Pick the Right LLM with Parallel Prompting

Organise Your Digital Brain with Chat... The Best Prompts Are AI-Generated Pro...

Pick the Right LLM with Parallel Prompting

‘The new workflow for me is I think with AI and work with my colleagues.’ – Satya Nadella, CEO of Microsoft

LLMs are changing fast. What worked well last week might be garbage to use today.

If you’re still using just one AI model for everything, it’s like playing a round of golf with just one club. You need to be able to select the best tool for the job.

golf club

This 3-minute tutorial shows you how to test multiple LLMs—side-by-side, in real time—so you can instantly see which one’s best for your task right now:

Watch: Best Way to Compare LLMs in 2025

You’ll learn:

How to run the same prompt through 3–5 models at onceWhat to look for in their responsesWhy you should be auditioning your AIs weekly

The best AI operators aren’t the most technically proficient, but they are the most adaptable. The tools keep changing, so we need systems that flex by constantly evaluating new tools, with feedback loops filled with the meta-skill of learning how to learn.

LLMs are changing so fast, we must dynamically adapt to keep up.

Which one is best today? Only testing will tell. Test your favourite LLM against other options, because the tools are always changing.

The apps on your phone are updating all the time. Compare your phone now to your phone 3 years ago. You ever pick up an old phone? The interface is off, the functions aren’t where you want them to be – we have to get used to these changes, or we stagnate.

If you’re using GenAI as a partner, you’re doing one of three things: Thinking, Making, or Doing.

One LLM is going to be better at helping you think, while another will be better at making things, and a third can excel at getting things done.

Stop hoping you picked the right tool. Start testing instead.

Parallel Prompting Tutorial

With some clever copy-pasting, you can let your tools audition for the job.

Watch Now (4:41)

New AI News This WeekOpenAI rolled back a recent update that was ‘sycophantic and annoying’ – because users treat thumbs-up like a likeMeta launched a standalone AI app built on Llama 4, integrating your WhatsApp, Instagram, and Facebook content into memoryDeepSeek released an advanced math-focused model, designed for proving mathematical theorems