There’s that moment—right before you leap off the edge—when the world holds its breath.
You’ve run through this scenario in your mind a thousand times. You’ve considered the aftermath, the noise, the chaos. And still, when every instinct cautioned you to stop, you jumped.
That’s exactly what Roy did.
He didn’t tinker in the shadows or pitch a safe startup idea to a room of polite VCs. No. He built a tool designed to cheat on software engineering interviews, live-streamed himself using it during an Amazon interview, landed the job, and then got kicked out of Columbia University.
And then? He raised $5.3 million.
Now, before you roll your eyes and mutter, “Another edgy tech bro,” hear me out. Because Roy’s actions aren’t just controversy for clicks. Buried beneath the chaos is a radical, almost poetic vision: What if AI is meant to amplify us to the point that the very idea of “cheating” becomes obsolete?
I’ve been writing about tech for over a decade, and I’ve seen trends come and go—blockchain booms, metaverse busts, and crypto winters. But this? This feels different. Not because of the funding or the drama, but because of the timing.
We’re on the brink of a cognitive shift so deep it may change the way we think, work, and perhaps, be human.
We’re on the brink of a cognitive shift so deep it may change the way we think, work, and perhaps, be human.
Let's talk about Roy. Not as a cautionary tale, but as a reflection.
The LeetCode grind. A ritual of modern suffering.
Picture this. You’re a brilliant coder. You’ve built apps, you’ve contributed to open source projects, you may have even shipped a startup. But it doesn’t matter when you walk into a big tech interview. You’re handed a whiteboard and asked to reverse a binary tree using only recursion. From memory.
Welcome to the LeetCode gauntlet.
For years, software engineers have been forced to memorize hundreds, sometimes thousands, of algorithmic riddles. Not because these problems reflect real-world coding. They almost never do. But because it has become a bizarre cultural ritual. A proxy for “intelligence.” A gatekeeping mechanism disguised as meritocracy.
Roy saw it for what it was — theater.
“I spent hundreds of hours grinding LeetCode,” he told me, voice tinged with the exhaustion of someone who has been through the machine. “Not because I cared. Not because it made me a better engineer. But because I’m competitive. If there’s a leaderboard, I want to be on top.”
Sound familiar? That’s the trap.
The system doesn’t reward skill—it rewards compliance. Thousands of hours wasted. Hours that could have been spent productively making useful things.
Roy stated, "I realized LeetCode has as much to do with real software engineering as asking how many ping-pong balls fit in a Boeing 747. It’s theater. And theater is expensive.”
So he built a backstage pass.
Using a translucent overlay that lived on your screen—watching your webcam, listening to the audio, analyzing your IDE in real time—Interview Coder could whisper answers during technical interviews. Not through copy-paste, but through seamless, context-aware AI assistance. You’d ask a question out loud, and the AI would respond—quietly, invisibly—like a genius friend sitting just out of frame.
Roy showed it. He recorded himself using it during an Amazon interview. He got the offer. He posted the video online with the caption: “This is how easy it is to game the system.”
Cue the outrage.
Amazon revoked his offer. Columbia launched an investigation. His co-founders panicked. For a month, the video barely scraped 15,000 views, while the world was in uproar.
Roy reflects on whether he has made a foolish mistake that will end up costing him for the rest of his life.
“I remember how much I sacrificed,” he reminisced. “I have sacrificed my education, career , my safety net, and for what, a few thousand views?”
But all of suddenly, it went supernova.
Tech Twitter exploded. Engineers celebrated, recruiters were angered, and academia receded. And all of the sudden, Roy wasn’t just the kid that was breaking the rules, he was the symbol.
“All press is good press,” he commented. “Even the hate drives attention, and attention is the oxygen for a startup.”
From “Cheating” to Cluely: The Bigger Vision
Here is where most people miss the point, Roy just didn’t want to help people cheat.
That was the hook, the spark. The real vision, a world where AI is not something to use, but something to think with.
Interview Coder was the first proof of concept for what he calls “ambient intelligence” an AI that is fully contextual, one that doesn’t live in a chat window. One that is aware of your screen, your voice, your context, and is ready to understand without prompts and commands.
That’s Cluely.
“The phrase ‘cheat on everything’ is intentionally ambiguous,” Roy explained. “Because once AI is integrated into every task, writing emails, running meetings, analyzing data, the word ‘cheat’ will cease to make any sense.
"Do you still remember the time when calling a calculator a cheating device made sense? ”
It is the same “death of literacy” when spellcheck software was invented. Now when spellcheck software is used, it is normal. We have offloaded spell checking to machines and have shifted our focus to meaning. Cluely wants to do the same at the level of cognition.
Picture a scientist at a lab working with Cluely. The AI listens and cross references every paper ever published on the subject, identifies a flaw in the experimental design, and whispers a suggestion in real time. Or imagine a journalist interviewing a source, while Cluely summarizes decades of that person’s public statements into a one-line insight.
It is the next logical step and that is what Roy is trying to build before Big Tech realizes it.
The Unexpected UX
Most AI tools are still reactive. You enter a command, wait, and get a response. Cluely transforms that. It is proactive, watches, listens, and anticipates."The biggest challenge wasn't the tech," Roy said, "It was taste. How do you make an AI overlay feel seamless? Not like a robot is watching you, but like an extension of your own mind?"
After designing many prototypes, he finally settled on the translucent overlay design--discreet enough to overlook, but dependable enough to use. And the performance? They are significantly reducing latency by locally hosting models, caching inputs, and behavior-based fine-tuning on the responses.
Cluely will soon be able to do much more than just respond to your questions. It will learn your tone and rhythm, and it will figure out your professional persona. It will recognize that you draft emails like David Sedaris, but you debug code like Linus Torvalds, and it will adapt to your style as necessary.
“This data becomes our moat,” Roy said. “Not just what you do, but how you think. No big tech company can replicate that without years of intimate user interaction.”
Most Disarming of All Is His Humility
For Roy, life has not been as complicated as it could be. He has had loving parents and a supportive home, and no real trauma. Even with the viral stunts, the funding rounds, and the Ivy League expulsion, he insisted that his life has been “easy.”
“It’s nothing. Kids are starving in Uganda,” he said plainly. “I got kicked out of Columbia. Big deal. I was going to drop out anyway.”
This isn’t bravado. It’s perspective. He believes we are living in the most opportunity-rich moment in human history.
If you’re in America, with access to the internet, and not in poverty, you are in the global elite. "You can build a billion-dollar company at 19, you just have to try," he said.
His advice? Broaden your horizons.
Not wildly thrust yourself into the unknown, but take one step at a time and stretch your limits. Start with the basics and build something great. A lot of people will not notice your work at the start, but that’s okay. Show it to them. With every new challenge you take unimaginably larger goals will start to feel like a walk in the park.
“I wasn’t this person six months ago," he said. "I wanted a FAANG job. A safe path. But once I took that first real risk—posting the Amazon video—everything changed. I realized the downside was tiny, and the upside? Infinite."
The Future Isn’t About Jobs—It’s About Thought.
Here’s the uncomfortable truth Roy forces us to face: If AI can do your job interview question, your job is already obsolete.
Not today. Maybe not next year. But soon. Instead of embracing it, we should lean in. “Why do we even need interviews?” he asked. “If an AI already knows every line of code I've ever written, every project I've shipped, every mistake I've learned from—why spend an hour pretending to assess me? Just match me to the role.
“Spend half an hour checking if we match.”
He envisions the future with AI taking the ‘what,’ while people focus on ‘why’. Scientists can expedite the process through discoveries. Artists shift to finding the more profound intricacies. Teachers go to mentoring instead of lecturing.
Roy exclaimed, “Just picture the potential if every researcher harnessed AI to multiply their efforts 100-fold. Cancer would be cured a decade sooner. Mars would be within reach in two years. The rate of human advancement would not just accelerate, it would explode.”
But all of this can only be possible if we stop seeing AI as a danger and start viewing it as a cognitive collaborator.
Be Remembered. Not Liked.
Roy is not concerned with earning your affection.
He desires to be remembered. Remembered like Steve Jobs, like Elon Musk. People who ignited discourse, shattered conventions and pushed society to adapt.
Roy said, “We all die. No one will reminisce your GitHub commits. No one will remember your LinkedIn endorsements in a thousand years. But if you shaped the way in which people think? That is immortal.”
He would stay polarizing. He would continue to post provocative tweets and push through the boundaries. In his mind, the online drama is just noise, and the only voices that truly count are the ones in real life: his parents, his future wife, and his unconditional lovers.
Everything else? Just fuel for the mission.
Also Read: The AI Bubble Was Never About AI: The Human Psychology Behind Tech's Greatest Illusion
Final Thought: The Leap Is the Lesson
Roy’s story isn’t about cheating. It’s about courage.
The courage to look at a broken system and say, “This is stupid,” then build a better one—even if it costs you your diploma.
We don’t need more people playing it safe. We need more people willing to jump—even when the net isn’t there.
Because sometimes, the fall is what teaches you how to fly.
And if Roy’s right? We’re all about to soar.
0 Comments