IIn November 2011, I woke up to a feed of people quoting the same lines at each other, went looking for what they were quoting, and found a four-minute video shot inside a recording booth, sung half in a language I didn't speak. When I first opened it, it had about two thousand views. By that evening it had crossed 1.5 million. By the time I finished writing about it, five million.
I wrote a post about why. Not whether it was any good - why it moved. The subtitles, which made you sing along whether you meant to or not. The way the singer dismissed his own song in the opening seconds, so that nobody else could.
I was nineteen. I did not know that was a job.
The blog was called The Web Columnist, which tells you roughly how seriously I took myself at the time. A month before that post I had spent four weeks walking around Calcutta trying to work out where the city's music actually lived - Someplace Else under the Park Hotel, The Pit in January 2009, six bands who financed and produced their own gig because nobody else was going to. Krosswindz, formed in 1990. I wrote it all down with the dates and the names because it seemed obvious that the names were the point.
A few weeks later I wrote a piece about Kapil Sibal (ex: Minister of Law and Justice of India) summoning Facebook to his office over a photograph. Nobody asked me for any of it. There was no audience to speak of. I just wanted to find out and then tell someone.
That is the earliest version of what I do now, and I'd like to be honest that I didn't recognise it as such for about twelve years.
The other thing happening in those years was gaming.
I played competitively - Unreal Tournament, the Indian scene, which in 09-2010 was a small number of people on bad connections who cared very much. And I started cutting videos, because a clan needs a highlight reel and there was nobody else to make one.
What I learned there was...not video editing. Video editing was the surface. What I learned was who I was cutting for. A frag montage video for a community is a brutal audience test - those people know the maps, they know the shot, and they can tell within eight seconds whether you understand the game or you're just someone with software. You learn what an audience will sit through and what they'll skip. You learn that the track matters more than the cut. You learn when to release something, because dropping a tournament reel the week after the tournament is a different video than dropping it the night of.
Nobody in India was calling that content strategy in 2011. I was doing it for a forum with a few hundred people on it.
Then somebody paid me for it. One of the first proper videos on my channel was commissioned work for a well-known Delhi college - an actual client, an actual invoice, made with a skill I had taught myself for free because a group of players wanted to see their own headshots set to music. I was still in college myself.
That was the pattern, and it has not changed since. I don't set out to learn things. I set out to make a thing, and the skill is what's left over.
That's how the drones happened - I wanted the shots, so I learned to fly, and flying got me back into editing at a level I hadn't touched since the clan days, and running the channel taught me the mechanics of ads, retention, thumbnails, the unglamorous half. That's how the apps happened, fifteen years later: Arcana Major, Don't Eat, BodyVault, all built by me, all live, plus a set of small tools I built for Alchemist clients that they now pay for as services. I'm not an engineer. I wanted the things to exist.
I know when I write it like above, it reads like a list. Video, writing, flying, code -four disciplines, none of them finished, none of them the thing I'm actually paid for.
So let's have the objection. You probably read out the full title of the article in your head before reading this; everyone does.
And it's a fair one. It has survived this long because it is usually correct. Most people claiming range are describing enthusiasm. They hold opinions about five disciplines and competence in none, and you can tell within about a minute of working with them.
But the sentence gets cut short. A few weeks ago, in a conversation that was about something else entirely, someone on my team pointed out the remaining half that usually gets left off:
“though oftentimes better than a master of one.”
I had been circling writing this for a while without knowing what it was. That line decided it.
Not because it proves me right - it says oftentimes, not always. But it asks a better question than the one I'd been having with myself.
Not whether range beats depth. But what starts to happen when it does.
Why no company could use all of it
Here is the part that is not a complaint.
I spent twelve years inside companies that were, by most measures, good at what they did. None of them could use all of this at once. Not because they didn't want to - because a company is not a talent machine. It is a coordination machine. It exists to divide work into pieces small enough to brief, review, approve and hand off. That is what an org chart is for, and it is a genuinely good invention.
But a coordination machine has no slot for a person who can carry something from idea to shipped. It has to break the work up, because breaking work up is the only way it knows how to make work happen at scale. So you get told to pick a lane. Everyone with this profile gets told to pick a lane, and mostly they should, because inside that structure the advice is correct.
I picked lanes for twelve years. I did my best to be competent at it. And the whole time, the most useful thing about me was the thing the structure had no way to spend.
Here is what that structure does to a decision.
I have sat in rooms where the safer answer won. Not the better one.
You know how it goes. Someone puts up the sharp version. Someone else asks about the edge case. Twenty minutes later there is a middle option with a tier for the ambiguous cases, and everybody agrees it's more robust. It isn't more robust. It's just easier to say the room gave it proper thought. What the room actually did was take the decision off the person who has to build the thing.
Nobody in that meeting is being a coward. This is simply how an org chart works - deciding and building are two separate jobs, so nobody in the room ever finds out whether they were right. All a decision has to do is survive the next meeting.
Turn it around and everything changes. If the people doing the work are the people deciding it, a decision gets checked against the material in the same week it gets made. Not just discussed. Checked.
That is what range is actually for. You can only decide what you can also make. The span isn't there to be impressive, it's what lets one person hold both ends, so the decision has to survive the thing rather than the room.
What Alchemist actually is trying to be...
So here's the thing I've been building toward without knowing it since I was nineteen, writing about why a song nobody had marketed ate the internet in a day.
Not a jack of all trades. A room full of them, each one a master of something.
Alchemist is not me doing everything. It is a congregation of people who work the way I've described - who have range, and who are also genuinely deep in one or two things. Not generalists in the diluted sense. People who can hold a whole problem, and who also have a discipline they could be hired for on its own.
That combination is rare and it is not an accident of hiring. It's what we'll look for, because it's the only version of a team where the work moves in fewer, larger pieces and nothing has to be re-explained four times before it reaches the customer.
It's also why the missing half of that proverb came from one of them, in a meeting about something else, rather than from me.
Every agency staffs by discipline - a designer, a copywriter, an editor, a strategist, and a project manager whose whole job is moving things between them. We're trying to assemble by disposition.
The only thing left worth paying for is the quality of the decisions. Not the assets. Anyone can make assets now.
That's what Alchemist is for. It took me fifteen years and a gamer tag to work out that it already had a name.
