AI tools can now create complete pieces of work from a short request. But a request such as “turn this report into a presentation” contains many smaller decisions. The AI must decide what matters, what the audience already knows, which point should come first, what evidence belongs on each slide, and what can be left out.
Even a detailed prompt may not solve this, because the quality of the result also depends on how much work you ask AI to complete before you stop and inspect it.
One useful skill to develop is choosing the right level of granularity. In plain language, that means deciding how large each piece of work should be. Make the pieces too small and you’ll spend the whole time directing the tool. Make them too large and you’ll receive a polished result with important decisions already buried inside it.
For a presentation, the first decision that shapes everything else is the argument: the main point the audience should leave with and the few ideas that lead them there. Once that’s correct, creating the slides becomes much easier.
In this article, I’ll show you how to give AI a report, notes, or other source material, make it propose the argument first, and stop for your approval before it builds the presentation.
Here’s what you’ll find inside:
A two-stage prompt you can use with your own material: AI proposes the main takeaway and supporting points, waits for your approval, and then creates the slide plan.
A final check before you build or present the deck: Make sure the argument survived, the evidence still supports it, and no slide is present only because the information happened to be in the source.
If you’re new to Zero to AI, the Start Here guide will help you choose what to try next.
Comments
Nothing yet. Say the first thing.
Sign in to join the conversation.