Shi Warren (@wslabs)

I Learned Where AI Should Stop Being Creative

I used to think the obvious way to improve an AI product was to add more AI. If one part of the workflow could call a model, why not let the model handle more of the workflow? That assumption held up

Oct 19 · 6 min readwslabs.hashnode.dev

I Stopped Treating AI Outputs as the End of the Workflow

When I started building AIMakeTattoo, I thought about the product as a group of separate tools. One workflow started with a text idea. Another let someone begin with a photo or reference image. Anothe

Sep 25 · 6 min readwslabs.hashnode.dev

I Built the Same “Modify” Feature Twice. The User Contract Was Completely Different.

When I first added a “Modify” feature to my tattoo product, I assumed the interaction pattern would be reusable. Both workflows looked similar on the surface: generate an image keep the result descri

Aug 28 · 5 min readwslabs.hashnode.dev

I Needed AI Image Editing. I Didn't Want to Build an Image Editor

I already had a working image generation flow. Users could describe a tattoo idea, choose a style, generate a result, and try again if they wanted another direction. That was enough until I started no

Aug 15 · 6 min readwslabs.hashnode.dev

Your AI Generation Endpoint Probably Needs a State Machine, Not Another try/catch

Most AI generation endpoints look simple in the first version. A request comes in, you call the model, wait for the result, and return it. Something like this: try { const result = await generateIma

Aug 7 · 10 min readwslabs.hashnode.dev

Shi Warren

About

Available for

Shi Warren's blogs

I Learned Where AI Should Stop Being Creative

I Stopped Treating AI Outputs as the End of the Workflow

I Built the Same “Modify” Feature Twice. The User Contract Was Completely Different.

I Needed AI Image Editing. I Didn't Want to Build an Image Editor

Your AI Generation Endpoint Probably Needs a State Machine, Not Another try/catch

Shi Warren

About

Available for

Shi Warren's blogs

I Learned Where AI Should Stop Being Creative

I Stopped Treating AI Outputs as the End of the Workflow

I Built the Same “Modify” Feature Twice. The User Contract Was Completely Different.

I Needed AI Image Editing. I Didn't Want to Build an Image Editor

Your AI Generation Endpoint Probably Needs a State Machine, Not Another try/catch

Shi Warren

About

Available for

Shi Warren's blogs

I Learned Where AI Should Stop Being Creative

I Stopped Treating AI Outputs as the End of the Workflow

I Built the Same “Modify” Feature Twice. The User Contract Was Completely Different.

I Needed AI Image Editing. I Didn't Want to Build an Image Editor

Your AI Generation Endpoint Probably Needs a State Machine, Not Another try/catch

Search Hashnode

Shi Warren

About

Available for

Shi Warren's blogs

I Learned Where AI Should Stop Being Creative

I Stopped Treating AI Outputs as the End of the Workflow

I Built the Same “Modify” Feature Twice. The User Contract Was Completely Different.

I Needed AI Image Editing. I Didn't Want to Build an Image Editor

Your AI Generation Endpoint Probably Needs a State Machine, Not Another try/catch