The Risks of AI Advancement
Does it ever feel like you have more employment worries than before? Or perhaps, do you feel threatened by the creation of something that’s more intelligent than human beings? It’s evident that the development of AI and the construction of a superintelligence come with numerous issues and, in the worst case scenario, could lead to the end of humankind.
You might believe that the thought of something taking our jobs is a new concern. However, we’ve been at this crossroads before. You can glance at history books or old newspapers and notice that workers have been feeling threatened by “new tools” for ages. The only element that’s changed is the type of machine provoking these worries.
As an example, throughout the early nineteenth century, 80 percent of US workers had a job in agriculture. However, that figure has fallen to less than 2 percent due to the machines that perform the primary duties associated with farming, livestock and working the land. Therefore, in 2013, when a study at Oxford University predicted that half of all American jobs were under threat of being automated over the next decade, people were right to be worried.
Despite the billions invested in the consumer technology of smartphones and apps, productivity has not improved significantly. Moreover, when you compare the growth in annual wages in the United States from 1991 to 2012, it was roughly half the increase that took place between 1970 and 1990.
Over the past few decades, the mechanization of the industrial workplace has been continuous. The large factories that were once crowded with people are now working on row after row of machines. However, even though the factory floors have 90 percent fewer people on them, it doesn’t mean that every worker has to despair about losing his or her job. This is because technology has routinely created new jobs as much as it takes them away.
Based on the dozens of studies Malcolm Frank analyzed in “What To Do When Machines Do Everything,” we can expect around 12 percent of US jobs to be automated over the next ten years, displacing around 19 million workers. However, these same studies also predict that these technologies should create 21 million new jobs, which would hold the unemployment totals of 2025 roughly the same as they are now.
A more significant concern, Nick Bostrom argues, is the unsafe development of a computer that is more intelligent than humankind, a superintelligence. We have to examine every possible scenario before bringing an uncommonly powerful force like SI into the world.
For instance, imagine that some sparrows adopted a baby owl. Having a loyal owl around might be very beneficial; the more powerful bird could guard the young, search for food and do any number of other tasks. However, here comes the problem: what if the owl realizes it’s an owl and eats all the sparrows? Therefore, a rational strategy would be for the sparrows to plan an excellent strategy for how to teach the owl to love sparrows, while also analyzing all possible outcomes in which the owl could become a negative force.
So what is our plan for making our superintelligent baby owl love us?
We can make safety a priority by creating the SI through international collaboration. Safety restrictions could be put in place and respected by all people and organizations who work on developing it. However, why would the competitive rush to design the first SI be a safety threat?
It’s because specialists would likely sacrifice safety to advance their process and wouldn’t share their work with others. That means that if an SI project went wrong and threatened humanity with extinction, too few people would know the machine’s design well enough to stop it.
On the other hand, if governments, institutions and research groups join together, Nick Bostrom argues that they could slowly build a safe and highly beneficial SI. That’s because groups could share their ideas for safety measures and provide thorough oversight for each phase of the design. Not only that, but an international superintelligence project would promote peace through its universal benefits. Just think about the International Space Station, an effort that helped stabilize relations between the US and the USSR.
There’s a new revolution coming, and the instrument at its core is a system of intelligence powered by self-learning software and massive amounts of data. Jobs are going to be taken away, as well as new ones created. However, our biggest concern should not be computers taking our jobs; it should be, as cliche as it sounds, computers taking over the world. We must prioritize safety over unchecked technological progress. The fate of our species depends on it.
Sources:
Nick Bostrom, Superintelligence: Paths, Dangers, Strategies, Oxford University Press, Reprint edition, 2016
Malcolm Frank, What To Do When Machines Do Everything: How to Get Ahead in a World of AI, Algorithms, Bots, and Big Data, Wiley, First edition, 2017
Kai-Fu Lee, “The Real Threat of Artificial Intelligence,” The New York Times, 2017
