• 100 School
  • Posts
  • đź’Ż Digest: Stop Overthinking AI

đź’Ż Digest: Stop Overthinking AI

This week: 3 live AI sessions to add to your calendar, the power of screenshots, and the meme that sum it all up

Author

Max Haining
November 09, 2025

Hey, it's Max đź‘‹

Harold posted this meme last week which I found funny but also uncomfortably accurate. At work, most people are still figuring out AI by themselves.

Half the people in your team are testing ChatGPT in a second browser tab, learning tricks off YouTube, and maybe sharing a screenshot in Slack if they’re feeling brave. The other half are pretending they totally get it already.

But nobody’s actually learning together.

90% of employees use AI tools in their personal lives, while only 40% have access through official company channels. So everyone’s out here solo-learning, guessing, copy-pasting prompts from strangers, and calling it strategy.

And it’s not just employees. A Forbes piece this week said something that’s been true for a while: a lot of leaders already suspect their teams know more about AI than they do which, if you’ve ever led a team through a tech shift, is a terrifying sentence.

As one CEO put it, the real issue isn't a lack of tools, but a "leadership readiness" gap where leaders are asked to set strategy for tech they haven't mastered themselves. If they stay quiet, the whole team stays quiet. This is at the core of why 95% of companies report no measurable business impact from their AI investments.

Middle managers, in particular, are stuck in the weirdest position. Fortune called them “the missing link” in AI adoption: too senior to experiment openly, too busy to learn deeply, and too human to admit they’re winging it.

So what happens instead is people experiment quietly, leadership waits for certainty, and everyone ends up “building alone.”

So… let’s fix that ⚡

Next week, Harold and Ciara are running a series of free, short, live sessions for anyone who’s tired of solo learning.

Here’s the lineup:

Something to try this week âś… 

Speaking of building with others, Ciara dropped something simple but effective on Twitter:

And she’s right. Instead of writing a long paragraph explaining what went wrong, you can just snap a screenshot, drop it into your chat, and let the model analyze the visual context.

Developers have been doing it for a while. GitHub Copilot Labs, Cursor’s Plan mode (which Ben mentioned in Ciara’s replies) and even Claude’s image-input features are built around that exact idea. But even more and more non-technical teams are realizing the same thing: screenshots speed up collaboration because they remove guesswork.

So next time something’s confusing, don’t write a wall of text. Screenshot it, drop it in your AI tool, and see how much faster the conversation moves.

Before you go ✌️

What about you? have you tried using screenshots to explain stuff faster to AI? did it actually help? tell me what you’ve tried, always love hearing how these small shifts actually play out in real workflows.

Cheers!

Max đź‘‹

P.S. Want to make your team & company AI-first? Let us help here.