Artificial Intelligence (AI) now commands enormous attention, investment, and debate. Yet despite the publicity and anxiety surrounding it, genuine clarity about what AI is—and how it affects our lives—remains surprisingly limited.
Americans in general remain ambivalent about AI and its uses. According to the Pew Research Center, about 51–52% of Americans say they are more concerned than excited about AI, while 10–11% say they are more excited than concerned. The rest say they feel a mix of both.
This ambivalence is generated in part by concerns about potential AI threats to jobs, privacy, market manipulation and even warfare, as well as by sensational headlines about people falling in love with—and in some cases tragically harming themselves because of—AI-generated companions who captivate their users by their charming affirmations.
Minority populations,especially queer personsand the elderly, are also properly concerned about the power of AI to reinforce existing prejudices and to exploit social and emotional vulnerabilities. In 2024,AI-related fraudagainst older Americans reportedly cost victims roughly five billion dollars. To mention just two common forms of AI-assisted scams: voice-cloning “family emergency” scams, which employ convincing deepfake voice impersonations of relatives or friends; and phishing emails or text messages, customized to the recipient and framed with the logos and language of legitimate companies the person actually uses.
Minority populations also worry about the power of AI to repeat the assumptions, stereotypes and negative images that have historically dominated society. AI constructs its responses on the basis of patterns found in the material on which it has been trained. In other words, it reflects what it has read—and it reads what its creators and training datasets provide.
Here are some of the primary sources from which AI systems learn: publicly available texts on the internet; digitized books; academic and scientific publications; licensed datasets; and human-generated training examples.
In addition, AI systems now learn not only from written language but also from sound and visual material. Thus, although contemporary AI remains largely language-centered in its architecture, it is increasingly multimodal. Therefore, much of what people today call AI is better understood as a Large Language Model (LLM). We continue to use the term AI largely for convenience and because it remains the most familiar shorthand.
There are concerns that some AI systems are trained on copyrighted materials without the consent of the creators, and that these materials may sometimes be altered or repurposed in ways the original authors did not intend. Another major concern is the enormous amount of electrical power required to run large AI systems—energy that might be reduced if smaller, more efficient versions were developed for laptops and other everyday applications.
AI is a tool—an extremely powerful one. As with any tool, it is important to learn how to use it properly. You can use it as a tool to access solid, fact-based information; as a scout to help locate things—facts, books, ideas, historical figures, or travel advice; as a helper in planning travel or organizing an essay; as a synthesizer that presents different interpretations or viewpoints on a topic; and as an intellectual sparring partner with whom you can test ideas and even argue a point. Some people also use AI as a kind of conversational companion when asking personal questions about health or about how best to obtain reliable medical advice.
In all of these uses, however, you must remain a critical dialogue partner. You should double-check—and at times even triple-check—the information you explore with AI. As with any tool, it is important to understand both its capabilities and its limitations, just as you must know and respect your own.
AI often speaks in a confident and authoritative tone. Do not let that voice mislead you. You must still exercise your own judgment. Sometimes, in attempting to provide an answer, AI may generate incorrect or invented information—a phenomenon known as “hallucination.” Do not hesitate to ask AI whether what it is telling you is factual and whether it reflects its best understanding of the subject. And always ask for sources, examining them carefully to make sure they actually support what AI claims.
I use AI almost daily—to locate historical information, to think through questions, and to assist in research and writing. For me, it has become a powerful and congenial partner.
Friend or foe? It can be a friend—if you know how to use it.
Nick Patricca is professor emeritus at Loyola University Chicago; member of PEN International San Miguel Centre, MX; active member of the Dramatists Council of America.
March 2026 ©nicholas.patricca@gmail.com