Member-only story
ChatGPT Canvas First Impressions and Comparison with Claude Artifacts — Video Included
Disclaimer, article made in collaboration with ChatGPT Canvas
Yesterday, OpenAI launched a new feature called ChatGPT Canvas, which turns ChatGPT into a more hands-on collaborator for both text and code documents. It’s not just a new user interface, but also a new model trained to excel at collaborating effectively — one that can decide when to make full-scale rewrites or targeted edits. This is something I’ve personally found challenging with tools like Claude Artifacts, which tend to always go for full rewrites. And that can be slow and… well, annoying. It can change parts you did not want.
Do not want a Medium account? Read here!
So, is Canvas any good? How do you judge? Well, you know me: I like to compare things. What is the closest thing? Claude Artifacts right? Today, we’re pitting ChatGPT Canvas against Claude Artifacts to see who truly shines when it comes to handling code and collaborating effectively. For the full comparison and some hands-on action, check out the video included below!
The Challenge
The best way to truly evaluate a tool is by applying it to a real-world problem. For today’s challenge, I chose a side project I’ve been working on — something that’s very close to me. For its landing page, I made a 3D animated logo I made earlier with AI, involving HTML, CSS, and JavaScript. It’s a cube that disassembles and reassembles, kind of like a digital magic trick. It works, but it’s a bit slow at times. So, my goal was to convert it to WebGL, the browser’s high-performance 3D rendering system akin to one used by games, to speed things up.
How well can these AI tools assist me in refactoring the project from HTML, CSS…
