Want smarter insights in your inbox? Sign up for our weekly newsletters to get only what matters to enterprise AI, data, and security leaders. Subscribe Now
When OpenAI launched GPT-5 about two weeks ago, CEO Sam Altman promised it would be the company’s “smartest, fastest, most useful model yet.” Instead, the launch triggered one of the most contentious user revolts in the brief history of consumer AI.
Now, a simple blind testing tool created by an anonymous developer is revealing the complex reality behind the backlash—and challenging assumptions about how people actually experience artificial intelligence improvements.
The web application, hosted at gptblindvoting.vercel.app, presents users with pairs of responses to identical prompts without revealing which came from GPT-5 (non-thinking) or its predecessor, GPT-4o. Users simply vote for their preferred response across multiple rounds, then receive a summary showing which model they actually favored.
Some of you asked me about my blind test, so I created a quick website for yall to test 4o against 5 yourself. Both have the same system message to give short outputs without formatting because else its too easy to see which one is which. https://t.co/vSECvNCQZe — Flowers ☾ (@flowersslop) August 8, 2025
“Some of you asked me about my blind test, so I created a quick website for yall to test 4o against 5 yourself,” posted the creator, known only as @flowersslop on X, whose tool has garnered over 213,000 views since launching last week.
AI Scaling Hits Its Limits Power caps, rising token costs, and inference delays are reshaping enterprise AI. Join our exclusive salon to discover how top teams are: Turning energy into a strategic advantage
Architecting efficient inference for real throughput gains
Unlocking competitive ROI with sustainable AI systems Secure your spot to stay ahead: https://bit.ly/4mwGngO
Early results from users posting their outcomes on social media show a split that mirrors the broader controversy: while a slight majority report preferring GPT-5 in blind tests, a substantial portion still favor GPT-4o — revealing that user preference extends far beyond the technical benchmarks that typically define AI progress.
... continue reading