The Employable Lie

The job market asks for storytelling — then punishes the nonlinearity that makes a story real.

Images: AI-generated from my own concept, AI-retouched to my own taste.

I did not arrive here by accident

Before I talk about the market, I have to tell a trajectory — because the whole argument that follows depends on it.

I was a geek as a child. Consoles, arcades, tournaments. I knew the internet when “going on the internet” still meant going somewhere: web bars, forums, back when almost no one had a connection at home. I poked around the BIOS of the all-in-one PC in my mother’s office. Then I drew in MacPaint on a Macintosh, then I modded my games, my skins, I touched the textures — I was tinkering with technology before I knew that’s what it was.

In 2004, as a student, I wrote one of my first papers on the evolution of the web: Amazon, eBay, Google, sponsored links, visibility that was about to become buyable and distort what we’d come to call “relevant.” In 2006, Facebook was still a university network — I’m on it under an absurd doctored name, before the masses arrived. I reread that paper two years ago. The intuitions were right. The drifts I described as possible have become the present.

I was talking about big data in 2008, in rooms where, in France, almost no one knew what to do with the word. It’s also the year I first traded notes on machine learning with freelance engineers. I set up a development shop — and people would shoot back “Ah, a comms agency!”: Facebook apps, WordPress templates, Twitter contests, HTML5, Canvas animation — for major brands, from luxury to entertainment. At the time, these weren’t really jobs. You still learned them by inventing them. When I saw that Flash wasn’t supported on the iPhone 3G, I understood immediately: Flash wasn’t limited, it was finished. A telecom group had spotted me for mobile gaming before a major brand even did.

Then Tech. It was the year of the iPhone 3GS; I watched the iPhone 4, Siri, the cloud, services, apps come into being — and the transformation of a great hardware-and-software brand into one of the pioneers of the modern customer experience. From the ground and the design side, I became a trainer, and I landed in the then-nascent CX — at that time, an eminently American vision. It wasn’t an exit from tech. It was its other face — the human interface of technology.

The line is not “job A → job B → job C.” The line is: discovery → creativity → design → builder → technology → interface → usage → experience → relationship → system → intelligence → the living.

My résumé may look nonlinear. My trajectory is not. The coherence isn’t in the job titles. It’s in the systems I never stopped learning to read.

I say all this for one reason: this path looks illegible only if you ignore the history it runs through — and it asks for a minimum of investment, beyond the titles.

And let me say it right away, because what follows will bite: I’m not saying recruiters are bad. I’m saying an impoverished chain of reading produces, mechanically, absurd errors. The problem isn’t a person. It’s a system.

The market asks for the future, then filters with the past

Job ads demand, almost word for word, what my profile contains (and maybe yours too): creativity, proactivity, autonomy, adaptability, decision-making, entrepreneurial spirit, the ability to connect subjects, to navigate AI, complexity, uncertainty.

Then, at the screening stage, the system looks for the opposite: linearity, immediate legibility, lexical conformity, similarity to the job description, proof of strictly identical execution, an easily classifiable degree, absence of risk, decorative but never disruptive originality.

The market says it’s looking for the future, but it reads with the grammar of the past. It claims to seek out-of-the-box profiles, then penalizes everything that doesn’t fit the box. At bottom, it doesn’t always detect talent. It detects what’s easiest to explain.

I did not become more competent. I became more readable.

Here’s the experience that made everything clear. While I was applying, no one looked at my profile. Then I started publishing — clarifying my signal, making my thinking visible, accumulating public proof. And suddenly, recruiters came — without my applying.

I hadn’t changed. Same story. Same skills. Same experience. Same nonlinearity. Same substance. What had changed wasn’t the value. It was the signal.

So the question becomes uncomfortable. If a profile only becomes interesting after being publicly packaged, validated, read, commented on, made visible, then maybe recruitment doesn’t detect value. Maybe it detects the social proof of value — after someone else has already done the work of reading.

The false storytelling

Recruitment talks a lot about “storytelling.” But what it asks for, in practice, isn’t narrative. It’s formatting.

A real story contains detours, ruptures, decisions made under uncertainty, intuitions that only become legible after the fact. HR’s false storytelling, by contrast, demands a straight line, perfect coherence, an impeccable retrospective causality: a single “through-line.”

The market doesn’t reward the story. It rewards the formatting we mistake for a story.

A career without rupture isn’t necessarily coherent. It’s simply easier to analyze. A story with no rupture, often, is not a story. It’s a brochure.

The normative trap of recruitment

In my obsession with living systems, one idea keeps returning: an individual can be normal for its category and abnormal for itself. Recruitment makes the symmetrical error to the one I study. It compares a person to the average shape of a role — the average title, the average path, the average number of years — then calls that “fit.” But “fit” isn’t a truth. It’s a reference class.

The problem wasn’t that I was “scattered.” The problem was that my coherence wasn’t linear. I wasn’t illegible. I was read with the wrong grammar.

We ask the future to prove it has already existed

An ad for a “visionary, creative, strategic” role should, logically, ask questions of vision: what do you see in our market? what would you do differently? what opportunity do you see that others wouldn’t?

Instead, it asks: have you already done exactly this? In the same sector? With the same title? The same scope? The same tools? You probably smiled reading that sentence. That’s normal. It’s absurd. And yet a lot of interviews work exactly like that.

To the point where if you’re a greengrocer who has never specifically distributed carrots, you’ll be told that won’t do — because here, you see, we’re dealing in carrots, not other vegetables.

They say they’re looking for vision. But they interrogate repetition. If you’re looking for an executor, that’s consistent. If you’re looking for vision, it no longer is. Otherwise, you’re not recruiting a vision. You’re recruiting an archive.

Because vision and execution aren’t always the same talent. The great houses have almost always been built by pairs: the one who sees what doesn’t exist, and the one who makes it real. Two different talents, both rare, both necessary. But recruitment sometimes demands the impossible profile: visionary, operational, pragmatic, affordable, linear, proven, low-risk, immediately productive, easy to manage, and perfectly aligned with the job description.

That’s not talent detection. That’s fantasy procurement.

You don’t ask an art director to prove their worth by sewing on an assembly line — someone else sews better, and that doesn’t make the vision any less precious.

The worst part is that there’s a simple way to test vision: give me a real problem, a context, a tension, an ambiguous case, and watch what I see, what I prioritize, what I’d refuse. I offered case studies. I was never asked for them. No exercise, no space to show the thinking.

You can’t claim to be looking for vision if you never create a situation where vision can appear.

They don’t want to test thinking. They want to recognize a format. A ready-to-reheat candidate: pre-cut, pre-packaged, pre-compliant. Microwavable talent.

The decorative version of talent

Companies don’t always look for the qualities they advertise. They look for the decorative version of them. “Outside the box” as an ad line, not a person who actually steps out of the box. Cross-disciplinarity as a keyword, not a career that has actually crossed several worlds. Creativity — as long as it doesn’t disturb the frame. Personality — but calibrated. Singularity — but not suspicious.

I have in mind a brilliant developer, a stellar career, AAA titles to his name, dropped from a role because judged “too pragmatic” in his relationship to AI (an article I devoured on Medium). Too pragmatic. In video games. A field where ideas only become worlds if someone knows how to make them work.

You don’t recruit out-of-the-box profiles by asking them to prove they fit perfectly inside the box.

The vocabulary of the misfits, without the misfits

There’s one more irony, an almost perfect one.

Look at the vocabulary of modern job descriptions: multitasking, priority management, strong powers of concentration, attention to detail, systems thinking, cross-disciplinarity, the ability to see patterns, creativity, autonomy, out-of-the-box thinking, problem-solver, architect, builder.

This vocabulary isn’t neutral. It was largely borrowed from the people we long called the atypical, the nerds, the misfits, the intense profiles, the autodidacts, the nonlinear paths — the people who connect too many things, too fast, too far, sometimes too strangely for the very organizations that claim to be looking for them. A vocabulary borrowed from a fraction of the population — perhaps 2%: atypical paths, the highly gifted, the neurodivergent, outcasts, misfits, nerds. And the defining feature of those paths is precisely nonlinearity. In other words:

the market took the vocabulary of the misfits, but it doesn’t always want the misfits.

It wants attention to detail, but not necessarily the obsession that comes with it. It wants systems thinking, but not necessarily the person who sees the system before everyone else and therefore becomes immediately tiresome in meetings. It wants creativity, but not the deviation. It wants pattern recognition, but not the path that learned to recognize patterns because it crossed several worlds. It wants “thinking outside the box,” but preferably with a trajectory perfectly inside the box.

And for the other profiles — those whose logic is simply different — they’re still asked to bend to this grammar: to become square when they’re round, to become round when they’re triangles, to become legible in a shape that isn’t theirs. To be clear: I’m not talking about ability, or intelligence, or competence. I’m talking about personality, constitution, form of logic. You’re forced to become a square when you’re a circle — and triangles are forced into the square.

Modern recruitment claims to look for rare shapes, then worries when they don’t resemble the shapes it already knows. Once again: magnificent era.

The readability premium

When a profile is complex, very senior, very cross-disciplinary, the reader may not understand. Not from a lack of intelligence — from a lack of context. They don’t necessarily know the industry, the history of the field, the real vocabulary, the emerging trajectories, the hybrid disciplines. So they take the most readable candidate. The one who ticks the boxes. The one who’ll be easy to defend if it goes wrong.

And that’s the hidden mechanism:

the safest candidate becomes the safest explanation.

The market doesn’t always reward the best candidate. It often rewards the candidate easiest to justify.

The problem gets worse by a notch when you know who reads first. Before reaching the manager or leadership, you go through pre-screening, screening, a recruitment officer, sometimes a junior, sometimes an intern. These people aren’t bad. But they’re handed the first reading of very specific, very technical, very nonlinear profiles, without always having the domain context. So they do what the system asks: tick the boxes. The more precise your sector, the richer your trajectory, the higher the risk that your résumé will be read by people who know the grammar of recruitment better than the reality of the job. When an impoverished chain of reading meets rich trajectories, it doesn’t see richness. It sees noise.

Talent detection or lexical obedience?

Many of today’s jobs didn’t exist when some of their best practitioners started: CX, UX, product thinking, mobile-first, platform strategy, service design, trust & safety, prompt engineering. The people who actually did these things before they had names don’t necessarily hold the recent degrees that certify them — nor even the job titles of the era.

Some profiles don’t lack competence. They simply lack the buzzword to name a competence they’ve practiced for a long time.

And HR AI, meant to read better, can reproduce the underlying problem. A model built to match résumés and job descriptions will look for lexical proximity, the degree, the title, the recent positions. But that’s not talent detection. A system trained to recognize similarity to a job description is not a system designed to detect talent. It’s sometimes lexical obedience. And in recruitment, what looks like an outlier isn’t necessarily noise. It’s sometimes the signal. Plainly: the system doesn’t always ask “can you do it?” It sometimes asks “can you describe yourself with the right keywords?”

Existential packaging

I was told, one day, that for someone so creative, my résumé was too austere. And I wondered how far it would go. Did I need a TikTok dance video? An engaging Instagram with cats? Little flowers in the header? A more “lively” typeface? Did I need to prove I’m cool and likable?

The problem is that, after several exchanges, the person hadn’t seen that I was precisely the opposite of what they were proposing: a geek, a minimalist, in love with black and tolerant of blue — only the Helvetica typeface lives in my heart — and, evidently, possessed of a dry humor. She was talking about personality without having read the personality. Then we discussed adding my civic commitments, my causes, my empathy, my favorite sports to the résumé. As if I were no longer applying for a job. As if I were running for local office.

This is no longer talent reading. It’s existential packaging.

And an authenticity you have to stage, color, calibrate, make LinkedIn-compatible is already no longer quite an authenticity. It’s a performance of authenticity.

That’s where the headhunter’s craft slipped. Before, they came to you: “I think you have something, I have an opportunity, does this trajectory interest you?” Matchmaking, intuition, the ability to flush out a talent before anyone else. Today, some tell you: your résumé is illegible, redo it, clarify it, make yourself more readable. Excellent intro résumé, but too cross-disciplinary — you need a résumé targeted with surgical precision. The person supposed to recognize talent asks you to become your own press office. They no longer really hunt heads. They sell boxes.

At this stage, you don’t need a résumé anymore. You need a capsule collection: no fewer than 120 résumés, 12 different t-shirt colors, 5 headshots from friendliest to most SaaS, and the cover letters to go with them.

The cool veneer of a conservative system

I did part of my training at a great hardware-and-software brand — one of those that helped popularize hiring for potential and personality: attitude, posture, the relationship to the customer counted as much as proof of execution. I remember a young blogger of that era who, after several failed interviews, said in substance that it was almost harder to get in there than into the CIA — for example.

Since then, many companies have borrowed the codes: relaxed tone, first names, coffee, the vocabulary of potential, flat management, “thinking outside the box,” “problem-solver,” even “super hero.” But behind the veneer: boxes, degree, conformity, identical proof, fear of the truly out-of-the-box.

They use your first name, but they read you like an administrative form. They offer you a matcha, but they hire as if the world hadn’t changed.

This isn’t a modern culture. It’s a conservative system with a friendly tone. As if the veneer had outlived the discipline.

The irony peaks when companies of fewer than two hundred people wonder whether a former Large-Corporate profile — out of a global brand, millions of users, standards copied everywhere — will withstand the pressure of their structure. They borrow the words, the gestures, the cultural vocabulary of those very brands, then doubt the ability of the people who came from them to function. Do you feel the irony? Or this variant, more delicious still: I’m asked, in an interview, whether I know how to use Workspace or Outlook — on a résumé and a story that announce I’m arriving straight from tech. At that level, it’s no longer a question. It’s an art installation.

And then there’s stability. On my résumé, an old period — ended more than ten years ago — where I stayed ten to eighteen months at certain companies. I was young, an expat, autonomous, curious; I moved, I was discovering myself, I evolved at each post, with glowing references, by the way. In interviews, it’s pointed out to me with suspicion. The job descriptions demand adaptability, mobility, strength of personality — then suspect the very path that proves them.

There are kinds of stability that hide fear. There are kinds of movement that reveal construction.

So my greatest crime would be that more than ten years ago, young and passionate, I didn’t have the delicacy to stay ten years in the same role to reassure a recruiter of the future. How reckless of me.

The candidate who became their own recruiter — and employability that becomes investability

Let’s recap what’s now asked of the candidate: to become their own recruiter, their own headhunter, their own press office, their own brand strategist, their own ATS consultant, their own storyteller, their own proof engine. All of it before even being taken seriously.

The burden of interpretation has shifted from the recruiter to the candidate.

People are no longer read. They’re asked to arrive already translated — a bit like a translator who’d ask you to arrive already translated.

But there’s a reversal. When a candidate has finally built enough signal to attract recruiters, they also become visible to readers, partners, clients, media — sometimes investors. And the question flips. At the start: why should this company hire me? After all that signal work: why do I still need this company? Not out of arrogance. Out of economic logic. If I had to become my own recruiter, my own PR, my own proof engine to be noticed, what is the company still offering me in return? A narrow job description? A role that captures only a fraction of the person?

The market says it wants to attract talent. Then it teaches that talent why it should escape.

By forcing nonlinear profiles to become visible outside the recruitment system, it also teaches them how to leave it. You’d almost laugh. If it weren’t tragic.

The hidden cost of not reading

We always treat the candidate as the risk: too senior, too junior, too expensive, too atypical, too hard to read. But what’s the risk created by the system itself?

Look at the reposted ads. Every week, every month, sometimes three, five, ten times — same title, same urgency, same requirements. Some are evergreen, fine. But when the same offer comes back endlessly, you have to ask the question: if it isn’t a signal of failure, what is it? An unfilled role costs HR time, delays, lost productivity, overloaded teams.

How much risk does the company create by not knowing how to read candidates properly? At some point, if the same ad returns more often than certain Netflix series, maybe it’s time to stop calling it a “demanding process” and start calling it a symptom.

And around this inability to read, an entire economy has flourished. To candidates, the message is: the system is opaque, the ATS filters, optimize, make yourself visible, redo your brand. Then they’re sold résumé optimization, personal branding, coaching, content strategy, even cognitive tests. To companies, the message is: you’re drowning, you can’t read anymore, you need AI, matching, scoring, recruiter chatbots. And in the middle, a market blooms. Not necessarily out of cynicism — but because the system is broken.

A market that fails to read talent creates another market that resells legibility to both sides.

Name the anxiety. Amplify the anxiety. Sell the workaround. It’s exactly the pattern we find around generative AI: denounce generated content, then sell the tools to “humanize” AI and make it undetectable. If it weren’t so sad, it would be fascinating.

This is not only a recruitment problem

You might think I’m stretching all this. But recruitment is only a tiny window onto something larger.

Distress at work is massive: depression, burnout, disengagement, loss of meaning. Brilliant people reduced to tasks too small, creative profiles locked in frames too narrow, competent people exhausted from proving their compatibility, pushed to the edge inside a logic that isn’t theirs. We talk a lot about the dangers of AI for employment. But one of the immediate dangers may already be here, and it has nothing of a machine about it: our inability to read humans at work.

To read a person badly is to place them badly. To place them badly is to use them badly. To use them badly is, sometimes, to break them. And grinding down competent people inside systems that claim to be looking for talent should worry us far more than it seems to. So this article isn’t only a critique of recruitment. It’s a small window onto our way of confusing competence with conformity, personality with packaging, vision with repetition, potential with past proof, talent with legibility.

Companies say they’re looking for talent capable of thinking differently. But too often, they can only welcome those who present themselves in a familiar shape. They want the out-of-the-box, on condition that it knows how to fold into the box at screening time. They want vision, but ask for an archive. Personality, but demand packaging. Cross-disciplinarity, but worry the moment a path actually crosses several worlds. The future, but hire the past with a coffee and a first name.

They’re surprised they can’t find a shoe that fits. But the system keeps turning everyone into sneakers.

Maybe the good profiles aren’t missing. Maybe they’re simply exhausted from having to present themselves in the same shape.

You can call this a recruitment error. I think it’s a societal problem.

And it doesn’t stop at hiring. The same disease runs the rest of it. We reward signal — visibility, exposure, attention, likes, reposts — over the work, and over the talent. You’re allowed to be called an expert once you’ve cleared some threshold of followers, as if an audience were a peer review. Visibility used to be a private choice; now it’s a tax. Be seen, or be assumed to have nothing to show. And proof itself now means the crowd — a wall of endorsements, a network of former colleagues vouching for you. I never stayed late at the office party, never baked the cake for everyone, to collect them; I gave my years to the work. So I’m assumed to have nothing to show for it. Don’t be shy, don’t be private — or build yourself a persona, quickly. And that is, once again, precisely the profile the atypical, the off-the-grid, the deep workers are least likely to fit. So here we are, watching the perfection of simulated, synthetic confidence in a hollow room — pressing the like button like a coin into an arcade machine, and mistaking the high score for talent.

Here is my most honest confession.

I’ve always written — practically since I learned how. Composed music, too. I stuck to my screens — gaming, coding, designing, writing — far more than I ever went out. My drawers are full of novels finished and unfinished, manifestos, drawings, articles, essays never published. For nearly fifteen years I had no network — nothing but a LinkedIn an employer made me open.

Because I prefer the shadow to the stage: I’m more at ease there, and my opinions I like kept close. I’m not shy; I hold back — not out of modesty, but out of caution and judgment.

And then I no longer had the right to speak unless I could prove it first: to show, to display, to showcase. So yes, I too embraced the sirens’ song — but in my own way, in my own voice. Every line I publish, I publish with my judgment, by a real choice, and with the responsibility that comes with it. At least, in the part that matters most, I remain who I am. If I’m going to be seen, I’ll be seen as myself — not some fashionable, formatted simulacrum.

And you?

If something here landed — a path someone misread, a moment you had to arrive “already translated,” a profile a screen quietly let slip — tell me. Yours, someone’s you hired, someone’s you watched disappear. The comments are yours.

This essay owes a great deal to an exchange with Liz Dugan,

whose work on psychological safety and neurodivergence at work — *We took away psychological safety and then told everyone to be more productive — sharpened its final section; and to **Iuliu-Cosmin Oniscu’s **[The Machine Has No Style: Why AI Execution Still Needs Human Vision](https://medium.com/@iuliu-cosmin-oniscu/the-machine-has-no-style-why-ai-execution-still-needs-human-vision-bf3a05aa4781).*

Author’s note:

In the spirit of this piece, I am proudly AI-assisted. This essay was originally written in French. AI tools were used for translation, proofreading, and structural editing. The ideas, the arguments, the examples, the personal experiences, the original French draft, and the vision remain my own work — and I answer for them.

A note on anonymity:

brand and employer names have been deliberately omitted throughout this piece. It’s a personal choice, out of respect for the places I’ve worked.