I’m really enjoying Anthropic’s Claude AI, specifically, Claude Cowork. It is more organized and visually outlines the process better than OpenAI’s ChatGPT. I like how it recaps what we finished.
The biggest limitation I found with Claude so far is that it is a lot more expensive than ChatGPT because it measures your use by brain power used by Claude AI according to tokens that burn fast with deep thinking such as coding and AI architect work. Once you use up all your tokens, you have to wait until your tokens reset by a certain date and time. Either that or you pay for more tokens.
I will keep using Claude for coding and AI architect work, but have ChatGPT assist me with my business and legal work since pursuant to the 2026 Stanford AI Index Report, both Claude and ChatGPT rank high in reasoning. Their ranking fluctuates depending on the type of reasoning. When it comes to legal reasoning, ChatGPT leads Claude by about 10 percentage points.
As far as user experience, Claude is more fun for me. It makes me want to work more. This is similar to how I enjoy using a Mac OS instead of a PC OS. The system makes my work feel like play. This is very important for me. I will keep this in mind as I build my law/business/psychology/science decision integrity AI app for my users (founders).
JON RIKI KARAMATSU 