Every designer and developer I know is quietly building their own AI toolkit. Here’s mine, as it stands today.
My Everyday Toolbox
ChatGPT is where I start when I need to think something through. It’s good for research, pulling together context quickly, and working through ideas before I’ve fully formed them. I use it like a first-draft thinking partner.
Claude handles most of my writing and code. It understands nuance better than I expected, and it’s become genuinely useful for client-facing copy, long-form content, and anything where tone actually matters.
Nano Banana 2 (inside Google’s Gemini) is where I go for image generation. It’s Google’s latest image model and the results have improved significantly. What I like most is the editing side: you can shift the mood, adjust the angle, apply a style from a reference photo. It’s less about generating from nothing and more about having actual control.
That’s the core stack. But there are a few others on my radar worth exploring properly.
The ones I am experimenting with
Midjourney is still the benchmark for art-directed, stylised image generation. If you care about aesthetic control and a distinctive visual language, it’s worth the learning curve.
Perplexity does something useful in the research space: search with citations. It’s more trustworthy than a straight chat response when you need to verify facts quickly.
Runway is one to watch for video. If motion content becomes part of your offering (and I suspect it will for a lot of us), this is where the interesting work is happening.
Notion AI is worth a look if you’re already using Notion for project management. Having AI baked into your workspace, rather than switching between tools, changes the workflow more than you’d expect.
The honest take: the tools are moving faster than most of us can test them. What matters more than the specific apps is getting comfortable with the workflow of prompting, reviewing, and editing. That bit is a skill, regardless of which tool you’re inside.
More on that soon.
