Philippe Khin

Honored to speak at @AsiaSocietyJP's "Startup Nation Japan: The rise of non-native entrepreneurs". It was a privilege to share the stage with fellow founders Oya Koc and Hao Tran, guided by Jesper Koll whose sharp questions and easy humor kept the conversation lively. Born in Cambodia, raised in France, and now building @SewaYou in Tokyo, I’ve learned that steady progress and strong local ties open doors for founders far from home. With SewaYou we’re working to help people in Japan meet new friends nearby and to make cross-border connections just as easy. If we can build even a small bridge, that’s success. Special thanks to the Asia Society Japan team and everyone who joined! Photos: © 2025 Taishi Yokotsuka for Asia Society Japan

nobody would ever hire me if they saw how i talk to claude

ads on social media might be the industry that makes the least sense. the feed keeps getting better, im reading sharp ai posts on X, then some anime garbage ad takes my whole screen and i scroll past in half a second. and somehow this still works? somehow it’s an industry?

still debating if ai helps your velocity? cool. i just merged a few PRs from a park bench on my phone

consumer apps are the final boss of shoving "ai" into the product. don't. nobody gives a damn, their ai is already chatgpt. your product doesn't need ai. your product development does.

bragging about the xxx million aum at your old vc job tells me nothing about your contribution. you could've been the guy refilling the coffee

that smell of startups shoving ai down your throat, half assing it just to have a slide in their pitch, then calling themselves ai native

the best note taking app might just be X drafts you're about to post, plus plain apple notes. a note you never act on is just noise you're dumping on future you to triage

the f these people doing with their life

good infra should just "disappear". set up postgres on @PlanetScale and forgot about it, and not a single hiccup in 8 months same for @render never even checked their status page once the best compliment you can give infra is that you forgot it's there

marketing in 2026 is just vibe shipping a few github repos wrapped around "claude -p" and letting the agents do the rest

8 years in. new face, same mission. a new era for @SewaYou

@jfozonx well, this is all from the company POV benefits, as a user, i just hate being redirected outside the app i’m currently using 🤷🏻‍♂️

If the alternative isn't fancy, all the more reason to just use the Google Form. Login can be required, but doesn't have to be; form owner decides. Errors/reliability are on the dev? Doesn't really matter who is responsible. It needs to work for the customer. It's an unreasonable expectation on devs to ask for an integration that is more reliable than Google Forms, anyway. Feedback going into a spreadsheet is also perhaps ideal, at least up to a certain scale. Patterns are easily recognized. You can add a column for the action taken. You can add a column for the issue tracker ID. Google Sheets living on the cloud means that everyone with access to the sheet will see that stuff.

you know a startup is run by a bunch of pms punching spreadsheets when the in-app feedback redirects to a google form instead of a native in-app form

@jfozonx fair on reliability. but the alternative isn't a fancy form, just a text input that posts to slack or opens a linear issue. and a google form makes me log in, half the time on the wrong account. the error you hit? that's on the dev, can't really blame the form for that 😛

@philippekhin Guilty as charged. Have you never submitted some feedback only to meet an error? Users don't care what the form looks like as long as their problems are being solved. Google Forms reliability is going to be much better than anything a typical startup can arrange.

things i used to be excited about but not anymore: library updates with new features / fixes / apis. i'm not the user of those libraries now, my agents are, and they don’t care 🤷🏻‍♂️

why don't bros at airbnb and amazon spend a few million tokens to vibe code their dark mode? post-ipo and still can't ship it. can't believe there aren't at least a few million users impulse buying at night or booking a place without getting their eyes blinded

doesn't matter how good your features or design are. on a consumer social app where users meet strangers, one bad experience with another user = one-star review guaranteed

the most desperate resurrection email campaign in software history

users always pay the price when apps try to expand: ugly ads to monetize free users spammy upgrade modals with a crappy lottery spin wheel useless bloated features to pad their session time when the goal is volume, even paying power users take a hit. the app just gets bloated

programming used to look cool highlighted code popping on a dark text editor, command lines running in a black terminal à la matrix, memorizing keyboard shortcuts and all now it's just "bro fix this" on an iphone while taking a dump