Note: This post's footer has had an April 2024 update for paid subscribers.

A frequent question I encountered during the AI-powered agent’s workshops was, “Which version or large language model should we use?”

I’ve made a Google Bard vs. ChatGPT comparison before (the listing copy smackdown), and I haven’t made it a secret that I’m happy right now to hand Open AI $20 of my hard-earned per month for ChatGPT 4 so that I can take advantage of extended idea generation and plugins that extend ChatGPT’s capabilities.

To date, I haven’t been able to test new models like Anthropic’s Claude as it’s not available in Australia (or Europe) right now; you can only use it from the US or the UK.

But there is a workaround if you want to go down that rabbit hole and test different prompts on different models to see which performs best for you.

Tool of the day

So, the hack here is to use POE (which stands for Platform for Open Exploration), which Quora recently released from invite-only Beta.

Here’s a quick video comparing the listing description task on the various models.

Happy hunting! 🚀

April 2024 Update - for premium subscribers 🤩

Subscribe to Premium to read the rest.