Sid Dobrin is not interested in treating artificial intelligence as a distant future problem. For him, the more urgent issue is already here, built into the ordinary systems people use every day without thinking much about them.
Dobrin, a professor in the Department of English at the University of Florida and CEO of Flying Fish AI, has spent years working at the intersection of AI, writing, education, policy, and consumer life. His new book, The Not So Perfect Machine: What Every Consumer Needs to Know About Artificial Intelligence, is built around a simple concern: most people are talking about AI as a workplace tool, a student cheating problem, or a civilizational threat, while far fewer are asking how it quietly shapes the lives of everyday consumers.
That gap matters because AI is no longer something people only use when they open a chatbot. It has become infrastructure. Dobrin compares it less to an app and more to a utility system. People do not opt into the water system every time they turn on a faucet, and they do not think about the power grid every time they flip a light switch. AI now works in much the same way.
A person may believe they are avoiding AI by refusing to download a tool or declining to use a chatbot, but the systems are still around them. Credit card transactions, grocery inventory, mobile payments, modern cars, utility billing, online pricing, shipping logistics, and recommendation engines all rely on algorithmic processes. Even a routine purchase, like ordering a pair of shoes, may involve predictive models analyzing trends, regional demand, pricing behavior, fulfillment paths, and inventory levels before the box ever lands on a porch.
That is one of Dobrin’s central points. Consumers tend to see the final event, while the system sees thousands of decisions leading up to it.
The problem is not only that AI is hidden. It is that many of these systems are designed to remove friction. In consumer life, friction used to be another word for hesitation. It was the moment someone checked a second price, paused before buying, reconsidered an impulse purchase, or questioned whether a digital message was real. In AI-driven markets, friction often becomes the thing platforms work hardest to eliminate.
Dobrin calls this part of the ecology of efficiency. The efficiency being optimized is not always the consumer’s. It is often the seller’s, the platform’s, or the system’s. A ride-share app, for example, may use urgency signals to determine what a person is willing to pay. A shopping platform may streamline a path to checkout before a buyer has time to compare. A recommendation system may keep feeding the next item because predictability is useful to the machine.
His argument is not that AI is evil or useless. In fact, he is careful to note that these technologies can be beneficial. They can make systems faster, more responsive, and more accessible. But he wants consumers to understand that convenience has a structure, and that structure is rarely neutral.
That is why language matters so much in Dobrin’s work. He pushes back against the idea that AI is a kind of sentient brain. To him, most of what people call artificial intelligence is better understood as complex computing, a high-speed statistical system that predicts likely outcomes based on massive amounts of data. Calling it intelligent encourages people to trust it in ways they might not if it were described more plainly.
He makes a similar point with generative AI. His “Taco Bell principle” explains how platforms can create the appearance of endless variety by recombining a limited base of ingredients. Generative AI does not create from nothing. It ingests enormous amounts of existing material, then rearranges patterns into outputs that appear new. The result may be useful, but it is still a remix, not independent thought.
That distinction becomes especially important when AI gets things wrong. Dobrin is skeptical of the popular term “hallucination,” because it makes the machine sound dreamy or confused. What is really happening is a statistical failure. The system selects something plausible within its model, even if it is factually false. In low-stakes settings, that may produce a strange image or inaccurate paragraph. In higher-stakes settings, it could affect a medical record, credit score, job screening, legal file, or insurance decision.
The same concern applies to synthetic media. Voice clones, deepfakes, and targeted scams have already moved beyond novelty. Dobrin points to cases where AI-generated calls or videos exploit emotional trust, making fraud feel like family, urgency, or authority. The danger is not only that people can be fooled. It is that the systems are designed to move faster than critical thought.
His most practical advice is also the simplest: pause.
Dobrin calls it the delayed response mandate. Before sending money, sharing alarming content, reacting to a shocking image, or responding to an urgent digital request, people should create a deliberate pause. The gap between seeing and verifying is where deception thrives. Slowing down gives human judgment a chance to reenter the process.
That idea connects to what Dobrin calls cognitive sovereignty, or the right to interact with digital systems at a human pace. In daily life, that might mean turning off one-click purchasing, manually reviewing checkout screens, refusing to let AI draft every message without review, verifying urgent requests through another channel, or maintaining enough personal knowledge about health, finances, and relationships to challenge what a machine presents.
For Dobrin, the goal is not retreat. AI is too embedded for that, and too useful to dismiss outright. The goal is to become less passive inside systems that increasingly depend on speed, prediction, and compliance.
The central question he wants consumers to ask is direct: Is this system serving me, or am I serving the system?
That question may be the beginning of a more realistic AI literacy. Not fear. Not blind trust. Just enough friction to stay human.
Want more Grit Daily Startup Show? Take a look at past articles, head over to YouTube, or listen on Apple Podcasts or Spotify.
