DALL·E 3 is an impressive image generator to say the least. Not only does it have the natural language prompting power of ChatGPT built-in, but its actual capabilities are seemingly unmatched. This is in stark contrast to image-gen models like Midjourney v5 and SDXL 1.0, which both promised better text capabilities but ultimately fell short. The only other comparable AI I’ve found in this regard is Ideogram.

Here are some examples from a quick chat with the DALL·E 3 conversation mode in ChatGPT:

But what’s the utility really of a tool like this? The first thing that comes to mind for me is logo design, since a logo is the part of every project that lends total creative freedom. With the aim of these AI tools to make creation more accessible to both those with and without the means/skills necessary to create art, I think creating a personal logo would be a great test for this. I can make logos, but the real question is should I when I can just ask an AI for ideas.

Let’s start with DALL·E 3:

Ideogram also gives comparable results: