Almost everyone talks about how computers and artificial intelligence are replacing human jobs. While this is serious, we’d like to talk about another aspect of artificial intelligence. As artificial intelligence performs more and more jobs, human society begins to lose the ability to choose its own destiny. We lose our self-determination. Let’s explore this topic a little further.
The human brain has a processing capacity of about 38 pentaflops. One pentaflop is one million billion (10^15) floating-point operations per second.
At one time people wondered if we could ever build a computer that could process information as fast as the human brain. Today current supercomputers already have a processing power greater than the human brain. By 2018 they will be at something like 3-5 times the processing power of a human brain.
Unlike the human brain, which is limited by the capabilities of our physical body, computing power has no natural limit that we know of. Computers have been doubling in processing power every 1-2 years for 50 years and the rate does not seem to be slowing.
Artificial intelligence (AI) is different from computing power. Artificial intelligence is a combination of hardware and software designed to function like animal intelligence. AI requires computing power to work, but it goes beyond raw processing power. It requires the development of extensive software, or computer code, to work.
The growth rate of computer software has been roughly the same as that of computer hardware, so AI is also doubling about every 1 to 2 years as well. AI development lags a bit behind computing power and AI has not yet reached the level of human intelligence.
Estimates are AI will reach the level of mouse intelligence somewhere around 2020. Humans are about 20 times smarter than a mouse. So if it continues to double every two years AI will be at the level of human intelligence somewhere around 2030. See Figure 1 where the vertical axis has units of mouse intelligence, so a 1 equals a mouse intelligence and a 20 equals human intelligence.
There’s no reason AI will stop at that level given current trends, unless we decide to limit it. At its normal growth rate by 2040 AI will be about 20 times as smart as a human. To AI at this level humans will be about as smart as a mouse seems to us. See Figure 2.
Human Self Determination
Self-determination is the process by which you control your own life. If you have 100% self-determination you are 100% in control of your own life. If you have 0% self-determination you have no control over your own life. As artificial intelligence makes decisions that were once made by humans, humanity loses its self-determination.
Artificial intelligence is already creating computer-generated news and financial reports, controlling robots, and providing advice on parole decisions and banking loans. These things were once done only by humans. As AI evolves it will do more and more jobs and make more and more decisions that were once made by human beings. As this happens humanity begins to lose its self-determination. See Figure 3.
There’s a tendency for people to talk about which jobs AI will replace. Many studies say a college education will be good insurance against being replaced by AI. It’s interesting to note that in 2017 an AI robot scored in the top 20% on the entrance exam to Tokyo University, the most prestigious university in Japan. The exam included a 650 word essay.
Unless we begin to think about ethical limits to AI at some point humanity will lose all of its self-determination. This is the point where humans are like mice who are being taken care of by artificial intelligence. As AI evolves it will also concentrate power in the hands of fewer and fewer people. Small numbers of people will be able to control large portions of technology and society. If we use history as our guide at some point a few small groups will attempt to control everything.
Once again we need to consider the ethical implications of AI. Tremendous power has always raised ethical issues. As we mature as a species we have been able to handle more and more powerful technology in an ethical way. Ai is simply the next jump in techniology that we need to learn how to ethcially manage.
Human Operating Systems
Other people say there’s nothing we can do to ethically manage artificial intelligence . They say it’s impossible for human beings to limit artificial intelligence.
We agree it’s probably not possible for the current version of human beings, that is to say late industrial human beings, to solve this problem. After all we’re the ones who created it. As Einstein famously said a problem can’t be solved at the same level it was created on.
In the industrial mindset bigger, faster, more efficient is the order of the day. Continuing to drive AI forward fits this industrial mentality. A best best case senerio for humanity, if we continue to drive AI forward without ethical management of its power, will be for humanity to end up on a nice nature preserve.
One way of looking at the big picture is to say that if humanity can’t solve this problem with our current operating system it’s the universe’s way of telling us it’s time to upgrade to the next human operating system.
What do we mean by human operating system? We mean the combination of culture, technology, and personal beliefs that we create around ourselves to organize human behavior. As you can see in Figure 4 about 40,000 years ago the first humans were hunter gatherers. With the hunter-gatherer operating system we had the ability to make some tools and we had the social skills to organize into small groups.
Then about 12,000 years ago we discovered agriculture and built a new operating system for ourselves. This completely reshaped us as a species. With the agricultural operating system we had the ability to organize into larger cultural groups and we learned how to work with the natural world so that we could feed ever larger numbers of people. This is the era of large empires. The humans who adopted the new agricultural operating system out competed those who remained behind in the hunter gatherer systems.
About 250 years ago humans began builing the industrial operating system. The industrial operating system knew how to harness fossil fuels and put them to work in machines to make the goods and services we needed. The industrial operating system gave rise to the modern world as we know it with smart phones and artificial intelligence.
Designing our Next Operating System
These previous upgrades in the human operating system were the result of steady progress and our ability to create new technologies and social structures. They unfolded naturally as humanity progressed. I doubt if anyone woke up one morning and said “tomorrow I’m going to invent agriculture” or “next week I’m going to invent a steam engine that will end up reshaping the world.”
Those were the good old days when we had time to allow things to play out naturally. When we didn’t have to worry about consciously planning our own evolution. Now there is broad agreement that the industrial era is winding down. Not because it’s not important but because it’s becoming automated and requires less human attention. The same thing happened with agriculture 250 years ago.
So there’s a new human operating system that’s rising. You often hear it referred to it as the rise of the creative culture. Whatever you call it’s underway. The new twist is that now we know we have 10-20 years to evolve into a creative culture that’s capable of ethically managing artificial intelligence. We need to ethically manage AI in the same way we’ve ethically managed any other powerful technology with the potential for misuse. In the same way we have ethcially managed atomic energy. Some things are too dangerous to our existence as a species to go unmanaged. Artificial intelligence is one of those things.
On the positive side we are standing on the shoulders of giants. We have the accumulated knowledge of humanity available at our fingertips. Our understanding of how the world works and how our brains and bodies work is growing daily. We can use this knowledge to drive our transformation.
On the negative side we’ve gotten out of balance as a species due to the rapid pace of technological development. Our social and cultural skills are lagging behind our technology. Many people and countries have no real goal other than growing their economy and possessions faster and faster. We need to evolve our social and ethcial abilities to catch up with our technical abilities.
Next Steps, AI Driving Human Transformation
What do we want the future to look like? How do we get there? These are big questions we are faced with. Ravi Ravindra, who teaches about the Bagvhad Gita, likes to say there are two kinds of problems. The first type are simple and can be solved by acquiring knowledge. We often assume all problems are like this.
The second type of problems are the significant problems. These problems cannot be solved by merely acquiring knowledge. To solve them you must transform yourself, you must move yourself to a higher level. You must turn yourself into someone who can solve the problem. Ethically managing AI is this type of significant problem.
For more information see the Category Trip Preparation.