I was listening to Oregon Public Broadcasting’s (OPB) TED Radio about machine learning. For those of you who don’t know, machine learning is the next generation of artificial intelligence and is distinguished by the idea that machines learn from their own mistakes and can therefore evolve more quickly than humans. One of the implications of machine learning is that many jobs performed by humans will soon be done better by computers.
The obvious jobs being replaced are things like truck drivers with self-driving trucks. Other functions displaced could be call centers, medical diagnoses and more.
The interviewer of the OPB TED Radio program repeatedly asked the experts if this was a good thing or a bad thing. Would machines ultimately displace humans from all work? Is there a role for humans in the future?
The interviewer clearly thought this was a bad thing, but the experts didn’t. The experts instead thought it would usher in a period of extraordinary wealth for everyone and that meaningful work would come from the integration of human capabilities with computers. They cited the example of chess in which a team of people and machines always beat either just the humans or just the machine.
The experts argued that people are good at framing questions that machines should try to answer and machines were good at analyzing the data to answer the questions. This got me thinking about several aspects of the role of humans in performing work.
My view is that while huge amounts of resources have gone into machine learning, relatively little has gone into better understanding and developing human learning. I have heard some scientists say that we use 10 percent of our brains. I don’t know if this is true, but just suppose that we are seriously under-utilizing our mental capabilities, which I believe to be true from Cerebyte’s work.
Certainly, there are neuroscientists studying brain function, but most of it is at a level of basic research and understanding—the brain responds to x or y stimulus. There are many people drawing conclusions about the implication of this research, most of which can be labeled as “junk neuroscience.”
Overall, there are substantially fewer resources being allocated to understanding and improving the way we think than there are for machine learning. So, I want to present a hypothesis: We are just at the beginning of understanding human brain power—like where machine learning was 15 years ago.
What happens if we use all our brain power?
In our work at Cerebyte, we have seen huge jumps in people’s ability to deal with complexity and to improve productivity in challenging, real world situations simply by applying the lessons of neuroscience to learning, as presented by Kahneman and Tversky, in their book, “The Undoing Project.”
I believe that in the future, after a traumatic period of adjustment, we will find that some of the concerns about the ascendancy of machines decreases as the capabilities of humans increase. For this to happen, we will need to be open to the idea that many of the ways we approach the world will have to change. It won’t be easy.
Is your organization thinking about how to optimize brain power? Or is your organization stuck on machine learning?