344
Productivity & Workflow355
Automation & Workflow224
Software Development250
Marketing & Growth192
AI Infrastructure & MLOps174
Writing & Content Creation203
Data & Analytics141
Design & Creative169
Customer Support131
Photography & Imaging156
Sales & Outreach125
Voice & Speech135
Education & Learning131
Operations & Admin87
A Wired guide explains simple ways to write better prompts in ChatGPT, including using photos, asking for shorter answers, and requesting clarifying questions.
In short: More people are learning that small changes to what they type into ChatGPT can lead to clearer and more useful answers.
A new Wired guide rounds up practical tips for writing better prompts in ChatGPT and similar chatbots, like Google Gemini. A prompt is simply the message you type in, like a question or an instruction.
The article says you often get better results by being more specific and by telling the chatbot what kind of output you want. For example, you can ask for “the 80-20” on a topic, meaning the few key points that give you most of the understanding (like reading the back cover before starting a long book). You can also set limits, such as “answer in 100 words” or “use a table,” which can make the response easier to scan.
Wired also highlights ways to change the chatbot’s behavior. One tip is to ask it to critique your idea like a curious 10-year-old, which can surface basic questions you may have missed. Another is to ask the bot to ask you clarifying questions before it answers, so you do less back-and-forth later.
The guide includes features beyond typing. In the ChatGPT phone app, you can add a photo to your prompt using the camera, then ask what something is or what a sign says in English. It also mentions “temporary chat,” a mode that does not keep the conversation after you are done.
As these tools spread, “prompting” is starting to look like a everyday skill, similar to learning how to search well on Google. It is still worth double-checking important facts, since chatbots can sound confident even when they are wrong.
Source: Wired