After nearly 1,000,000 prompts... We're trying something new
Nearly 1,000,000 prompts improved. Thousands of people around the world.
And one thing keeps coming up: Most people still don't know how to get started.
Not the tech crowd. Not the devs.
Or the PMs using claude as a second brain. 😅
The travel agent opening ChatGPT for the first time.
The teacher building course materials from scratch.
The financial advisor trying to automate his day.
The field marketing team building their first agent from scratch.
We got super obsessed with this gap.
So we're trying something new: AI Workshops for non-tech teams. 📣
Taking everything we learned from @Pretty Prompt.
and making it jargon-free.
Not for the people who already get it.
For everyone else (99% of the world)
Would absolutely love some thoughts from the community!
Replies
this is genuinely one of the most underrated problems in the space right now.
everyone's racing to ship vibe coding tools and no-code agents, but if you watch someone non-technical actually try to use them, it falls apart fast. my friends, family, they're not struggling with the tools, they're struggling with the mental model. and no amount of better UI fixes that.
what you're building bridges that gap in a way most people aren't even thinking about. rooting for this one (personally).
Pretty Prompt
@riya_pariyar Thanks so much! I'd love to get your feedback, based on what you see on your friends and family! Would you be open to a quick call? It would be super helpful! my email ilai (@) pretty-prompt.com
This feels right. The non-tech gap is less “write a better prompt” and more “know what to bring into the conversation.”
For workshops, I’d anchor around 3 habits:
1. Bring source material first: examples, constraints, current process, audience, bad outputs.
2. Ask AI to explain assumptions before generating.
3. Treat the first answer as a draft to inspect, not a result to trust.
That framing keeps it jargon-free and also prevents the most common disappointment: people expecting magic from an empty prompt.
Pretty Prompt
@jim_jeffers Great points, I'm collecting quite a lot of resources to try to distill the most important points, not just from what people say, but actually from what they do and how they use AI today. Always open for a chat if you're up for it!
Appreciate the invite. A pattern I’d be curious to compare notes on: what people actually do right before a “bad prompt.”
In workshops, that might be the highest-signal observation — are they missing source material, vocabulary, examples, constraints, or just confidence to iterate? Each one needs a different teaching moment, and only one of them is really a prompt-writing problem.