“ There are no strings on me”. Ultron was a superior artificial intelligence(AI) program hell bent on wiping out the human race. He struck fear into the hearts of millions as he created his own weapon of mass destruction. Although this plot is pure fiction, our fear of AI technology one day ruling over us lowly humans is not as without premise. For the past few decades, the promise of AI technology has been nothing more than disappointing. However, in recent years we have seen great strides in the development of a truly superior AI tech. This progress has led us to a discussion that we must have now. Are we willing to trust AI and what are the risks of doing so.
We have seen so many movies from Terminator to iROBOT where the advancement and implementation of AI technology in our daily lives ultimately leads to chaos. Before we look into the plausability of that scenario, we need to understand what is AI. AI is basically intelligence acquired by machines. This intelligence is usually associated with the ability to solve problems and adjust to changing environments. A very good example would be self-driving cars where the program has to constantly assess its surroundings to determine the best course of action to take. Now many critics argue that because machines don’t need to rest like humans do, if they are able to constantly learn and adapt to its environment, who is to say they won’t learn that they are smarter and more superior than us. Once the machinese figure that out, they could easily take humans on. With their superior intelligence, it would be as plain as David vs Goliath.
The situation is further exacerbated when we integrate AI into our daily lives. If we were up against clunky robots like in the Avengers, we might still have a fighting chance. AI has become increasingly ubiquitous in modern society. Soon it may be an integral part of education, healthcare and even our military. Imagine if the world’s nuclear launch codes were in the hands of an AI program that might go rogue. The age of the internet of things has arrived and if AI programs are able to control almost all our devices, they would ultimately control us. This is the risk we face when we allow AI into our lives. The risk of course is limited as scientists and engineers are not idiots too. They can code backdoors into AI programs and shut them down when they go rogue. However there is always a risk of the AI machines learning of this backdoor and blocking it before we get to use it. This scenario may seem outrageous, but the low probability is coupled with unimaginably huge consequences.
Even assuming the best case scenario, where machines live under the rule of humans forever, the ending isn’t happily ever after. Many argue that AI will take away ALL our jobs. Not just menial jobs, not just some jobs, but all jobs. During the Industrial Revolution, mechanized labor put hundreds of thousands out of jobs. A lot of people kept their jobs because at the time, machines did not have what humans did, intelligence. With AI, every job is at stake. A machine could be a lawyer, doctor, or even a police officer. We would see massive structural unemployment because how could a human compete with a machine which never gets tired, hungry or careless at times.
The world has been blown away with what AI can do, but they should be worried about what AI may do. This article serves not to be a party pooper, but a reminder that all big leaps have consequences. We need to engage in a clear discussion about how to proceed with AI before the strings come off.