Are artificial intelligence (AI) and superintelligent machines the best or worst thing that could ever happen to humankind? This has been a question in existence since the 1940s when computer scientist Alan Turing wondered and began to believe that there would be a time when machines could have an unlimited impact on humanity through a process that mimicked evolution.

When Oxford University Professor Nick Bostrom’s New York Times best-seller, Superintelligence: Paths, Dangers, Strategies was first published in 2014, it struck a nerve at the heart of this debate with its focus on all the things that could go wrong. However, in my recent conversation with Bostrom, he also acknowledged there’s an enormous upside to artificial intelligence technology.

You can see the full video of our conversation here:Since the writing of Bostrom’s book in 2014, progress has been very rapid in artificial intelligence and machine and deep learning. Artificial intelligence is in the public discourse, and most governments have some sort of strategy or road map to address AI. In his book, he talked about AI being a little bit like children playing with a bomb that could go off at any time.

Bostrom explained, “There’s a mismatch between our level of maturity in terms of our wisdom, our ability to cooperate as a species on the one hand and on the other hand our instrumental ability to use technology to make big changes in the world. It seems like we’ve grown stronger faster than we’ve grown wiser.”

There are all kinds of exciting AI tools and applications that are beginning to affect the economy in many ways. These shouldn’t be overshadowed by the overhype on the hypothetical future point where you get AIs with the same general learning and planning abilities that humans have as well as superintelligent machines. These are two different contexts that require attention.

Today, the more imminent threat isn’t from a superintelligence, but the useful—yet potentially dangerous—applications AI is used for presently.

How is AI dangerous?

If we focus on what’s possible today with AI, here are some of the potential negative impacts of artificial intelligence that we should consider and plan for:

Change the jobs humans do/job automation: AI will change the workplace and the jobs that humans do. Some jobs will be lost to AI technology, so humans will need to embrace the change and find new activities that will provide them the social and mental benefits their job provided.

Political, legal, and social ramifications: As Bostrom advises, rather than avoid pursuing AI innovation, “Our focus should be on putting ourselves in the best possible position so that when all the pieces fall into place, we’ve done our homework. We’ve developed scalable AI control methods, we’ve thought hard about the ethics and the governments, etc. And then proceed further and then hopefully have an extremely good outcome from that.” If our governments and business institutions don’t spend time now formulating rules, regulations, and responsibilities, there could be significant negative ramifications as AI continues to mature.

AI-enabled terrorism: Artificial intelligence will change the way conflicts are fought from autonomous drones, robotic swarms, and remote and nanorobot attacks. In addition to being concerned with a nuclear arms race, we’ll need to monitor the global autonomous weapons race.

Social manipulation and AI bias: So far, AI is still at risk for being biassed by the humans that build it. If there is bias in the data sets the AI is trained from, that bias will affect AI action. In the wrong hands, AI can be used, as it was in the 2016 U.S. Presidential election, for social manipulation and to amplify misinformation.

AI surveillance: AI’s face recognition capabilities give us conveniences such as being able to unlock phones and gain access to a building without keys, but it also launched what many civil liberties groups believe is alarming surveillance of the public. In China and other countries, the police and government are invading public privacy by using face recognition technology. Bostrom explains that AI’s ability to monitor the global information systems from surveillance data, cameras, and mining social network communication has great potential for good and for bad.

Deepfakes: AI technology makes it very easy to create “fake” videos of real people. These can be used without an individual’s permission to spread fake news, create porn in a person’s likeness who actually isn’t acting in it, and more to not only damage an individual’s reputation but livelihood. The technology is getting so good the possibility for people to be duped by it is high.

As Nick Bostrom explained, “The biggest threat is the longer-term problem introducing something radical that’s super intelligent and failing to align it with human values and intentions. This is a big technical problem. We’d succeed at solving the capability problem before we succeed at solving the safety and alignment problem.”

Today, Nick describes himself as a “frightful optimist” that is very excited about what AI can do if we get it right. He said, “The near-term effects are just overwhelmingly positive. The longer-term effect is more of an open question and is very hard to predict. If we do our homework and the more we get our act together as a world and a species in whatever time we have available, the better we are prepared for this, the better the odds for a favourable outcome. In that case, it could be extremely favourable.”

Introduction