Benedict Evans used two slides and a handful of lines to visualize the biggest question in tech right now. In today’s edition, we’re dissecting his strategy and laying out four rules to simplify a big idea
Benedict Evans publishes twice-yearly presentations on macro tech trends that are famously minimal—no visual or verbal clutter. In his most recent deck, AI Eats the World, two slides stand out.
The opening slide shows five platform shifts side by side, each represented as an S curve—the universal shape of how technology gets adopted. Slow start, rapid growth, then a plateau as it becomes infrastructure. Mainframes. PCs. Web. Smartphones. Four complete arcs in gray. Generative AI in red, dashed, still in progress.
Five slides later he returns to the same visual. This time the AI splits into a second possible direction with a question mark:
He could have used projections, market data, analyst forecasts. But in a story FULL of data, he chose a conceptual, data-free visual instead.
The result is two slides that say more than most decks say in fifty.
Four rules to simplify a big idea
Capturing a big idea in a minimal visual is one of those things that's easy to appreciate and hard to do. We've all sat in presentations where the point was buried under layers of data, bullet points, and extraneous graphics. And we've also all been on the other side of that—so deep in our own material that we can't see what to cut.
Evans' slides are a useful reminder of what's possible when you 1. Identify the only thing they need to understand from each slide and 2. Zoom out far enough to see the visual pattern that makes that obvious.
His framing has two parts. The first: AI is the next major platform shift in a long line of them. The second: nobody agrees on how big. One view says it follows the same adoption arc as every platform before it. Another says it's something categorically different—in his words "less like a new device or network, more like electricity or fire. Something that fundamentally changes humanity."
So how do you take something that complex and make it land in two slides? Here are four things you have to believe and do to get there:
1. Believe that complexity obscures insight (the mindset)
If you really believe that complexity obscures insight, it changes what you think your job is. You understand that your job is not to transfer everything you know. It is to create the conditions for understanding. That is a fundamentally different task. This is the starting point from which you can see
2. Use the sharpie filter (the process)
Software is a high-resolution environment. It forces you to worry about details that don't matter yet. A thick marker on a small sticky note is the opposite. There isn't room for layers of data or complicated visuals. There's only room for the core relationship. Start by asking: is this thing bigger than that thing? Is this thing causing that thing? Is this thing coming after that thing? Pick the simplest shape that answers the question a circle, an arrow, a line going up. You're not making art. You're figuring out what you're trying to say. If you can't sketch it in three seconds, you don't have a visual problem, you have a clarity problem.
3. Establish a rhythm to highlight the break (the technique)
When one thing looks different from everything else in a set, you can't help but look at it. That’s the Isolation Effect. Your brain is wired to treat the "odd one out" as the most important thing on the page. To use it, you need two things:
A Set. A group of elements that look identical. This establishes the "background."
A Break. One single element that violates that pattern.
The contrast does the work. Because you’ve established a rule with the first few elements, the one that breaks the rule becomes the only thing that matters. You don't have to explain that it's important; the visual gap between the pattern and the outlier makes the point for you.
4. Build the idea piece by piece (the delivery)
When you put everything on the screen at once, you’re basically giving the audience a map of the whole destination while you're still trying to explain the first turn. They aren't necessarily ignoring you, but they are processing the "solution" at their own pace instead of yours.
The real value of showing things in stages is that it lets you build an argument. You establish a "normal" or a baseline that everyone agrees with. Then, when you click and show the disruption, the change stands out because it has something to react against.
Wrap-up
The root problem of complicated slide visuals is often that they start with the data they have instead of the idea they're trying to land. So we end up building slides that reflect our research process rather than our argument. This is an incredibly easy mistake to make especially when you're deep in your own material. What Evans reminds us is that the job isn't to transfer everything you know, it's to create the conditions for understanding.
Simplifying the visual is hard because it means accepting that the work you did to get there doesn't need to be visible in the final answer.
P.S. Evans' superpower isn't his research or his data. It's his strategic judgment about what to say and how to say it. As AI continues to reshape entire areas of human expertise, the ability to make people understand, feel, and act on a complex idea remains a uniquely human advantage.

