All articles tagged with #prompting
Axios's Jim VandeHei explains how to get more from AI by teaching it to remember you: beyond working memory in a single chat, cultivate lasting memory by flagging what to save, auditing stored memories, mining past chats for patterns, and using topic-focused workspaces (Projects and Gems) to keep context. The guide also provides prompts to save preferences and review memories, enabling smarter prompting and long-term AI-driven workflows.
A Penn State study using ChatGPT-4o shows that increasingly rude prompts yielded higher accuracy (about 84.8% for very rude vs. 75.8% for very polite, with ~80.8% for very polite), challenging earlier work that politeness boosts performance. The researchers note that tiny prompt wording changes can drastically affect outputs and cautions against deploying hostile interfaces in real-world use, while acknowledging the findings are not a license to insult AI.
Writing effective prompts is crucial for ChatGPT to produce accurate and relevant responses. Prompts should include a specified topic, background information, limitations, and a clear goal. Users should be mindful of biases and review and verify responses. Providing ChatGPT with system-level directives can also help direct its behavior. Following these tips can improve the outcomes of interactions with ChatGPT.