Subject Line: Roundup #3: Non-obvious prompting advice
Hey! It’s Stew with Type, the AI-first document editor.
In this month's roundup email, I wanted to share some prompting advice that may help you get much better writing support from most language models.
This email also contains a comparison of the Claude and GPT families of models and some news from Type HQ.
Let’s get into it!
Language models are like smart interns – bright, but in need of adult supervision. I've learned this the hard way, starting back in 2020 when I first started using the earliest GPT-3 model.
Since then, I've discovered a few prompting techniques that reliably help most AI models write and think better. You can read my full learnings here, but these are the key bits of advice:
If you want more detail on each, check out the full post here.
When using AI to write, you've got two popular families of models that you can work with: Claude and GPT. But which one should you use, and when?
We just published a detailed comparison of how Claude and GPT's power and speed models perform at various writing tasks. This short guide covers the relative strengths and weaknesses of each as it pertains to writing.
Here's the summary:
You can read the full comparison here.
Some news! We just raised a $2.8M seed round at Type. We'll be using this new funding to accelerate our mission to help every person and team express brilliant ideas.
This milestone wouldn't have been possible without the enthusiastic feedback and support from this community, so thank you. You should notice some major improvements in the product over the next few weeks and months, so stay tuned.
That's all for this week - thanks for reading!
Best,
Stew Fortier
Co-Founder & CEO
Type.ai
Join 50k+ doing their best work