Is AI Good For Humanity? Will AI and Machine Learning Make Humans Better?

Over 20 years after the new millennium and science fiction seems to become less fiction every single day. We may not have commercial flying cars just yet, but technology has advanced far beyond our wildest dreams in many other places. Photographers, cinematographers, and tech geeks alike will all have an appreciation for what it means to have a full terabyte of storage space on a micro SD card or a thumb drive. Augmented reality lets us use our own phones to visualize potential furniture purchases in our own homes using nothing but the device in our pocket. Maybe the most mysterious and exciting advances of all, however, is Artificial Intelligence.

It can be a scary thought for many: machines with a mind of their own. Truthfully, however, this is a bit of a mischaracterization of what artificial intelligence actually means. That is not to say that caution is not important. Let’s explore a little bit about what AI is, how it can be good, and how it can be bad for humanity.

What is AI?

Artificial Intelligence can be difficult to understand, in part, because it does not have a very cohesive definition. Britannica defines AI as “the ability of a digital computer or computer-controlled robot to perform tasks commonly associated with intelligent beings.” Paraphrased, you might say that AI is essentially “computers that seem smart,” which is not very precise.

Because of this, AI refers to a very broad range of algorithms, technologies, and ideas that fit into this description. One such idea is that of machine learning, which is often used as a synonym for artificial intelligence, but is actually only one branch of this tree. The illustration that machine learning gives us is still useful for understanding these terms, however. By feeding a machine learning algorithm large amounts of data as a reference for “training,” the algorithm will extract patterns from the information and be able to make decisions and predictions that rival and sometimes surpass human intuition or insight.

Machine learning “mimics” the way that humans learn—recognizing patterns from past experiences—and artificial intelligence mimics human intelligence in a broader sense, but how can this be helpful?

The Good

Machine learning and AI more broadly have already shown to have significant benefits for humanity in a wide range of circumstances. For businesses, AI customer service chatbots have come into a boom for online brands. With a properly trained and implemented algorithm, these autonomous bots are not only available 24/7 for customers at the click of a button, but they are likely able to access information for the customer much faster and more efficiently than what a person would be able to do. It’s easy to argue that there is simply no replacement for humans in customer service, but at the very least for relatively quick and simple tasks, these AI widgets improve the quality of life for the customer and allow the brand to focus on more challenging issues.

On a much more vital level, machine learning and AI can and has been implemented to literally save lives. When it comes to medical diagnoses, there is often some crucial (educated) guesswork at play. While House M.D. might be highly dramatized, the truth is that many different illnesses can cause similar symptoms, and mapping the right symptoms to the wrong illness can certainly be a matter of health, life, and death. Powerful AI algorithms are simply able to crunch more numbers than we are, and consider more details to a greater degree of precision, which can help eliminate some of this guess work and give medical professionals an extra tool to help save lives—just as DeepMind showcased.

The Bad

These algorithms are able to process and analyze data exponentially faster than any human would be able to and, in a way, this is both a strength and a weakness. Although this leads to incredible insights, it also means that AI is not able to create original ideas—it can’t have thoughts of its own. The implication here is that AI is not some pure, transcended consciousness that knows us better than we know ourselves, it is simply an extension of the information that we already have, the information that we are feeding the algorithm. As a result, the conclusions that AI programs might draw are bound to be tainted with our own biases and flaws.

An important and recent example comes with policing. Using AI to predict and prevent crime may sound like a great idea for public safety, but where this data comes from needs to be considered. It is no secret that marginalized communities in the United States are over-patrolled and over-policed and, as a consequence, account for higher arrest rates. Arrest rates, however, are not crime rates. This means that the only data we can feed these algorithms has a highly racist bias and, as such, an AI crime-stopper program would only replicate these biases.

These are the kinds of issues and limitations that come with artificial intelligence. It’s not that “AI is racist,” it’s that humans collect and shape our own data. When that data reflects our internal prejudices and biases, the AI will follow those trends by command.

The Verdict

So, is AI good or bad for humanity? The answer is boring: it depends. Artificial intelligence is a powerful tool that can help people accomplish things that were simply unimaginable before, and that can help give us insights that we would have never seen. At the end of the day, however, AI is just that: a tool. As with any tool, it’s important to know how and when to use it. This much is still a very human decision to make, and so it is up to us whether artificial intelligence will serve as a superpower to save lives, or as just another way to replicate harmful patterns.