The Case for Being Slightly Delusional

Two Bets I Almost Did Not Make

When I signed up for Shreyas Doshi’s Product Sense course, the rational voice was loud: I was an internal technical product manager for data platforms at Expedia. I was not building user-facing products. The course was designed for people shipping consumer experiences, not someone managing metadata pipelines. Every reasonable assessment said this was not for me.

I signed up anyway. Let me see what happens.

The eight-day course did not teach me how to build consumer products. It did something I did not expect: it rewired how I see every product and service I use. It built a muscle for what good looks like. I started noticing design choices in apps I had used for years without thinking. I started asking different questions in roadmap discussions. One course, taken on a slight delusion that it might apply to me, had a lasting impact on how I perceive the world. No rational cost-benefit analysis would have predicted that return.

That experience taught me something about the value of ignoring the rational voice. A year later, I ignored it again.

When I decided to leave Expedia for Capgemini, the doubts were louder this time. Product company to consulting company. Focused depth to broad client work. Every career advice thread on the internet says do not go from product to consulting. The rational voice had a long list of reasons to stay.

I talked to a close group of people I trust. Their feedback gave me enough confidence to take the jump. Since joining Capgemini, I have been constantly learning and growing in ways the previous role could not have offered. I have worked on data agent architectures, AI Governance from a practitioner lens across multiple industries, and the messy reality of what works and what does not work when you deploy these systems at enterprise scale. The move expanded my surface area in ways that staying in one product company, focused on one internal platform, never would have.

Both bets were slightly delusional. Both paid off in ways I could not have predicted. And both followed the same pattern: the rational voice said no, a quieter voice said “let me see what happens,” and the outcome exceeded what either voice anticipated.

That pattern has a name. Psychologists have been studying it for nearly fifty years.

A 1977 Psychology Paper Goes Viral (Without Credit)

Somewhere around late 2022, millions of people on TikTok started telling each other to be delusional. The phrase “delulu is the solulu” originated in K-pop fandom circles around 2013-2014, where it mocked fans who genuinely believed they would date their favorite idol. By 2023, TikTokers had repurposed it into something broader: a philosophy of radical self-belief. Instead of tempering your expectations, embrace unrealistic goals. Act as if the outcome is already yours. By late 2023, the hashtag #delulu had crossed 4.3 billion views, later surpassing 5 billion by 2025.

The most visible example was Sabrina Bahsoon, known as “Tube Girl,” who filmed herself dancing unapologetically on the London Underground in 2023. She behaved as if she were performing for an audience of millions. Then, because the internet works the way it does, she actually was.

On the surface, this looks like motivational nonsense wrapped in Gen Z slang. But the reason it resonated with tens of millions of people is that it accidentally rediscovered findings that psychologists have been documenting for nearly fifty years. Nobody scrolling through TikTok knows the citations. The citations exist anyway.

The Science Behind the Meme

In 1977, Albert Bandura published “Self-Efficacy: Toward a Unifying Theory of Behavioral Change” in Psychological Review. His central finding: self-efficacy is one of the strongest psychological predictors of persistence and behavior change, alongside skill, resources, and context. Bandura identified it as “a highly accurate predictor of degree of behavioral change,” with efficacy expectations shaped by four sources: performance accomplishments, vicarious experience, verbal persuasion, and physiological states. The first one matters most. Actually doing hard things builds the belief that you can do more hard things.

A decade later, Shelley Taylor and Jonathon Brown went further. Their 1988 paper “Illusion and Well-Being” in Psychological Bulletin argued that mentally healthy people systematically hold three forms of positive illusion: they overestimate their own abilities, they believe they have more control over events than they actually do, and they hold unrealistically optimistic views of the future. The controversial part: these illusions did not correlate with worse outcomes. They correlated with better mental health, greater resilience, and higher productivity. Taylor and Brown linked positive illusions to “the ability to care about others, the ability to be happy or contented, and the ability to engage in productive and creative work.” The “positive illusions” thesis is influential, but the size and conditions of its benefits are still debated in later literature.

Tali Sharot’s neuroscience research at University College London added the biological layer. Many studies find that a majority of people, roughly 80%, exhibit an optimism bias, consistently overestimating the likelihood of good events and underestimating the likelihood of bad ones. As Sharot has noted in interviews, this pattern holds broadly across populations. It is not a flaw in human cognition. It appears to be a feature, one that persists because it provides a functional advantage.

Psychotherapist Alison McKleroy connected the dots explicitly when she described the delulu trend as “almost like a self-efficacy tool.” The meme is doing what Bandura described in 1977. It is just doing it in 15-second videos instead of peer-reviewed journals.

Where the Meme Becomes Dangerous

TikTok optimizes for clean narratives. Reality is not clean.

The Winners You See and the Thousands You Do Not

Survivorship bias is the obvious gap. We see the people for whom delusion paid off. We never see the thousands who projected identical confidence and got nothing back. BLS establishment survival data implies roughly one in five businesses close within a year, about half within five years, and around two-thirds within ten (exact figures vary by cohort, and these track establishments, not individual entrepreneurs). In VC-backed samples, first-time founders have roughly a 21-22% success rate, where “success” is narrowly defined as reaching an IPO. For every delusional founder who became a case study, there are dozens who just became a statistic. The meme celebrates the winners and pretends the losers do not exist.

The privilege dimension is invisible but real. “Just be delusional about your goals” is different advice depending on who is hearing it. Someone with a financial safety net can afford to bet on themselves for years and be wrong. A first-generation professional carrying student loan debt cannot absorb the same level of risk. Telling everyone to “just believe” without acknowledging the material conditions that make belief safe or dangerous is advice that works selectively, then claims to be universal.

The Slide from Confidence to Fantasy

The slope toward magical thinking is shorter than it should be. The distance between “believe you can do hard things” and “the universe will manifest your desires if you think positively enough” collapses quickly. Rhonda Byrne’s The Secret sold over 34 million copies (publisher-reported) on essentially this promise. There is no credible evidence that thinking about outcomes causes them to materialize. Belief without execution is not confidence. It is fantasy with a vision board.

And there is an irony worth noting. The Dunning-Kruger effect, first documented in 1999, found that people scoring in the bottom quartile of competence (12th percentile on average) estimated themselves at the 62nd percentile. A 50-point gap between belief and reality. The mechanism: lacking skill also means lacking the metacognitive ability to recognize you lack skill. Delusion, in this framing, is not a strategy. It is a symptom of not knowing enough to be appropriately humble. Though even this finding is now contested. Critics, including Gignac and Zajenkowski (2020) and a McGill University analysis, have argued the effect may be largely a statistical artifact of how self-assessment data is analyzed. That does not mean miscalibration never happens; it means the original framing overstated how cleanly the data supports it.

What Delusion Actually Does When It Works

Strip away the meme language and the motivational aesthetics, and there is a mechanism here worth understanding.

The world systematically rewards people who act with more confidence than their evidence strictly supports. This is observable in hiring (confident candidates get offers over equally qualified but hesitant ones), in fundraising (investors fund conviction as much as competence), in leadership (teams follow people who project certainty, even when that certainty is partly performed), and in negotiation (the first person to name a number anchors the entire conversation).

This is not mysticism. It is incentive structure. Confidence creates opportunities, opportunities create experience, and experience eventually builds the competence that was missing at the start. The loop is real. The question is whether you enter it through genuine belief, strategic performance, or blind overestimation, because those three paths lead to very different places.

The version that works: projecting confidence about the destination while being ruthlessly honest about the path. Believing the outcome is achievable, then doing the actual work to close the gap between belief and ability. The confidence gets you in the room. The discipline keeps you there.

The version that fails: believing the outcome is inevitable regardless of effort. Confusing the feeling of certainty with the presence of competence. That is not strategic overconfidence. It is denial wearing ambition’s clothes.

Why This Generation Specifically

There is context for why “be delusional” resonated with Gen Z and not with previous generations at the same age. This is a generation where 27% report their mental health as fair or poor, compared to 15% of millennials and 13% of Gen X. Only 47% consider themselves “thriving,” versus 59% of millennials. Anxiety and depression are among their most commonly reported diagnoses.

For a generation saturated in self-doubt, economic uncertainty, and algorithmic comparison, “be delusional” is an overcorrection that at least points in a useful direction. It is a crude tool for a real problem. After a decade of hustle culture telling them to grind harder, followed by a pandemic that dismantled the premise that grinding leads anywhere predictable, a reframe from “work harder” to “believe bigger” feels psychologically refreshing. Even if it is equally incomplete.

TikTok’s algorithm amplifies the incompleteness. “Be delusional” fits in a caption. “Develop calibrated self-efficacy through evidence-based goal setting and iterative competence building” does not. The medium selects for the oversimplification.

The Reframe Worth Keeping

I have been on the other side of this equation. When I joined the Kellogg Executive Education Chief Product Officer track, the first virtual meet and greet made me question whether I belonged. The cohort was full of VPs, Senior Directors, and Heads of Product joining from across geographies. I was a Senior Product Manager at Expedia. Two or three levels below nearly everyone in the room. But as the program progressed, something shifted. My combination of technical depth and product instinct, go-to-market thinking, jobs-to-be-done frameworks, building with user empathy, turned out to be a perspective the room was missing. By the time we hit the business strategy simulation, I was contributing at a level that had nothing to do with title. The belief that I could belong arrived before the evidence did. The evidence followed.

The opposite of delusion is not realism. It is paralysis. More careers stall from excessive caution than from excessive confidence. The person who waits until they feel fully qualified before applying. The one who runs another analysis instead of proposing a direction. The professional who hedges every statement until nobody in the room trusts them.

Bandura’s research is clear on this: belief precedes competence more often than it follows it. You frequently need to believe you can do something before you have evidence that you can. That belief is not a substitute for the work. It is a precondition for starting it.

The TikTok trend stumbled onto something real. Self-efficacy research supports it. Positive illusion theory supports it. The optimism bias appears to be functional, not pathological. Where the trend fails is where all memes fail: at nuance. It skips the part where belief has to be converted into skill through sustained effort. It ignores who can afford to be wrong. It conflates confidence with competence, when the relationship between the two is sequential, not interchangeable.

The useful version of “delulu is the solulu” is not “believe hard enough and reality will comply.” It is closer to: believe before the evidence arrives, then build the evidence before the belief expires.

That is a less catchy tagline. It is also the one that actually works.