Artificial Intelligence and Consciousness
Last year, a pretty amazing thing happened at a lab at New York’s Rensselaer Polytechnic Institute. A robot demonstrated self-awareness for the first time.
The experiment was called the Wise Man test and it worked like this: the lab took three robots and programmed them all to believe that two of them had been given “dumbing pills” that prevent them from speaking - but they didn’t know which of them received it.
He then asked them which ones got the dumbing pill, and the robot answered, and then realized that because it was able to answer, it couldn’t have gotten the pill.
This is something no computer or robot has been able to do before.
Artificial Intelligence has grown by leaps and bounds just in the last few years, with the self-awareness test I just talked about but also the computer AlphaGo beating a world-class Go champion earlier this year.
These are not small things.
It was always believed that if you could combine the computational power of computers with the creative and innovative thinking of a human brain, you’d have something with unlimited intelligence that could solve all of our problems.
Or wipe us off the face of the Earth.
One of the pioneers in consciousness research is Sir Roger Penrose.
He is the Emeritus Rouse Ball Professor of Mathematics at the University of Oxford, and spent his early career tacking cosmological questions relating to general relativity such as the cosmic censorship hypothesis and the Weyl Curvature hypothesis. He has been awarded the Eddington Medal of the Royal Astronomical Society and the Wolf Foundation Prize for Physics, both of which he shared with Stephen Hawking, and in 1994, he was knighted for his contributions to the world of science.
Beginning in the 1980’s, he turned his attention to the questions of consciousness and the brain, using math and physics to try to understand how that creates our consciousness.
He’s written two books on the subject, including The Emperor’s New Mind: Concerning computers, Minds, and the Laws of Physics, Shadows of the Mind, A Search for the Missing Science of Consciousness.
The main point that Penrose makes, and the thing that has earned him a bit of notoriety, is his conclusion that consciousness can’t be explained by conventional physics.
His assertion is that the structure and connections of neurons in our brain simply does not physically hold the computational capacity to create the consciousness we experience.
He argues that there must be something else at play, at an even smaller level, and he believes that exists in quantum space.
So along came Stuart Hameroff, an anesthesiologist, who became intrigued by Dr. Penrose’s work, and brought to Penrose the idea that there are structures inside brain neurons called microtubules, and that they may hold the key to allowing the brain to compute within the quantum space.
Microtubules make up the cytoskeleton inside the cell and hold it together. They’re tiny - only nanometers across - and Hameroff suggested that the tiny spaces inside the hollow tubules allows for qubits, or quantum bits of information, to communicate with each other.
Working together, they created the Orch-OR model of consciousness, which states that consciousness in the brain originates in processes inside of neurons, instead of through the connections between neurons.
Interestingly, a study in 2014 at the National Institute for Materials Science in Japan discovered quantum vibrations in microtubule latices that proponents of Orch-OR believe is evidence in favor of the model.
But this theory is hotly contested, with most critics arguing that the brain is too “warm, wet, and noisy” to create a coherent quantum system. This has become known as the Warm, Wet, and Noisy argument.
As AI research continues, we’re finding that the key is not in simple processing power, but in neural networks.
There’s a fascinating book called Who’s in Charge? by neuroscientist Michael Gazzaniga who has worked with split-brain patients over the last 20 years or so.
Split-brain patients are people who have had their left and right brains disconnected from each other. Our hemispheres are connected by a bundle of fibers called the corpus collosum and back in the 70’s the discovered that if you cut the corpus collosum, you can actually stop seizures in people who were getting them on a daily basis.
So there are hundreds of people who have had this surgery done and it’s produced some very unexpected results.
For example, the visual cortex usually resides in the left hemisphere, so when people were shown images of say a chicken in the right hemisphere, they literally thought they weren’t seeing anything. Their right brains couldn’t process it in a way that allowed them to put it into words.
But, when a pencil was put in their left hands, their left hands would draw a chicken. And then, when asked why they drew a chicken, they would respond with, “well I ate chicken yesterday so that must be why."
We also have a number of subconscious modules in our brain that cause us to take actions and decisions before our conscious brains can decipher it.
The story here is that what we perceive as a holistic conscious experience is a collection of a dozen or more conscious modules in our brain making decisions for us which are later justified by an interpreter module.
In other words, our conscious experience… might just be an illusion.
For now the answer to the hard problem of consciousness remains out of reach, but as research into artificial neural networks and quantum computing continues to progress, we might be on the cusp of some very relevatory stuff. Stay tuned.