AI: Beyond the Headlines

“Artificial intelligence could lead to extinction, experts warn.”
“300 million jobs will be lost or degraded by artificial intelligence.”
“Why the AI revolution may wind up killing capitalism.”

These are actual headlines that have appeared in the popular press over the past couple of years. I’m sure they generated plenty of clicks for their publications, but I also doubt these articles did much to educate readers about the true potential of AI. Understanding and effectively leveraging this emerging technology is going to be critical to business success in the coming years, but leaders should watch out for these common misconceptions:

Misconception #1: AI is All LLMs

While large language models (LLMs) like ChatGPT tend to dominate headlines, there is a wide world beyond generative AI tools. Consider how Netflix personalizes your movie recommendations, or how Amazon predicts your next purchase; these are examples of AI-powered recommendation engines. Other forms of AI are fueling pharmaceutical research and product design. Often, businesses will find their competitive advantage not in the form of headline-grabbing chatbots, but rather from focused applications that solve specific business challenges.

Misconception #2: Every AI Capability Adds Value

When people discuss using AI to assist with new drug discovery, they often focus on the technology’s ability to rapidly generate large numbers of new drug candidates. But if you actually talk to experts in the field, they’ll tell you that there are already more than enough potential new drugs. The bottleneck comes in testing and validating these drugs, which is something that AI isn’t very helpful for (at least, not yet).

Misconception #3: ‘Intelligence’ Always Means the Same Thing

When I teach my “Product Innovation in the Age of AI” class at MIT Professional Education, I start by asking my students to define intelligence. Even though these are highly skilled mid-career professionals, they typically struggle. That’s because intelligence really isn’t just one quality, but rather a set of behaviors and capabilities. When we break the term down into its components (including deductive, inductive, and abductive reasoning) we’re better able to see what specific abilities AI tools need to tackle a given task.

Misconception #4: Artificial General Intelligence (AGI) is Coming

No doubt, you’ve seen at least a few headlines warning about the “dangers” of AI tools surpassing human intelligence, becoming self-aware, and turning on their creators. You can stop worrying. Scientists don’t even have a unified theory of human intelligence yet, let alone the ability to replicate this intelligence in computer form. If AGI is even possible, it’s centuries away from becoming a reality.

Misconception #5: AI is a Job Killer

Sure, some jobs will become obsolete as AI tools improve, but the idea that these jobs will be permanently “lost” assumes that we live in a static world, which we don’t. We live in a world with a highly dynamic economy, and jobs constantly evolve in response to new technologies. In fact, in the nine years since a 2013 study claimed that AI would “destroy” 47% of jobs, the U.S. economy actually added 16 million of them.

Misconception #6: AI is Useless

While some incorrectly claim that AI can essentially do everything, others overshoot in the other direction, claiming that AI can do practically nothing. These statements are often made by people who’ve tooled around with ChatGPT for a few hours, discovered that it can’t write an Oscar-winning screenplay, and decided this means the tool doesn’t have any value at all. In fact, many creative professionals already are using AI tools in their jobs to great effect.

Misconception #7: AI Will Make Us Smarter

Contrary to what some people seem to believe, AI tools won’t make humans smarter. However, they will make us more knowledgeable, more quickly. Already, I use AI tools to rapidly expand my circle of knowledge. But here’s the trick: This actually works best for subjects that I already know quite a bit about. If I ask a generative AI tool for an overview on a new topic in my field, I can cross-reference this new information with what I already know to ensure that it is credible (rather than the result of an AI “hallucination”). This means, essentially, that you already need to be “smart” at something to take it to the next level with AI tools. But when you combine existing human expertise with the emerging capabilities of AI, that’s a recipe for game-changing applications.

As the pace of innovation in AI accelerates, the ability to distinguish hype from reality becomes increasingly vital. Professionals must invest in continuous education to understand AI’s potential, avoid pitfalls, and harness its benefits. Lifelong learning has never been more critical than in the age of AI. With the knowledge and tools gained through professional education, individuals can confidently navigate this transformative landscape and position themselves—and their organizations—for success.

Erdin Beshimov is a lecturer at the Massachusetts Institute of Technology, a Senior Director of Experiential Learning at MIT Open Learning, and an instructor of the MIT Professional Education course Product Innovation in the Age of AI. He is devoted to helping learners from across the globe study innovation, technology, and entrepreneurship, founding MIT Bootcamps and the MITxMicroMasters program, and co-developing MIT’s open online courses in entrepreneurship.

Before joining MIT, he served as Principal at Flagship Pioneering, where he focused on water, energy, and materials ventures, and is a co-founder of Ubiquitous Energy—a solar technologies spinout from MIT.