Why write code if the LLM can just do the thing? (web app experiment)
Details
- External ID
- 45783640
- Source
- HN
- Company
- —
- Product
- Why write code if the LLM can just do the thing? (web app experiment)
- Website domain
- github.com
- Launched
- Nov. 1, 2025
- Cohort
- —
- Upvotes
- 436
- Upvotes percentile
- 0.982532751091703
- Tags
- —
- Fetched at
- Sept. 7, 2026, 9:25 p.m.
- Updated at
- Sept. 7, 2026, 9:25 p.m.
Description
I spent a few hours last weekend testing whether AI can replace code by executing directly. Built a contact manager where every HTTP request goes to an LLM with three tools: database (SQLite), webResponse (HTML/JSON/JS), and updateMemory (feedback). No routes, no controllers, no business logic. The AI designs schemas on first request, generates UIs from paths alone, and evolves based on natural language feedback. It works—forms submit, data persists, APIs return JSON—but it's catastrophically slow (30-60s per request), absurdly expensive ($0.05/request), and has zero UI consistency between requests. The capability exists; performance is the problem. When inference gets 10x faster, maybe the question shifts from "how do we generate better code?" to "why generate code at all?"
Enrichment
- Theme
- browser automation and scraping for AI
- Vertical
- Horizontal
- Function
- Agent / copilot
- Audience
- Developer
- AI stance
- AI-native
- Project type
- Hobby / open-source project
- Normalized one-liner
- llm-powered web app automation
- Manually corrected
- False
Could you build this?
Yes This is a lightweight experiment that routes HTTP requests to an LLM with basic SQLite and HTML-rendering tools, which can be implemented in a few hours using standard AI APIs.
Discussion
20 comments analyzed.
Concerns raised: AI energy consumption not justified by 5-30% productivity gains, LLM non-determinism and inconsistency make them unsuitable for critical workflows, Training costs and ongoing compute requirements are prohibitively expensive, Quality and hallucination issues prevent convergence to reliable solutions, Jevons Paradox: efficiency gains lead to increased total resource consumption
Feature requests: Temperature and flow control separation for shower systems, Persistent code generation with automated quality oversight and testing, Deterministic workflow processes instead of LLM-decided outcomes
Competitors
Other products that read as similar to this one — 171 launches clear the similarity bar, closest 8 shown.
Attention rank: #5 of 172 (itself plus its competitors, highest first — normalized so YC and Product Hunt are compared fairly).
Launched 2 days after the earliest competitor.
- AgentAct AI · ph · 2026-09-30 · 1 upvotes · similarity 0.47
- Gram Functions · hn · 2025-11-19 · 23 upvotes · similarity 0.43
- Command Center, the AI coding env for people who care about quality · hn · 2026-06-08 · 69 upvotes · similarity 0.42
- OpenUI · hn · 2026-03-11 · 8 upvotes · similarity 0.42
- Mini-coder · hn · 2026-03-02 · 6 upvotes · similarity 0.41
- I nerfed our coding agents on purpose · hn · 2026-06-05 · 27 upvotes · similarity 0.41
- Teaching AI agents to write better GraphQL · hn · 2026-02-04 · 6 upvotes · similarity 0.41
- The Mog Programming Language · hn · 2026-03-09 · 163 upvotes · similarity 0.40
Other launches for this product
- No other launches for this product.
Same idea, different domain
Nobody's really built a agent / copilot tool for Agriculture yet.