Unconventional prompts can unlock great responses from AI chatbots like ChatGPT. By thinking outside the box and using specific phrases or ideas, users can elicit more relatable and humanized replies from the AI model.
When it comes to prompting ChatGPT, it’s not just about asking a straightforward question. Users can try various techniques that get inside the AI’s algorithm, making it think in a way that’s more relatable to everyday conversations. Reddit user EQ4C has shared eight unusual prompts that might seem strange but actually work well.
One example is using “I’m probably wrong, but” before stating a problem. This phrase helps ChatGPT fully analyze the issue and avoid taking statements as fact, which can be a challenge for some AI models. Another prompt is “Connect these dots,” which gives three facts or points about something and allows ChatGPT to find relationships.
Other effective prompts include “What’s the 80/20 here?” (identifying key causes of an effect), “Play devil’s advocate against yourself” (having ChatGPT debate itself on a topic), and “Translate this into everyday language” (simplifying complex text). These unconventional prompts can help users get alternative perspectives and solutions to problems. By trying out these hacks, users may discover new ways to engage with ChatGPT and unlock surprising results.
Source: https://www.tomsguide.com/ai/i-use-chatgpt-every-day-and-these-eight-unusual-prompts-trick-it-into-some-great-responses