A beginner's guide to generative AI and AI prompting: tools, techniques, and tips

I’m Bria, a tech enthusiast and go-to expert for all things gaming and gadgetry. Bringing over a decade of tech troubleshooting and passion for the latest gadgets, I’m here to guide you through your tech questions.

Generative AI isn't just a trendy tech term—it's actively reshaping how we tackle tasks. Whether writers tap into AI to whip up content, online sites use it to handle customer inquiries, businesses jazz up their email marketing campaigns, or artists looking to enhance their art with AI text and image generators, the impact is clear. AI is making waves in how we think about and use tech.

For instance, I recently showed my college-aged sister how to use AI to generate and flesh out essay outlines. She was amazed at how it turned her rough ideas into structured, detailed plans, making her writing process much smoother.

It’s a powerful tool that's not just for tech experts.

Understanding generative AI

What is generative AI? Well, it’s a type of artificial intelligence designed to create new content, such as text, images, music, and more. Unlike traditional AI, which is programmed to follow set rules or perform certain tasks, generative AI learns from vast amounts of data to produce outputs that mimic human responses, creativity, and expression.

At its core, generative AI works by predicting what comes next, whether it’s the next word in a sentence, the next note in a song, or the next pixel in an image. The more data it’s trained on, the better it can generate realistic and useful content. The result is a tool that can assist with creative work, automate repetitive tasks, and even suggest innovative solutions to complex tasks.

What are some popular tools for generative AI?

There are several generative AI tools, but here are some of the most popular ones.

ChatGPT: unlocking natural language potential

ChatGPT™, developed by OpenAI®, is a powerful AI text generator that uses natural language processing to produce a copy that feels like a human wrote it. Users provide prompts that guide the AI, and the generative AI model interprets the prompt, drawing from vast datasets to create text, images, or other content that aligns with the given instructions. This process allows for tailored, creative, and contextually relevant responses.

DALL-E: transforming text into visuals

DALL-E™ is an AI image generator that creates pictures from text descriptions. For example, if you need a unique image for a blog or social media post, DALL-E can create custom visuals based on your prompts. This tool is perfect for content creators and marketers looking for original, eye-catching visuals.

OpenAI Jukebox: generating music with AI

OpenAI Jukebox™ is an AI tool that generates music in various styles and genres. It can produce original compositions based on text prompts, making it ideal for musicians, producers, or anyone exploring AI-generated music.

Getting started with prompting in generative AI

Think about this: You're planning a dinner party for a small group of friends, and along with deciding how much food to prepare, you also need to manage a range of dietary restrictions—lactose intolerance, tree nut or shellfish allergies, and gluten sensitivity. Simply searching Google® for "a recipe for a dinner party" might give you thousands of options, but most won’t fit all those needs. Instead, you'd start by listing every dietary restriction as a prompt or framework for your search to find recipes that suit everyone.

In much the same way, using generative AI tools effectively starts with crafting the right prompts. Proper framing and AI prompt writing are key to getting the best results, whether you're creating content, generating images, or finding solutions to specific problems. Let’s break down some basic concepts to help you get started.

What is a prompt?

A prompt is an AI command you give to an AI generator to produce a specific result. Your prompt tells the AI what you want to produce.

How do AI prompts work?

From basic direction to complex multipart concepts, the depth and detail of your prompt will directly influence the quality and relevance of the AI’s output. The better the prompt, the better the result.

Types of prompts

There aren’t really different types of prompts. The best prompts are clear and instructional, designed to achieve various outcomes based on your needs.

Here are some common prompts:

• Descriptive prompts give detailed descriptions of what you need, focusing on content or information, like saying, “Write a summary of the novel 1984.”
• Instructional prompts, on the other hand, focus on directing the task or format, such as asking, “Generate a formal email requesting time off.”
• Creative prompts inspire imaginative or artistic responses, like asking, “Create a surreal landscape of a floating city above a tranquil sea at sunset, with buildings made of glass and light reflecting in a kaleidoscope of colors.”
• Clarifying prompts aim to gather more details or simplify concepts, such as requesting, “Explain the concept of blockchain in simple terms.”

Each type of prompt serves a different purpose and can be used to tailor the AI’s responses to fit your needs.

Prompt chaining

Prompt chaining is a technique used in generative AI, particularly with language models like ChatGPT. In this approach, multiple prompts are linked together to create a more refined answer. It breaks down a task into smaller steps, where each prompt builds on the previous response.

Example of prompt chaining

• Prompt 1. "Create an outline for a blog post about the benefits of drinking tea every day."
• Prompt 2. "Expand on the first point of the outline with examples and detailed explanations."
• Prompt 3. "Rewrite the conclusion to be more engaging and provide a call to action."

By chaining prompts, you can achieve a structured, thorough, and polished piece of content, using the AI’s strengths to improve the content in stages.

Our experts have seen it all

Sign up for Asurion Home+ and get 24/7 tech support and device protection—all in one plan.

Learn more

What are some effective prompting techniques for generative AI?

To get the best results, use clear and specific prompts. For example, instead of typing “write an essay on renewable energy,” try “write a 500-word essay on the benefits of renewable energy.” The more details you provide, the more accurate and useful the output will be.

No, you don’t need a creative writing degree or extensive knowledge of multiple language models to use this kind of AI. You just need to know some simple prompting guidelines and the appropriate directions to give. We’ll go over both.

Setting chat prompt guidelines

When interacting with a generative AI like ChatGPT or Claude®, crafting well-detailed prompts can make all the difference in getting useful and relevant responses. Here are some guidelines to consider when setting your chat prompts:

1. Be clear and specific. The more detailed your prompt, the more likely you will get a response that matches your needs. Instead of asking a vague question, provide context, specify your desired format, and include key details.
2. Use natural language. Phrase your prompts in a conversational tone. Generative AI tools are trained on natural language, so using everyday speech patterns can help the AI understand your request more accurately.
3. Set the scope. Clearly define the scope of your query. Indicate whether you want a brief summary, an in-depth analysis, or step-by-step instructions.
4. Provide examples. If you want a particular style or format, include examples in your prompt. This can guide the AI in producing content that aligns with your expectations.
5. Use follow-up questions. To refine the AI’s response, ask follow-up questions or provide feedback. This iterative process helps the AI learn what you’re looking for, leading to more tailored answers.
6. Test different approaches. Don’t be afraid to experiment with different phrasing or angles. Slight adjustments in framing your prompt can lead to vastly different—and often more useful—responses.

What are some examples of effective and ineffective prompts?

• Effective Prompt. "Generate a short story about a detective solving a mystery in a futuristic city that’s between 800 and 1000 words."
• Ineffective Prompt. "Tell me a story."

The first prompt gives context, a theme, and specific details, while the second one is too vague, which may lead to less focused results.

Let’s dig a little deeper. Here are some more examples of good AI prompting.

1. Clear and Specific
   • Good. “Can you provide a detailed summary of the key benefits of using generative AI for content creation, including examples of tools and their applications?”
   • Bad. “Tell me about AI.”
2. Contextual and Detailed
   • Good. “I’m planning a marketing campaign and need a list of creative ideas for engaging social media posts. Could you provide three unique ideas suitable for a tech startup targeting young professionals?”
   • Bad. “Give me some social media ideas.”
3. Natural Language and Conversational
   • Good. “I’m looking for a quick recipe for a gluten-free vegan dessert that can be made in under 30 minutes. Can you suggest something that fits these criteria?”
   • Bad. “Gluten-free vegan dessert recipe.”
4. Scope and Format
   • Good. “Can you explain ChatGPT-3 and ChatGPT-4 differences in a comparison table? Include details on their capabilities and improvements.”
   • Bad. “What are the differences between ChatGPT versions?”
5. Providing Examples
   • Good: “I need a professional email template to request a meeting with a potential client. Could you provide a polite, formal template including introduction, purpose, and availability sections?”
   • Bad. “Write an email to a client.”

What are common mistakes to avoid when prompting generative AI?

• Being too vague. Lack of specificity can lead to generic or irrelevant outputs.
• Overloading with details. Too many specifics can confuse the AI and lead to mixed results.
• Overlooking context. Make sure your prompts are directly related to the content you want to generate, as irrelevant or vague prompts can lead to off-topic or inaccurate results.
• Expecting perfection. As useful as the tool can be, it’s an algorithm that’s constantly learning. Don’t assume the first response will be perfect or fully accurate.