates your perception of reality.
The reality you think you’re experiencing is just what’s left over after your brain has filtered most of it out.
Back to those monstrous celebrities. What’s happening is that my consciousness is being saturated with input and pushing up against its 41bps limit. Consequently, my ability to perceive reality is beginning to break down and melt my world into a Salvador Dali painting.
If I spent enough time in this carefully engineered state, I would likely go insane. At least you’d think I was insane when I described my strange world to you because it would not resemble your reality at all. You see celebrities. I see monsters.
In the celebrities-become-monsters experiment, my consciousness has exceeded its Shannon limit.
Named after the “father of information theory” Claude Shannon, the Shannon limit (or “capacity” as it is also referred to) is the maximum rate at which data can be transmitted through a communications system with zero errors. This rate can be mathematically calculated. Our brain is a communications system and the Shannon limit of human consciousness is 41bps. Exceed that and reality starts becoming full of errors.
The Shannon limit of human consciousness is important because we’re actively pursuing the development of technologies that will push us far beyond it, perhaps with drastic consequences.
One such technology is being developed by Neuralink. Founded in 2016 by Elon Musk, Neuralink is developing “ultra high bandwidth brain-machine interfaces to connect humans and computers.”
According to Neuralink’s website, “We’re designing the first neural implant that will let you control a computer or mobile device anywhere you go.”
How does Neuralink propose to achieve this goal? By drilling a hole in your skull and inserting micron-scale threads of electrodes into your brain that connect to an implant they call “The Link”.
The threads are so fine and the surgery so precise that Neuralink is building a robotic neurosurgeon to “reliably and efficiently insert these threads exactly where they need to be.”
When RoboDoc is done doing surgery, you’ll be able to use the Neuralink app “to control your smartphone, keyboard and mouse directly with the activity of your brain, just by thinking about it.”
Initially, the purpose of Neuralink is to enable people with neurological injury or disease that has resulted in paralysis to control computers and other electronic devices.
Musk’s vision of Neuralink goes beyond that though.
“Over time I think we will probably see a closer merger of biological intelligence and digital intelligence,” Musk told an audience at the World Government Summit in Dubai in 2017.
“It’s mostly about the bandwidth, the speed of the connection between your brain and the digital version of yourself, particularly output,” Musk said. “Some high bandwidth interface to the brain will be something that helps achieve a symbiosis between human and machine intelligence and maybe solves the [AI] control problem.”
The “AI control problem” is the issue of how to avoid inadvertently building an artificial superintelligence (ASI) that will turn around and harm its creators.
Control problem evangelists such as Musk, believe that we must solve the control problem before an ASI is created. Failing to do so could result in the creation of an ASI that becomes self-determined, seizes control of its environment and makes decisions that are detrimental to the survival of the human race.
Musk is one of the earliest technologists to issue a clarion call for regulating the development of Artificial Intelligence (AI), famously comparing our pursuit of AI development to “summoning the demon” and the technology as “a fundamental risk to the existence of civilization”.
Neuralink seems to be a “if you can’t beat them, join them” strategy to merge with AI before, as Musk warned, “people see robots going down the street killing people”.
That’s one potential future. Another is one in which we build a high bandwidth brain-machine interface (BMI) so that we can interface directly with our electronic devices, the Internet, and all the hidden AI systems and machine-learning algorithms humming away in the background.
But in order to do this without going insane, we’ll have to figure out a way for human consciousness to exceed its current Shannon limit. Otherwise, the high bandwidth flow of data into our consciousness via a BMI will result in errors that begin warping our reality, sending us careening toward psychosis.
Perhaps this is already occurring to one degree or another in the accelerating technologically driven world we’re creating in which each of us are daily uploading an average of 34 gigabytes of information to our brain via electronic devices.
And while our pursuit of AI could turn out to be “summoning the demon”, perhaps the potential monsters we should be most concerned about are the ones that have already achieved self-determination, seized control of the environment and are currently making all the critical decisions that will determine the survival or the destruction of the human race.
