personal_asset
Five Years Ago, the Job "Flash Developer" Disappeared
Flash developers vanished five years ago, not because the technology was bad, but because the platforms turned together. This time it is AI's turn.
On April 29, 2010, Steve Jobs posted an open letter on Apple's site titled "Thoughts on Flash."
It listed a few reasons: too many security holes, too much battery drain, and interactions designed for a mouse that do not move to touchscreens. The sign-off was blunt — the iPhone does not support Flash and never will.
Ten years later Adobe announced it themselves: from December 31, 2020, Flash Player would stop receiving updates and support entirely. Google, Microsoft, and Mozilla pulled it from the browsers at the same time.
Ten years passed in between, but the result was the same: an entire profession, gone.
What Flash actually did is unclear to many people now.
Between roughly 2005 and 2012, splash animations on web pages, corporate site intros, and browser games mostly ran on it. Knowing Flash and being able to build interactive animation was a marketable trade, listed as its own line item in job postings.
The year I entered the industry, my previous company's old site still had a fragment of Flash splash animation nobody had bothered to delete. Nobody used it, nobody spent effort replacing it. It just sat there like a fossil.
It was not beaten by a stronger technology. What killed it was Apple, Google, and Microsoft turning together, pouring resources and entry points into HTML5. No single company could hold onto Flash against an entire ecosystem turning.
How good a technology is has never been the only variable in how long it lives. Whether platforms keep accepting it counts just as much.
The disappearance of Flash developers did not mean those people disappeared.
Some moved their interactive animation craft into game engines and switched to Unity. Some picked JavaScript back up and continued in frontend. Some left the industry entirely.
The ones who caught the turn were not those most fluent in Flash. They were the ones who had casually learned something else on the side, with a foot in two boats. Mastering exactly one specific technology leaves you most exposed at the moment of the switch.
This script is not unique to Flash. IE6-specific compatibility hacks and heavily jQuery-dependent code went through the same process: not sudden death, just nobody asking about it anymore.
Now it is programmers in general who are nervous about AI, and the logic rhymes: several platforms pushing in the same direction — model vendors, IDE vendors, cloud vendors, all selling the same thing, driving down the cost of the step that translates requirements into code.
But compared with Flash there is one clear difference. Flash had a death date you could write on a calendar: December 31, 2020, cut in one stroke. This time there is none.
What is being compressed is not one specific technology but the price of a finer-grained step, and the pressure is continuous with no finish line. The absence of a clear death date makes it harder to guard against — nobody will remind you on any given day that now is the time to turn.
A window of advantage always closes, whether or not the end date is on a calendar.
In ten years as a programmer I have seen more than one of these turns. What actually travels across them was never a specific language or framework. It is the feel for judging how much longer this wave holds and where to step next.
That sounds like a platitude, but compare it against the Flash cohort and the difference shows: among people with the same ten years in the field, those willing to learn one more thing and keep one more exit were visibly calmer when the turn came.
When that open letter was published, nobody thought a profession would disappear because of a letter. In hindsight the signs were already there.
The thing worth watching is not which new AI feature shipped. It is how your own company's job requirements read six months from now. That is the real signal for how far this turn has gone.