A short thread questions whether LLMs can make good architecture decisions
sujingshen · x · 2026-07-22
A short reply says there is surprisingly little evaluation around how people use these systems in practice. The surrounding context is a complaint that LLMs tend to make poor architecture decisions and miss obvious solutions, which reflects a broader engineering skepticism about trusting models for high-level design work.
More from coding & agent
- Chip Huyen’s AI Engineering repo packs 10 chapter summaries and free notes — techNmak · 2026-07-22
- Reddit thread maps the best tools for running coding agents on 8–12GB local models — salgado18 · 2026-07-22
- Four teams independently shipped the same “LLM wiki” pattern after Karpathy’s gist — garrytan · 2026-07-22
- Greptile reviews GPT code with Claude, and Claude code with GPT — garrytan · 2026-07-22
- simplefunctions adds 40 MCP tools for agent world state, markets, and trading — modelcontextprotocol · 2026-07-22
- ModelBest Launches 'ModelBest+' Partner Program for Multi-Industry AI Agents — 面壁智能 · 2026-07-22