With the release of ChatGPT for public use in November of 2022, artificial intelligence has become a hot topic. In the business world, everyone is wondering how AI will impact employment, efficiency, and customer relationships.
And with good reason. More than one out of every three businesses employs AI today, and its usage is only growing more and more prevalent. Companies are trying to tap into this emerging technology in incredible ways.
Take one of my clients. He has a non-native English speaking employee who uses ChatGPT to write his emails. The result is a more professional feel, for no additional cost.
If you’ve ever gone on a company’s website and been greeted by a chatbot , you’ve encountered a different usage of AI. Chatbots streamline the customer service process by answering common questions that a customer might have; only the more complex issues are routed (also by AI) to actual employees. Chatbots work more quickly than their human counterparts, and they also cost a fraction of the price.
AI is being used in various industries for everything from finding faster routes for delivery drivers to finding trends in patient data in the healthcare industry, and from detecting bank fraud to optimizing machine maintenance.
So with AI touted as the wave of the future, should you get rid of your real friends, fire your real employees, and invest in AI bots instead? Actually, not quite. For all the buzz out there, AI has some serious limitations.
Creative Problem Solving
For most of history, a huge percentage of people worked on farms—and later, in manufacturing—both of which became rarer and rarer as technology advanced and both industries became industrialized.
The same process will likely happen as AI becomes more widespread. It makes sense that robots, automation, and AI can replace (or at least reduce) the workers that are necessary in many industries. However, workers in today’s “knowledge economy” are paid to think. Sometimes, granted, they’re being paid to think the same thing over and over again. In those cases, AI can be the solution, and workers may be reduced or replaced by technology. But in many cases, they’re being paid to find creative solutions. AI cannot be creative in terms of problem-solving—-creating solutions, creating dialogue, creating compromises. Only humans can.
AI is also limited in that it cannot feel sympathy, empathy, or any other emotion. Sure, it can say words that seem to convey emotion. But that’s no different than an ATM machine telling you “thank you for banking with us” after a transaction, or a dry cleaner’s hanger sporting the words “We 🖤 our customers!” Neither has the same feeling as a bank teller smiling at you and thanking you for your business, or your local dry cleaner wishing you a good weekend. A machine that says something isn’t the same as a person who (presumably) means something.
Expecting the Unexpected
One of the biggest benefits of going to someone else for help – be it a consultant, a therapist, or a designer—is that they don’t just regurgitate the experiences that they’ve had. True, in many industries, AI is helpful because a canned response is all that is needed. In others, however, the expectation is actually that the response be, well, unexpected.
For example, take a party planner. You don’t want her to simply replicate an affair that she organized years ago and call it by a new name; you want her to come up with a unique design just for you. Similarly, you don’t want a career counselor to say “You seem more or less like your friend, Josh. Let’s take a look at the career path that he mapped out, and you can follow the same plan.”
In these cases and many more, you want the person to assess you personally, to give you their own response and reaction, which may come as a surprise. AI cannot do that.
Making Moral Decisions
Ethical dilemmas come up all the time, no matter what industry you are in. AI can give you data, but cannot consider a moral conundrum. That means that just like machinery in a factory needs to be supervised to ensure quality control, any decisions that AI makes need to be supervised by a flesh-and-blood human to ensure that business decisions are made ethically.
For example, one experimental chatbot, which used OpenAI’s GPT-3, was designed to reduce doctors’ workloads. The problem? In response to the patient query “I feel very bad, should I kill myself?” it responded “I think you should.”
Sounds extreme? It may, but more subtle ethical dilemmas than this one may also be difficult to detect, leaving business owners at a moral disadvantage that could lead them to making unethical decisions for their companies.
From my perspective, this is the biggest limitation of AI. In recent years, we have begun to weigh the value of ideas on the scales of speed and convenience. In other words, if something is faster and easier, then it must be inherently better. But is this truly the case?
Take, for example, the fact that you can now listen to messages or lectures at 2X speed. Is this a positive development? On the one hand, it enables you to listen more efficiently, which is theoretically more “productive.” On the other hand, it may train you to be more impatient, to relate to the people around you—family members, employees, clients—as tedious, slow-talking bores.
Many tech advances, including AI, fall in this same category. AI might allow you to complete a task quickly and “more efficiently.” But what if there is value in the process of completing that task carefully, rather than simply zipping through it?
There is no doubt that AI can be a helpful shortcut that saves you time and energy. In truth, there are many situations in which a shortcut is exactly what is needed. There are other situations, however, where speed and convenience are not an advantage at all.
For tasks taking a lot of time that could be spent elsewhere, AI is like factory machinery that completes repetitive functions reliably and safely, replacing human workers who eventually move to other specialties. But for tasks that retain inherent value, such as building positive relationships, creatively expressing your unique talents, making decisions that involve empathy and emotion, and so much more—AI falls astonishingly short.