Ask someone who reviews decks for a living what's changed since 2023, and they won't mention AI first. They'll mention time. A pitch review that used to get twenty unhurried minutes in a room now gets four minutes on a call, sandwiched between two other meetings, with someone half-reading the deck before the call even starts. The slides haven't necessarily gotten worse. The window they're judged in has gotten brutally shorter.

That's the actual story of the last three years — not that presentations became "more AI-powered" or "more visual," though both are true. It's that the culture around how decks get made, reviewed, and trusted shifted underneath the format itself.

The review cycle got compressed, and nobody adjusted the deck for it

In my experience, most teams still build decks for the meeting length they used to have. Twelve slides, a slow build toward the ask, a Q&A section at the end for the questions that never quite land because the room's attention has already drifted. That structure made sense when a stakeholder gave you thirty focused minutes. It makes a lot less sense now that the same stakeholder is skimming your deck on a phone between two other calls before deciding whether the live meeting is even worth keeping on the calendar.

The teams that adapted didn't necessarily make shorter decks — some of the best ones I've seen got longer, with appendix-style backup slides for exactly this reason. What changed is where the argument lives. The first three slides now have to survive being read in isolation, without the presenter in the room to fill gaps verbally. If you've ever wondered why so many "what makes a pitch deck work" guides insist the opening has to state the problem in one sharp sentence rather than easing in with a mission statement, this compressed-attention shift is why — it's covered in more detail in how to build a pitch deck that actually gets funded, and the logic holds well beyond fundraising decks.

AI changed where the effort goes, not how much effort exists

Here's where most trend pieces oversimplify. The assumption is usually "AI made decks faster to build, so people spend less time on them." What I've actually watched happen is closer to a redistribution than a reduction. The time that used to go into formatting — aligning boxes, picking a chart type, writing a first-draft outline — has genuinely shrunk. But the time spent second-guessing whether a slide's argument actually holds up, because a generated draft can look polished and still be wrong, has grown to fill a lot of that gap.

The scale of the shift is real, even if the effect on total effort is murkier than the marketing suggests. McKinsey's Global Survey on AI found that regular generative AI use in at least one business function roughly doubled, from about a third of organizations in its 2023 survey to 65% by early 2024 — a pace of adoption most enterprise software never sees in twelve months. That's not a "some teams tried a chatbot" number. That's most of the market rewiring a workflow at once.

What it hasn't done is remove the part of the job that was never really about formatting to begin with — deciding what the deck is actually arguing, and whether the evidence supports it. That work got no faster. If anything, reviewing an AI-drafted argument for gaps takes a different, less mechanical kind of attention than building the argument from scratch used to.

The camera stayed, and it quietly rewrote what "good" looks like

Hybrid work was supposed to be a phase. Three years on, it's just how a large share of decks get delivered, and the culture has adjusted accordingly in ways that are easy to miss because they happened gradually rather than all at once. A slide density that read fine projected on a boardroom wall is often unreadable at meeting-window size on a laptop with three other tabs open. Presenters who built their delivery style around reading a room's body language have had to relearn what feedback even looks like when most of the room is a grid of small video tiles, half of them with cameras off.

A Decktopus survey of over 300 business professionals and students found that the share of presenters creating a deck at least once a week — a strong proxy for how embedded remote and hybrid presenting has become — is now well over half, up sharply from roughly monthly cadences a few years earlier. Presenting isn't an occasional formal event for a lot of teams anymore. It's a weekly operating rhythm, and weekly rhythms get judged by different standards than quarterly showpieces do — faster to produce, more forgiving of rough edges, but also more disposable if a deck doesn't earn attention in the first thirty seconds. The mechanics of that specific shift — what actually changes about pacing, eye contact, and pauses once the room becomes a screen — are worth a closer look in how to present on video calls without losing your audience.

Consulting-style rigor stopped being a consulting-only habit

This is the shift I'd flag as underrated in most "trends" write-ups, because it's not visible in any adoption statistic. Three years ago, structured frameworks like action-titled slides or a strict one-message-per-slide rule mostly lived inside strategy firms and the teams that had worked with them. Now I regularly see product managers, ops leads, and founders who've never set foot in a consulting firm applying the same discipline — writing slide titles as claims instead of labels, building the argument before opening a design tool at all.

That's a genuine improvement in a lot of decks. But it's worth being honest about the trade-off, because it doesn't show up in the trend pieces praising the shift: rigor borrowed without the underlying habit of testing the argument first tends to produce decks that look disciplined and still say very little. A tightly formatted slide with a confident-sounding action title isn't automatically a correct claim — it's just a well-dressed one. The structure McKinsey teams actually rely on, and where it tends to break for people who copy the surface without the process, is worth understanding properly rather than cargo-culted — how McKinsey designs their presentation slides goes into where that gap usually shows up.

What actually carried over from three years ago

Not everything shifted. The presenters who were good at reading a room in 2023 are, for the most part, still good at it now — the skill just had to be re-pointed at a different medium. And the core failure mode hasn't changed at all: a deck with no clear argument still fails, whether it's dense with AI-generated charts or hand-built the slow way. Faster tools didn't fix bad thinking. They just made bad thinking easier to dress up convincingly, which is arguably a worse problem than the one they solved.

If there's one practical takeaway from watching this shift happen, it's this: before you adopt whatever's changed this quarter — a new AI tool, a shorter deck length, a stricter slide template — ask whether it's actually solving the compression problem your audience is living with, or just adding a layer of polish on top of a review cycle that was already too rushed to notice the difference.