You've heard that artificial intelligence (AI) will soon be a major force in our society. But how transformative will it be? Will thinking machines take over the majority of human jobs, enabling most people to live lives of leisure? Will these machines cure all human diseases, resolve all human conflicts and serve us cold lemonade on hot days? Conversely, could this technology become a major threat to humanity?
By now, many people have formulated some basic assumptions about AI, and many of these are quite outlandish. I've listed a few of the more popular notions below. After providing a brief description of each, I'll explain why they're just plain wrong.
You might have heard that teams of computer scientists are capable of building software that can learn. Once created, these programs start to take in all sorts of information from the Internet and from their surroundings. In essence, they adapt cognitively, getting smarter all the time, and they keep drawing brilliant and original conclusions.
In reality, humans still shape every detail of the manner in which these purported AI systems operate. Indeed, I can attest from personal experience that this software is challenging to code and maintain. Experts must labor over these programs and continually feed them information. Without the ongoing guidance of talented human beings, they'd eventually become useless over time.
In other words, today AI isn't truly adaptable in a generic sense. No such program is close to making any cognitive leaps beyond what humans teach them to do. Despite what you may have been led to believe, when presented with a situation that it's never encountered before, AI software often won't know how to respond, and more than likely, will not even attempt a response.
Thus, even considering the current state-of-the-art, these programs can only function within a relatively narrow set of parameters.
Some of today's most sophisticated AI programs are basically chatbots and virtual assistants. They're tools that people can employ to perform specialized tasks faster, and with much greater accuracy.
For example, such software can help medical specialists by reviewing millions of pages of research material or medical records in a short period of time, extracting relevant pieces of data and using that information to make judgments and predictions. While impressive, those actions aren't fundamentally different from what computer programs have been already been doing for decades.
You might even say that using AI these days is analogous to using a common calculator. After all, calculators also let you solve certain types of problems much faster and with less likelihood of making mistakes, just as these virtual assistants do.
The very concept of Ethics in AI has recently been greatly overhyped. Sure, some people fear AI, but there have always been those who've worried about new technology. Keep in mind that, when railroad travel first emerged, many believed that riding in a train would be fatal and could cause your body to melt.
In terms of AI, it's only natural that some would be jittery about the idea of machines purportedly becoming smarter than humans ― not to mention more powerful and more numerous. What if they suddenly realized that they could band together and wipe out all of humankind? (At the very least, this scenario has made for exciting movies.)
On a smaller scale, what if some intelligent machine simply decided to make mischief once in a while? For example, what if your robot assistant at home posted your tax records or medical history online?
Before you let your fantasies carry you away, it's important to remember that machines are just machines. Morally and ethically, they're neutral. All these programs can do is take the processes that their human developers have defined and automate them. Machines have never been capable of any interest in people's lives and affairs - no matter how "knowledgeable" they may become, it's highly unlikely that they ever will.
Throughout human history, it's often been frightening to consider how people might misuse potent forces. As with fire, or the atom, AI could be used for enriching our lives or destroying them. Yes, AI will force people to reexamine their ethics, and over time, lawmakers must develop extensive policies regarding those programs. But human principles, motives and behavior should always be at the forefront of these discussions.
When it comes to artificial intelligence, I think it important to challenge our basic assumptions, especially when these assume the trappings of conventional wisdom. In the process, I hope I've been able to help others in reexamining their biases, presumptions and even anxieties regarding AI. Using this method, we can consider AI in a new light and find new ways of evaluating this innovative and helpful software.
What are your hopes and concerns for the future of AI? Drop me a line - I look forward to learning more from you.