For the last half a century, a question that has bugged scientific minds is ‘which between human and machine is smarter?’ When Deep Blue defeated the world human chess champion, at the time, it seemed the machines had upped humans. Coupled with the rise of AI, in some quarters at least, machines did appear to have dwarfed the intelligence of man.
However, the machines were not going to have a smooth win. After the Kasparov defeat, debates soon followed centering on different questions. What really constitutes intelligence? Do computers possess intelligence? If computers did possess intelligence, do we currently measure such intelligence accurately? Many questions, many solid debates, but unfortunately, the consensuses were not as many.
In this mesh of debates and questions, some have started asking “what if we are doing this all wrong?” Perhaps, we are blind to the fact that humans and machines both have unique, complementary strengths. Which means we should not be contrasting them. Rather we should be considering the viable possibility of developing a supremely intelligent cyborg from merging humans and machines.
Yipeng Yu and his team of researchers set out on this course when they compared the problem-solving capabilities of rats, computers and rat-computer (cyborgs).
What the team first did was to train six rats to run a series of unique mazes within a one-week period. Then they placed microelectrodes in the brain of the rats, particularly in their somatosensory cortex and the medial forebrain bundle. The team then coerced the rats to complete the maze run with a sip of water, fragrance of peanut butter, and stimulation of the medial forebrain bundle (which is a key node of the brain’s reward system).
When the rats were set, the scientists tested the rats on 14 new mazes. They monitored the rats’ paths, strategies and total time spent completing the mazes.
The research team also developed a maze-solving computer algorithm to solve the exact same 14 mazes run by the rats. The algorithm used the left-hand and right-hand wall-following rules.
The research team then complemented the biological intelligence of the rat with the artificial intelligence of a computer algorithm.
How this worked was that when the rats needed help in choosing a direction, the algorithm intervened by stimulating the left and right somatosensory cortex of the rats to prompt them to move left or right respectively. In so doing, the rats avoided dead ends, loops and took unique paths as well.
To contrast the performance of the rats, computer and rat cyborgs, the team evaluated how many locations each visited in the course of completing the maze, how many each visited the same location (steps) and the total time each sent to complete the maze.
From their findings, the cyborgs and computers were roughly at par in the number of steps taken. Interestingly, the cyborgs took less steps than rats, evidence that cyborgs are more adept at problem solving that their fully biological counterparts.
In addition, the cyborgs upped their counterparts (computers and rats) in terms of the number of locations visited. In other words, the cyborgs visited fewer locations.
Finally, from the findings a specific maze was just as challenging for the rats’ as it was for the computer and the rat cyborg. The researchers found this out when the number of steps and locations covered showed strong correlations between the types of beings across various maze layouts.
The potential of cyborgs and ethics concerns
From the research findings of Yu and his colleagues, it does appear that the notion of optimal intelligence residing exclusively in either man or machine is untrue. Rather, the integration of man and machine does hold more promise in the drive for optimum intelligence. And why not? Artificial computing systems have the luxury of speed and logic, while biological neural systems like the human brain have remarkable cognitive abilities.
Nonetheless, obvious concerns arise. Could cyborgs put humans at risk for some reason? Is it possible that we are actually going over the top by altering human behavior with a machine?
Whatever the concerns are, matching man and human is so pragmatic that we have already started using such connected systems. Surprisingly, these systems are even in widespread use. For decades, we have utilized human brain-computer interface to help restore movement, vision and communication.
In fact, while your brain may not have a direct connection to a computer using electrodes, it is fair to say you occasionally function as a human cyborg more often than you realize. Some of us augment our navigation abilities while driving with GPS. A word-processor spell-checker helped enhance the author’s writing of this article. A digital planner literally maintains your mental health by organizing your busy schedule.
Therefore, while drawing the line between harmless lifestyle enhancement and dangerous mind-control can be a difficult endeavor, perhaps it is about time we embraced the computing abilities of machines as complementary and beneficial to the natural powers of intelligence we possess.
The details of the research can be found in PLOS ONE.