AI tools have become very popular and used in the recent years. They offer different features which help people in day-by-day activities and make their tasks easier to complete. In my opinion, one particularly interesting feature is the AI image generator.
In my professional experiences, I have often created presentations, sometimes related to very specific topics. It was difficult to find topic-precise images to properly represent these subjects and situations. To address this, I have sometimes used different AI tools to generate specific images.
I have tried three AI tools for this purpose: ChatGPT (using DALL-E 3), Gemini (through Nano Banana) and Canva AI. Each of these AI tools has its own advantages. For instance, both ChatGPT and Gemini integrate communication directly in their model, making it easier to properly describe the prompts, specify picture features and details and request any adjustments. While working with these tools, I have noticed interesting differences. For example, many times ChatGPT asked for more detailed information before generating the image, in order to better understand the user’s idea. In contrast, Gemini begins generating images directly, without requesting additional details. However, in both cases it is possible to adjust elements, such as gradient, colours or any visual features.
Canva AI has its advantages too. For example, when requesting image generation, it generates multiple options simultaneously, allowing the user to select the favourite one. Furthermore, the user can directly integrate the chosen image in presentation or create a presentation starting with it.
Personally, I prefer using ChatGPT (DALL-E 3) for image generation because it offers continuous communication, and more accurate and high-quality results. However, my opinion could be a little biased since I am using the ChatGPT premium version, and thus have access to potentially better features and content, while for the other tools I only use the basic subscription. This is something that needs to be further analysed.
What do you think about AI image generation tools? Have you ever tried any?
