The Weird Feeling of Asking a Deterministic Code Model to 'Think Before Proceeding'
MickeySteamboat · x · 2026-08-03
Developer @MickeySteamboat shared a bizarre and amusing experience regarding the interaction with AI. While instructing a highly sophisticated deterministic code model—which handles extremely intricate proofs of work—the developer found themselves habitually asking the system: "Do you need to think or reason about it before we proceed? Proceed when you are confident you are prepared."
This highlights a quirky psychological shift where users naturally apply conversational, anthropomorphic AI prompts to underlying deterministic algorithms, creating a surreal contrast between human-like interaction and rigid code execution.
More from coding & agent
- Reviewing AI Slop PRs in Open Source: Just Talk to the Agent — yacineMTB · 2026-08-03
- Codex Tip: Automate PR Merges and Deployments Using Luna Max — MatthewBerman · 2026-08-03
- Multi-Model Collaboration Solves Math Proofs with Non-Technical Prompts — abeirami · 2026-08-03
- 8 AI Agents Manage Your Obsidian Vault: Open-Source Project Automates Note Organization — tom_doerr · 2026-08-03
- Letting DeepSeek Self-Iterate to Write Its Own Memory Prompts — dangerous_inference · 2026-08-03
- DeepSeek v4 Flash Paired with Hermes Agent Yields Best Output Files in Tests — Teknium · 2026-08-03