Jiayi

Building toward an agent-native future for app growth.

[Jiayi @jiayiwhere]

@james406 A better way of saying you can’t pay the rent

[Jiayi @jiayiwhere]

Jev→ Laya → Laya-MLX → Kev in 3 days is wild guys.

Jev brought attentions to decision models.

Laya made the idea open-weight + multilingual.

Laya-MLX made it run locally on your Mac in ~7–14ms.

Kev basically said “what if we just build this on Qwen?” — and did.

Closed → open → local → built on existing models.

All happened in 3 days lmao.

[Arpit Bhayani @arpit_bhayani]

first came jev. then laya. then laya-mlx. and now kev.

all system one decision models. all within a span of 3 days. never seen a race this fierce before.

let us breathe, folks. let us breathe.

[Jiayi @jiayiwhere]

@suraj_sharma14 Money aside, the freedom to create and give is a reward in itself.

[Jiayi @jiayiwhere]

@anupamrjp Building became abundant. Attention stayed scarce.

[Jiayi @jiayiwhere]

@suraj_sharma14 @perplexity_ai Day 2 of building my X account to 1K.
Already learning that thoughtful replies to the right people > endless shitposting.

[Jiayi @jiayiwhere]

@Gaus450 $0 → $1 might be harder than $1 → $1k. Keep going.

[Jiayi @jiayiwhere]

@marclou Go viral gets you traffic. A real problem gets you revenue.

[Jiayi @jiayiwhere]

@Sandro7ge My trigger is when research stops creating new insight and starts creating more tabs.
If AI keeps giving me versions of the same answer, I go find someone who’s actually done it.

[Jiayi @jiayiwhere]

Solo founder lesson:

Spent 3 days researching something.

Asked someone who actually knew what they were doing.

Got the answer in 20 minutes.

New strategy:

AI for what AI knows.
Experts for what experts know.
Me for what I know.

Turns out “solo founder” doesn’t mean “personally reinvent every wheel.”

[Jiayi @jiayiwhere]

@Gaus450 Because attention isn’t a problem you can solve with more compute.

[Jiayi @jiayiwhere]

@alexcooldev The best onboarding is the one users barely notice.

[Jiayi @jiayiwhere]

Exactly. Human attention still runs on the same hardware. Meanwhile, the growth stack is still 10 different tools you have to stitch together yourself.

[Suraj Sharma @suraj_sharma14]

@jiayiwhere Marketing feels pre-AI because the other side of the screen didn’t get a context window. They still decide in a second.

[Jiayi @jiayiwhere]

App building in 2026:

Idea at lunch.
Shipped by dinner.

Spent the next 3 months figuring out why nobody is paying.

[Jiayi @jiayiwhere]

@suraj_sharma14 Ate. Marketing still feels pre-AI.

[Jiayi @jiayiwhere]

@Pranto39 Depends on what you’re launching and who your users are. I’d use Reddit/X to talk to potential users before launch, then share what worked, what didn’t, and what you learned along the way. People on X love following the journey when there’s something useful to learn from it.

[Jiayi @jiayiwhere]

@Byysid How are you measuring this back to actual app revenue — just views/installs, or downstream subscription revenue too?

[Jiayi @jiayiwhere]

@frederickjames Curious — once the stack is set up, what’s the biggest bottleneck for you now to growth: understanding what’s happening, or deciding what to actually change next?

[Jiayi @jiayiwhere]

@frederickjames Curious — once the stack is set up, what’s the biggest bottleneck for you now to growth: understanding what’s happening, or deciding what to actually change next?