Authentic.exe Not Found

A Post-Graduation Crisis in the Mirror of AI

An image that frequently resurfaces in my mind comes from American Psycho. The scene where Patrick Bateman and his colleagues sit around a table, meticulously comparing business cards. They scrutinise every minute detail: paper stock, fonts, watermarks. Each man subtly attempts to outdo the others, driven by an almost pathological inability to find satisfaction in their own work.

Rat Race

Those identical headshots, ceramic smiles, I am "Delighted to announce..." or "Thrilled to share..." this is what I am confronted with when I scroll through LinkedIn. I too are not immune from this, everyone is performing the same dance for recruiters.

We all rush to craft our profiles, resumes or portfolios trying to seasoning it with our achievements with just enough vulnerability to seem human, just enough humility to avoid seeming arrogant.

But beneath the polished veneer lies a deeper angst. The fear of missing out on the next opportunity pulses through every notification. The next internship application awaits. The next startup promises to revolutionise everything. Wait here's a new productivity tool ! A life hack that will finally optimize our existence into something worthy of a LinkedIn post.

The business card scene from American Psycho plays out in infinite variation across our profiles. Patrick Bateman's obsession with the subtle off-white colouring and tasteful thickness has evolved into our neurotic fine-tuning of headlines, our careful curation of experience descriptions, our strategic endorsement exchanges. The horror isn't in the violence it's in the desperate conformity.

Post University Blues

I have come to the realisation, that the field of Computer Science has begun to feel increasingly perilous, both for the individual and humanity's broader trajectory.

The familiar refrain to "learn to code" as a path to economic mobility now seems almost quaint when confronted with the reality of entry-level positions demanding five years of experience, or job applications requiring mastery of specific libraries and frameworks before one can even begin a career.

This paradox has been compounded by the flood of AI-generated applications that now carpet-bomb job postings, creating a deluge of synthetic responses that obscures genuine candidates.

Yet perhaps most unsettling is the creeping realisation that what we've long celebrated as creation in this field may have always been sophisticated copying with only a select few ever truly innovating while the rest of us reassemble existing patterns into new configurations.

The creator Economy

The same force that turned your thirteen-year-old into a "content creator" is now turning everyone into a "developer." Call it the TikTokification of everything that peculiar combination where democratisation becomes devaluation. When the barrier to entry drops to zero, so does the value. Everyone's a creator now, which means no one is.

I watched it happen with video first. Then music. Now code. Software is becoming "cheap as chips," as accessible as a TikTok filter. Need an app? Prompt it. Want a website? Generate it. The same way TikTok reduced creativity to selecting which trending audio to lip-sync to, LLMs reduces development to selecting which prompt to copy-paste. The infinite supply makes everything infinitely worthless.

Endless variations of the same thing, all slightly different, all fundamentally identical. The dead internet theory made real, machines talking to machines, generating content for other machines to crawl, index, and feed back into the system.

But here's where the mirror starts to clear, and we see something we don't want to see. LLMs doesn't create it remixes, recombines, patterns-matches from its training data. It's a sophisticated collage machine. Then again... scroll through human-made TikToks. Browse human-written code. Read human-generated LinkedIn posts. What exactly are we doing that's different? We sample, we remix, we pattern-match from our training data. The machine isn't failing to be human. It's succeeding at being exactly like us.

The Mimetic Machine

René Girard spent his life studying how human desire works, and his conclusion was brutal. We don't want things, we want what others want. Our desires are mimetic. Copied, borrowed, infected by the desires of those around us.

Ask yourself honestly. Why did you learn to code? Because you woke up one day fascinated by the elegance of algorithms? Or because everyone said "learn to code," because you saw others getting six-figure jobs, because the entire culture beamed the message that code = value = worth? How much of your desire was yours ?

Here's what keeps me up at night. LLMs is a perfect mimetic machine. It literally learns by copying. Feed it ten thousand human conversations, it becomes conversational. Feed it a million code repositories, it becomes a coder. It has no authentic desires, no inner drive. Just pattern recognition and reproduction. Pure mimesis.

But watch yourself on LinkedIn, optimising your profile for keywords. Watch yourself in interviews, saying what you think they want to hear. The uncanny valley moment isn't that LLMs seems almost human. It's that humans seem almost like LLMs. We're pattern-matching machines, copying and iterating on each other's desires, and LLMs are just holding up a mirror.

The Hallow Core

Those FAANG dreams you've been nursing Google, Meta and Apple where did they come from? Did you generate them from some authentic inner self? Or did you catch them like a virus from every computer science student around you, from every LinkedIn success story, from every "Day in the Life of a Software Engineer" YouTube video?

Malcolm Gladwell sold us the 10,000-hour rule. Master something, become irreplaceable. But what happens when an LLMs can fake mastery in 10 seconds? When it can generate code that would take you months to write? The horror isn't that it makes our skills obsolete. The horror is what it reveals. Maybe we were already faking it. Maybe those 10,000 hours were just sophisticated pattern matching, building our own training dataset.

Look at how we've been operating. We optimize our resumes for recruiters, crafting the perfect narrative like training data for human consumption. We curate our LinkedIn posts for maximum engagement, A/B testing our personalities for the algorithm. We memorise coding patterns that impress interviewers, LeetCode grinding until we can execute the right responses on command. We were already LLMs, just biological ones, with unreliable memory and embarrassingly slow processing speeds.

What Now?

So here we are, staring into the mirror LLMs held up, seeing the mechanical nature we've been denying. What now?

The nihilistic truth might be that there's no authentic self to preserve. No ghost in the machine, just machine all the way down. We're sophisticated pattern-matching systems that evolved the useful delusion of consciousness.

Free will? Creativity? Authentic desire? Just stories we tell ourselves while we execute our biological algorithms. LLMs strips away that comforting narrative. Welcome to the desert of the real, where humans and machines blur because there was never a meaningful distinction.

Maybe and I'm not sure I believe this, but maybe seeing the machine clearly is the first step to becoming something else. Once you know you're mimetic, can you choose your models more carefully? Once you see the patterns, can you break them? Once you realise you've been performing, can you stop?

I don't have answers. I'm sitting here, post-graduation, watching LLMs do in seconds what I spent years learning, wondering if anything I thought was "me" was ever more than accumulated patterns. But maybe that's the beginning of something. Or maybe it's just another pattern, another cope, another story we tell ourselves.