5 creative ways to use your LLM

Important points

• You can use image generators like OpenAI's DALL-E to generate logos, wireframes, branding, or designs.
• LLM is great for generating boilerplate code and debugging existing code.
• LLM makes education more accessible than ever, and you can use LLM to generate flashcards, quizzes, and other learning resources.

Over the past 18 months, LLMs (Large-Scale Language Models) have made dramatic advances, and even better, are constantly improving. These improvements are sometimes locked behind paywalls, such as GPT-4 and DALL-E, but new media are now also available. LLM can now be used for everything from AI content generation, image processing, and video editing, as well as hosting software generation and debugging tools. The current rate of change within AI is so incredible that it's easy to miss out on opportunities that AI could potentially save you time, money, and effort. Let's take a look at how to get the most out of these tools.

1 Prototype your design quickly with LLM

Image generators like DALL-E are terrifyingly impressive

Text to image generators have come a long way and are constantly being improved, so there are many different ways to use them. You can either generate the entire image from scratch (though it may look a little “weird”) or you can use it to generate specific parts of the image, which you can then modify further. It's also perfect for generating mockups, wireframes, and prototypes, which can be further interactively adjusted using prompts.

Some image generators are free, but if you're a creative professional, you may already be paying or have access to one using Adobe Firefly. Free generators, such as those supported by his Copilot in Microsoft Designer, can be good, but they may be of low quality or have a distinct “AI” look. DALL-E is included in your ChatGPT Plus subscription and is one of the best. Google Gemini can generate images, albeit with restrictions in place following recent controversy, but as of this writing he is not available in the EEA, UK, and some other countries. Different generators often have a unique “feel” or aesthetic and have different features, so test a few or look at a few samples online before committing to a subscription. We recommend checking it out (or you can stick with the free option) like! )

2 Learn something new with an LLM

You can ask your LLM even your most embarrassing questions

In my view, the educational aspect of an LLM is an underrated benefit that it offers to those who are curious and want to ask questions. They are really good study companions (though not for writing essays for you, of course). You can ask LLMs about any topic, whether you're looking for a quick overview or digging into the details. Just ask why, how, and what causes what you're learning about. It's also great for creating flashcards, practice questions, and subject matter quizzes. For example, you can ask ChatGPT to generate a short quiz about a topic you're studying, but you can ask it to omit the answer. Then write your answer in the prompt box and let ChatGPT grade it for you.

Obviously, the LLM is not perfect. They are known to hallucinate and can confuse complex topics. For things that are important to your education and career, we recommend focusing on more academically reliable reference materials, such as textbooks and class slides.

3 Generate your logo and brand using LLM

LLM is perfect for finished products and inspiration

LLM can provide inspiration when generating custom images to represent your brand. Logos and brands inherently follow patterns. While every brand wants to look unique, they also want to be recognized in their field alongside other similar brands, so they tend to follow design trends. By explaining what you're looking for (e.g. logo, business card) and what your brand stands for (e.g. mission statement, value proposition), you can at least gather some new suggestions to consider.

It helps to be as detailed as possible here and to be clear about what you're looking for. This may include prompts asking you to select specific colors, text, or styles. Providing more context about your business can also be helpful. You can generate several different versions of an image for inspiration and let your graphic designer combine the elements of your choice to complete the final product. However, some of these image generators seem to really struggle with output containing text (strange spellings or hallucinatory phrases). That's why we recommend generating your logo or branding graphics, then adding text and additional details using a traditional image editor. or combine multiple elements.

Four Generate templates for other software using LLM

ChatGPT Plus has generated a Word document template that you can download

This really surprised me! ChatGPT Plus (on GPT-4) was able to generate an Excel document budget template and output both the document file and the Python used to generate it. This wasn't a very complex format, but you can easily enhance the output by making small changes to the prompts or adding more specific instructions.

ChatGPT is already great for writing cover letters and resume parts, but this one definitely takes it to the next level. We've already shown you how to use ChatGPT to improve your Excel, but generating complete templates makes it even more powerful and useful. Obviously, these templates can be limited, but if you're facing writer's block or just don't know how to create something quickly, these templates can be a great starting point. Masu. This will get you moving in the right direction (or maybe you won't like it, but you'll be motivated to make adjustments)!

Of course, this could also extend beyond the Office suite. LLMs like Gemini and ChatGPT are great at generating boilerplate fields for almost any app that accepts templates. Because of the way templates are generated using scripts, you are not limited by file type. As long as you can write the code to build the template, it should work.

Five Translate, debug, and generate code using LLM

LLM is amazingly good at debugging problems in your code, and even better at generating code.

This is clear. LLM is great for debugging code. It still tends to be confusing for large projects, and you can easily get sidetracked by hallucinating features and functionality that don't actually exist, but it's great for debugging. For best results when debugging your code using LLM, be sure to provide as much detail as possible, including specific error messages. However, be careful here, as LLM typically saves all prompts and data submitted for potentially training other models. Therefore, if you are in a professional environment or are using confidential information, please avoid passing that information on to your LLM.

Another less obvious use of LLM for code is code transformation. Let's say you wrote your API specification in Python, but you also want to support the Java SDK. LLM is great at this, and you can accomplish most of your goals by rewriting your app's endpoints in another language. This use case is likely to be less error-prone than other more novel coding purposes. However, you can also use LLM to generate boilerplate code. No more searching GitHub for information to help you start your project. LLM takes care of that for you, including configuring package managers, dependencies, and even basic CI.

Use your LLM imaginatively to save time and effort

LLM is becoming more powerful all the time, and like any powerful tool, it takes some time to learn how to use it properly. We've focused here on what you can do with some of the big LLMs that are readily available today (ChatGPT and Gemini), but as more specific tools are released, we'll be able to creatively expand on LLMs. There will be more features available. In the meantime, we can continue to enjoy leveraging the clear potential of these amazing tools.