A junior-high dropout who can't draw won an Apple Design Award
A self-taught outsider with a junior-high education who cannot draw built the first non-game app from China to win an Apple Design Award — by running into the wall again and again and never letting a feeling of ‘something's off’ slide. Taste is not a gift; it is a habit built by long training.
The video won't play here. Listen to the audio instead:
The argument · tap a timestamp to hear it
The product metaphor came first; the brain science was retrofitted later
In 2022, when his daughter was a little over two, Qiang Ge noticed something while out walking with her: she would spontaneously say the English words for the paving tiles, plants and buildings she saw over and over on the street, but had no reaction at all to words a translation device read out mechanically. He also watched her play with stickers — peeling them off, placing them, spacing them, all done deliberately — and out of that came a metaphor: peel real-world objects off like stickers and attach language information to them so it stays. Only later, when a friend gave him a book on brain science, did he learn this matched the principle of contextual memory. At the time it was pure intuition.
— Qiang GeSpeed is the best experience; the animation only exists to serve it
At first, image recognition plus cutout took ten-odd seconds, and Qiang Ge felt no user could tolerate that — but he does not come from a technical background and could not solve the compute problem. So he thought back to the full motion of his daughter peeling a sticker: peel it off, pick it up, stick it on. He turned that natural sequence into an animation. The moment the user presses the shutter, recognition has already started; by the time the animation finishes, the result has appeared. His conclusion: speed is the best experience, and all those flashy burst animations are only there to serve speed.
— Qiang GeThe years when he didn't know the word ‘design’ were the real education
In junior high, Qiang Ge spent every day at internet cafés, using Photoshop to make QQ Zone skins and pick out music for them — but at the time he had no idea the word ‘design’ existed; he only knew he was making things. After dropping out, he kept posting work on forums, and a knockoff-phone company in Hangzhou hired him on the strength of that work. It was only after coming to Beijing that he learned design was a profession with a whole system behind it, and the colleagues around him put him under enormous pressure.
— Qiang GeLooking for blue once a week trained a reflex he still has
At his first company, the boss gave him a weekly exercise: this week, go find blue, then sort what you find by how it feels to you — professional, mysterious, sci-fi — and print it as a poster to pin up at your desk. He was also told to trace the shapes of concept cars in Photoshop and generalize the visual feelings involved. Qiang Ge says the training built a conditioned reflex in him: now, whenever he sees any object, he unconsciously starts thinking about why it was designed the way it was.
— Qiang GeTalking through how you solve the problem beats handing over the answer
Qiang Ge admits that leading a team was painful in the early days: he was always dissatisfied with a proposal but could not say why, the team did not understand him, and it even led to mental-health problems. Later he deliberately stopped saying ‘I feel like’ and switched to laying out the goal, the assumptions and the possible outcomes, walking through his entire thought process out loud. He says that once he moved from only giving conclusions to giving the process of solving the problem, the team actually did better work.
— Qiang GeA design plateau isn't too little practice — it's internalization that hasn't finished
In 2016 Qiang Ge ran a training program in his team built around absorbing, internalizing and outputting, comparing a person to a computer: input the signal, process it, output. He used the plateau in weight training as an analogy for a design plateau: the bottleneck is not insufficient training volume, it is that internalization has not completed — your eyes have taken in too many good things and your hands cannot keep up. The only way through is to keep making things; once the signal has been processed, the work naturally moves up a level.
— Qiang GeAI raised the bar for building products rather than lowering it
Qiang Ge uses AI heavily to test new features, and results come fast — but he found that things had actually slowed down: choosing became difficult, attention got scattered, and feedback arrived too fast to internalize. He says AI raised efficiency yet made the work itself slower, and the bar for building products is now higher, because there is less deep thinking. He has changed how he works, giving AI a limited set of choices instead of constantly trying new things.
— Qiang GeHe isn't good at anything else, so he never agonizes over whether he chose right
Asked whether design was an active choice or the only option, Qiang Ge says the latter: there is nothing else he is good at, and he cannot even squeeze out the time to get into vinyl. He rarely agonizes over whether a choice was correct. The biggest choice of his life was leaving the company, and his test was to ask himself whether he would regret it in two years — the answer was that not leaving would definitely be a regret. He says doing design is painful and joyful at once, and he is willing to keep at it.
— Qiang GeIn their own words · checked verbatim
Take, for example, watching a magician do a trick. He picks up a top hat, and suddenly there's a rabbit inside. You feel delighted. But once you know how the magician did it, are you still delighted?
就比如说我们看到一个魔术师在变魔术的时候。他拿了一个礼帽，里面突然变成个兔子。你会觉得很惊喜。但是你要知道魔术师怎么变的时候，你还会惊喜吗？
Qiang Ge19:30
Back then I genuinely didn't know what those two words, ‘design’, meant. In my head the concept didn't exist. I just knew how to make. I made that Zone page.
我当时就真的不知道什么是设计这个两个词，我在我概念里面没用。对我就知道做。做了这个空间。
Qiang Ge26:42
What I call breathing room is: it looks like it could still go one step further. But it doesn't go further, and I can still hand it in.
所谓的呼吸时间是。他好像还可以再前进一步。但是他不前进，我也能交差了。
Qiang Ge48:29
You might have tried a hundred methods. And at the end you got that cover open — the engine side cover, I mean — with no manual, no screwdriver, nothing at all.
你可能试过100种方法。到最后你打开了这个盖子，就是这个发动机侧盖，你没有说明书，没有螺丝刀，没有什么东西的时候。
Qiang Ge1:00:59
So AI raised efficiency, right? But it seems to have made the work itself slower and slower. Because choosing has become so important.
就是AI提升了效率，对不对？但是他好像。让事情本身越来越慢了。就是因为选择变得很重要了。
Qiang Ge1:06:06
Figures
| Time from launch to winning the Apple Design Award | three or four months | 1:00 |
| Image recognition wait time | ten-odd seconds | 12:19 |
| Versions scrapped and rebuilt during development | five or six | 21:32 |
| Official product launch | February 2025 | 10:19 |
| Breathing room left before handing work in | two or three days | 47:28 |
| Time it took to finish Zen and the Art of Motorcycle Maintenance | half a year | 58:49 |
Glossary
- Flashcard
- A study card with a piece of knowledge on one side and its explanation on the other.
- Feature phone
- A non-smartphone that can still take photos, get online and play music.
- Quality
- The standard of quality in Zen and the Art of Motorcycle Maintenance that resists being put into words.
How to listen
Designers, product managers and indie developers — and anyone who earns a living on instinct and taste but cannot articulate why something is good.
The final 1 minute — the closing ad and the next-episode trailer — is skippable.