Generative AI, also called large language models (LLMs) or GPTs, is stirring a lot of hype and excitement. They are definitely adept at impressing people, able to produce natural language and impressive pictures within seconds.
But they can do more, and if used wisely, they can save you time and effort. Let’s take a look at several ways you can use different AIs to save your time and help you achieve higher quality for your work tasks.
These examples will work on most of the popular generative AI services, such as ChatGPT, Claude, Perplexity, and Copilot.
Tip 1: Practice some caution
It’s important to appreciate that LLMs may be very powerful and impressive, but they are still machines. No matter how much it looks like an LLM can think for itself, it does not. It simply produces what it thinks are accurate results, even if those results are entirely wrong.
You, the human, are the difference, so always check and edit the output to ensure you are pleased with the results. The LLM helps you produce better results, but it can’t replace you.
Tip 2: Create context
Some people see generative AI/LLMs as little more than a search engine. However, while they are useful for searching information, there is a big difference. A conventional search engine works on basic queries—for example: “Highly rated business laptops”. But with an LLM, you can go much deeper by setting the context—for example, “I want a highly-rated laptop that costs R15,000 or less. It should have a long battery life. It should be very portable.” This context helps the LLM understand what you expect from it.
You can use context in many ways, such as “Find information only from research reports” or “I am explaining AI to someone who knows nothing about computers. Write an explanation that will make sense to them.” Setting context helps the AI focus and makes it easier to get useful responses when you give follow-up instructions or questions. You can also use this method in other ways, such as “Create a list of overnight courier companies in my city, find information on their average prices, and create a list of their website addresses.” (Here are tips to improve your AI prompts.)
Tip 3: Check your work
One problem with spending time working on something like a report, article, or letter is that you grow familiar with the text, and it becomes harder to spot mistakes or shortcomings. A way around that issue is to show the work to someone else and get their feedback. But that’s not always feasible.
Fortunately, you can do the same with an LLM. First, set the context: “I am writing an article on AI tips to help people be more productive. I am listing the tips in a helpful way with brief and high-level explanations that are not technical. The goal is to help readers quickly find tips that improve their interactions with LLMs. Take a look at the following article and provide feedback on it.” Then, paste your copy and look at the feedback.
(An important consideration: some LLMs will retain the information they receive and could surface it somewhere else, so remove any sensitive or confidential information before sharing it with a public AI service.)
Tip 4: Change writing tone or focus
You’re writing a formal-sounding letter, perhaps to an important client or an institution like a bank. Before you know it, you’re spending more time on how to write the letter than on its content. It can also be the opposite: you want the letter to be more casual, but your writing just feels so stiff and formal.
LLMs can help in several ways. You can tell the LLM to rewrite the letter to be more casual or formal. You can list all your main points and instruct the LLM to write a first draft in formal or casual tones. Another method is to define the audience: “This letter is to tell people about the upcoming staff party, so make it friendly and exciting.” You can also use this tip to change the focus of the writing: “This letter is about the staff party, but it leads in with the time and date. Rewrite it to begin with the snacks and refreshments.”

Tip 5: Use voice mode
You can interact with LLMs by typing in your requests. But several allow you to talk to them, and a few will even talk back! These features are incredibly useful when you want to give more elaborate instructions or enjoy hearing feedback to your requests. Several LLMs, such as Claude, Gemini, and Copilot, have these features on their mobile apps.
However, ChatGPT also offers both features in its web chat. On the ChatGPT website, you can click on the microphone icon and speak through your computer’s microphone. You don’t have to be precise—just talk naturally and let your thoughts flow. ChatGPT is proficient at following them. Alternatively, you can have a full-on conversation with ChatGPT by clicking on the voice mode button (next to the mic icon). Now the AI can talk back.
These features are very useful. For example, you can narrate to-do lists, your thoughts after a meeting, or a dump of ideas, and the AI will take notes, then structure the content for you.
Bonus Tip: Experiment
Don’t limit yourself to one LLM. Different LLMs have specific benefits and drawbacks. Perplexity is outstanding at search, Claude is excellent with language, and ChatGPT is a pretty solid generalist. Try out the same ideas on different LLMs, experiment with prompts, and see what is possible!
Generative AI isn’t about replacing human creativity—it’s about amplifying it. These tools can transform how you approach writing, research, and problem-solving, but only if you use them thoughtfully. The key is finding the right balance between leveraging AI’s capabilities and maintaining your judgment. With practice, these techniques will become second nature, saving you hours while improving work quality. Your future, more productive self will thank you.


