// Hacker Noon · 10 February 2026

90% Failure Rate: Why GPT-4o Can't Optimize Code (And What We Built Instead)

GPT-4o fails 90% of the time at code optimization. Why zero-shot prompting doesn't work and how to build an agentic search workflow instead.

Hacker Noon

@hacker-noon · Saurabh Misra

hackernoon.com

Hacker Noon@hacker-noon