He was averaging 4 hours and 44 minutes of sleep. He’d wake up just hours after falling asleep. His brain was, quite literally, “buzzing.” And while he was at the peak of his career as a software developer, Glenn Sanford’s mind and body were saying: “enough.”
Sanford is the founder of eXp Realty, a multibillion-dollar real estate business, as well as CEO and founder of SUCCESS® Enterprises, connected to exP World Holdings. He is well-known for his innovative approach to using AI to build and streamline enterprise operations, including SUCCESS Labs™. And now, he’s ready to open up for the sake of other developers and creators, especially young entrepreneurs unaware of the addictive traits of AI, and the dopamine trap that can come with the job.
Sanford isn’t alone. In a 2025 report from Creators 4 Mental Health, North American content creators were surveyed, and researchers found this population has “significant mental health challenges that go hand in hand with their roles as content creators, but the data shows they largely lack any significant support.” Specifically, 10% have experienced suicidal ideation, 69% struggle with financial instability and 65% feel obsessive over content performance sometimes or often as a result of their roles. Overall, 62% have burnout.
In addition, a systematic review of existing studies of software engineering professionals found their well-being was linked to important work outcomes. Higher well-being among software engineers correlated with better performance, productivity and creativity and presumably with lower turnover and fewer negative outcomes.
For Sanford, the mental health aspect quickly became a physical health concern, as his body told him to slow down, and also prompted him to dig deeper into the dopamine trap of content creation.
Why creators get caught in a dopamine loop
Sanford says there was a specific high he didn’t realize he was seeking—one that would come from “just one more prompt” into his AI tool. The survey points to this as “obsessive checking of post analytics,” but for Sanford, it was a way of life starting last summer when he was tasked with the role of developing a personal development community. In August, with the release of ChatGPT-5, he says they were able to build an “entire replacement to Facebook” in about a month and were working toward building what he calls the “Netflix of personal development.” But, along the way, while trying to add features and functions, errors and issues would surface repeatedly. AI operates in such a way that it only takes a minute or two for the feedback and the fix. That process of trial and error, followed by a quick solution to build stronger features, was where the dopamine trap originated for Sanford. “The feedback loop was so rapid, I kept wanting to go back, [asking], Can I make it do that?”
He’d work on these prompts and resolutions all day late into the evening and finally started noticing that literal brain buzz. “I was waking back up thinking, I need to do this. I just need to add this feature or this function,” he recalls. “Even though I knew it wasn’t healthy, it was almost impossible to stop… just one more prompt.”
Soon more than his head was buzzing—his heart started racing, and he found himself experiencing atrial fibrillation, commonly known as “A-fib,” a serious condition that can require treatment.
Why creators are particularly at risk
In a world where AI is integrated across most industries, research is just starting to emerge about the benefits and drawbacks, including which roles are most likely to have a risk of a dopamine trap, mental health struggles and “addiction” to success.
Creators are increasingly at risk not just because of those fast feedback loops but also because it’s a field full of stiff—and quick—competition. “As a type-A competitor persona, there’s the idea of winning. Getting there before competitors do. Making it. That internal drive,” Sanford says. “If I don’t do this, somebody else will do this and take the lead.”
Tommy Wood, PhD, assistant professor of pediatrics and neuroscience at the University of Washington School of Medicine and author of The Stimulated Mind: Future-Proof Your Brain from Dementia and Stay Sharp at Any Age, says the motivation and reward cycle we experience with dopamine is powerful.
He explains dopamine is released when we anticipate a reward, helping motivate us to do the work we believe will pay off. When the reward arrives, dopamine spikes reinforce learning by signaling which actions led to the outcome, such as unexpectedly strong user engagement. But when results are hard to predict, dopamine responses are even stronger. For creators and developers, this unpredictability can fuel a cycle of constantly chasing the next success, with rising expectations after each win.
Elana Hoffman, PhD, a clinical psychologist, explains this part as “moving goalposts,” in which you constantly need more and bigger input to get the same “hit” just as you would a drug. “The danger comes when a person is driven by the dopamine and loses track of the value in the work,” she says. She is especially concerned about younger creators because they’ve grown up on screens and might already have a need for immediate gratification while struggling with delayed rewards. “If they are noticing that they are having a hard time focusing on things that don’t involve the dopamine hit of screens, if they are needing more reinforcement to feel the same impact, if they are constantly craving the dopamine hit—that might be a sign,” she says.
Keeping the ‘creativity’ in ‘creator’
AI is helping to level the playing field for creators. Tara Lynn Townes, CEO and founder of Bullying Buddy Consulting, didn’t have a technical background but learned that she too could develop an app. For her, this was empowering. “Every ounce of the app is calculated, from the design to the infrastructure to building it out. There are so many different phases. They were all amazing and intriguing, but they definitely created frustrating times,” she says. “Then there’s good old-fashioned beta testing—what they call ‘breaking the app’ in the technical world. That process by itself is taxing because you want it to work, and it’s not working. Then you uncover something else. They call them ‘bugs.’” This process led to overwhelm and burnout for her too.
She worries that developers will get stuck in the “repetitive cycle: being excited, being motivated and creating something you imagined, then seeing it come to fruition.… All human beings seek validation—especially younger ones.” Her advice to them? “That frustration in itself is real, so try to remain at baseline and keep it positive and exciting. Don’t do it as a chore—view it as something you’re creating, like an art project.”
Recognizing the risk and reward—and seeking recovery
Sanford was jolted from this cycle when he was presenting his ideas to his CEO, and a few people noticed him acting “manic.” He was talking fast and remembers “people couldn’t get a word in edgewise” as his obsession became apparent. In addition, his multiple visits to the emergency room with heart issues were wake-up calls that he had “pushed it too far.”
Once he started realizing the extent of the issue, he went through a period where he couldn’t even look at a computer without feeling his diaphragm tightening up. With the help (coincidentally) of AI, he started designing a better life for himself—one with magnesium and blue light glasses, more exercise and some walking breaks.
Wood calls this “diversifying inputs,” including getting stimulation from learning a new skill, playing a sport or spending time with friends. He also emphasizes the role of rest.
“When we’re sleep deprived or stressed, our dopamine signaling changes so that we become more focused on ‘now’ rewards rather than ‘later’ rewards. As a result, we default to behaviors that we know reliably provide us with motivation and hits of dopamine,” Wood says.
When the work still has to get done
As content creators will tell you, the work still has to get done. This is where Wood recommends regaining control through daily structure. For example, Sanford now limits his time coding to three to four hours per day, a stark contrast from the 12 to 16 hours he was putting in before.
“If your work relies on performance metrics that inherently have some variability and uncertainty, one way to maintain an element of control is to make sure your days have some structure to them. This can also help to decrease stress. Think about chunking the day into defined blocks of work with true breaks in between,” Wood says. Breaks might be as simple as watching a hilarious short, changing location or doing breathwork.
Sanford also shares that people should consider how many times in their life they have it in them to do a big build like this and what the ramifications might be. If you’ve already done it, do you really need to do it again? The physical and mental toll might not be worth it.
There’s a “certain price to be paid” for success in any field, Sanford says. “I technically ‘made it.’ It’s not like I need to do anything to ‘make it.’ But there’s always a piece of you that says, Can I do it again? Maybe you should say, Not this time.” Instead of putting himself in crisis mode again, he says his future focus will be setting up the next generation to do it in a more manageable way.
Photo by DC Studio/Shutterstock
