https://www.oneusefulthing.org/p/an-opinionated-guide-to-which-ai
“GPT-4 is the most full-featured model. It has access to web browsing (so it is no longer stuck with just the knowledge it had at training), it is multimodal (which means it can “see” and listen to speech), it can create images, it can code and do data analysis, and it can talk back to you (at least in the phone app). There are lots of other nice features, like the option for a privacy mode that doesn’t share your data. It also has enough memory (its “context window”) that it can do pretty complex tasks”
Ethan Mollick suggests that you just use GPT-4. But not the free version. I’ve said before that I prefer Claude to any other LLMs but that’s in the free category and I’ve never experimented with GPT-4.
OpenAI has only just re-opened access to GPT-4 via ChatGPT Plus subscriptions (there had been an short-term pause on new subscriptions), so if you’re OK to pay for the service then that might be the simplest option. However, you can still experiment with this model for free through the Bing search engine (only if you use Creative or Precise mode). But I think that Microsoft adds other bits and pieces to the interface i.e. it’s not ‘pure’ GPT-4.
Microsoft has also just changed the branding around their AI services and it looks like Copilot is the name for GPT in Bing search. The branding on the page says “Copilot…with Bing” while the browser title says it’s “Bing Chat with GPT-4”.
Confused yet? Me too.
I will almost certainly start experimenting with GPT-4 from OpenAI (ChatGPT Plus). Given how useful I find Claude, which is slightly better than GPT-3.5, it’ll be interesting to see if ChatGPT Plus is worth the cost.