When Ada Lovelace programmed the first “computer” in the early 19th century, she never envisioned it to possess the capability to program itself. Now, we see virtual assistants and self-learning machines everywhere. Even in your email, Gmail organizes your mail into categories and learns to mark important mail. In fact, Spotify and Amazon use machine learning to determine which products you would like based on your past purchases. A more surprising use of artificial intelligence (AI) may be that on a Boeing 777, human pilots only fly for approximately seven minutes, with most of the flying done using artificial intelligence.

While the potential of general artificial intelligence strikes awe in some, the devastating possibility and effects of singularity lead others to stay far away from artificial intelligence’s enticing cry. Many are worried about the ability to create intelligent machines for warfare or to utilize any method to accomplish their task. With intelligent machines, any state or organizational actor who uses artificial intelligence in combat will induce other states to do the same as well. Thus, the future of warfare may be battlefields full of robots, sort of like the Terminator series playing out in reality. The more likely and most pressing scenario is machining putting human lives at risk while trying to do their jobs. For example, if robots see humans as an obstacle in completing an objective such as making paper clips, robots could seek to eliminate humans or transform the entire human race into paper clip-making slaves.

In the short term, we are already feeling the implications of artificial intelligence. The current narrow AI is already replacing blue-collar jobs. We can already see that in Korea, where chefs are being replaced by robots. As Uber begins to test out self-driving vehicles, we can assume that human drivers will be replaced in the near future by machines. Even worse, technology companies gather personal information in the name of delivering a personalized experience and better ad targeting. These same invasions of privacy with AI can be exploited by governments to establish or reinforce autocracies. For example, the Chinese Communist Party is working with Chinese technology companies to create a social credit system, rewarding those loyal to the Communist Party and punishing dissidents. In America, inherent bias present in human programmers or human sample sizes influences artificial intelligence, with machines only amplifying existing biases. For example, a Google image classification algorithm grouped two African American individuals as gorillas and another machine learning application was biased towards African Americans in predicting recidivism and in granting parole. This unacceptable racism results from machines being unable to sort biases from impartiality. The limits of technology for utilitarian purposes are clearly outlined in these cases.

While artificial intelligence may seem prevalent and alluring to many, its short- and long-term harms are clearly worse than the benefits it brings. While the convenience artificial intelligence brings seems to be everywhere, the privacy, security, and ethical issues it raises force us to reconsider the continued development of artificial intelligence. At the very least, we should establish firm boundaries to protect us from the seeping influence of artificial intelligence in our daily lives.