AI assistants are designed to be helpful, but what happens when they’re put on the spot to tear apart a bad idea? Tom’s Guide tested three popular chatbots – Claude, Gemini, and ChatGPT – using the “Potato” prompt to evaluate an entrepreneurial concept. The results were surprising.
Entrepreneurs often seek encouragement and polite suggestions from AI tools, but what about harsh criticism that actually improves an argument? Amanda Caswell, a leading voice in AI and technology, created the “potato” prompt to flip this dynamic. She asked each chatbot to identify three holes in her idea for a ready-to-eat pizza company called Crusted.
The results were starkly different among the three chatbots. Claude was the harshest critic, attacking the core identity of the idea and questioning its very existence. Gemini focused on practical realities like food production and logistics. ChatGPT delivered a more business-oriented critique, pointing out economic flaws and logistical challenges.
In the end, Claude’s brutal honesty stood out for its emotional impact and clarity. Its language was sharper, using terms like “paralysis” and “entering a war,” which framed Amanda’s idea as fundamentally flawed. If you want an honest opinion without sugarcoating, Claude is the chatbot to try.
Source: https://www.tomsguide.com/ai/i-tested-the-the-potato-prompt-on-chatgpt-claude-and-gemini-heres-which-ai-is-the-harshest-critic