As presentation improves, AI has raised the bar for communication

While it is often contended that “AI has raised the floor on what good work looks like”, it hasn’t made people any better at explaining their thinking – if anything, that gap is becoming more obvious. Lyndon Nicholson, founder and CEO of Future Group, argues that as it becomes easier for anyone to produce polished presentations, reports and recommendations, the value is shifting away from producing the material and towards being able to actually stand behind it.

A few months ago, I watched a manager present a beautifully built case for a new initiative. The deck was sharp. The narrative flowed. The data visualisations looked like something an agency like ours would have charged five figures for. Then someone on the leadership team asked a simple question: “What happens if this doesn’t work?”.

The room went quiet, not because the question was hostile, but because the presenter had never actually had to sit with it before. The deck had done the thinking’s job so well that he’d never needed to.

I’ve spent twenty years building and selling presentation solutions into enterprise clients, and the last two watching AI tools rewire how that work gets done. The change has been real and often genuinely good. People who used to hand in cluttered, rambling decks now produce something that looks considered. Reports that took days now take an afternoon. The floor for polish has risen dramatically, for almost everyone, almost overnight.

But I keep noticing the same gap, time and again, client after client: the work has got better. The people presenting it haven’t got any more ready to defend it. For consultants and other professional advisers, whose value ultimately rests on the credibility of their thinking, that’s becoming an important distinction.

That’s not a small distinction. For most of my career, building a presentation was how you found out what you actually thought. You’d sit with a messy argument, try to write it down clearly, and discover the hole in your logic somewhere around slide six. The struggle was the point. It’s where the thinking happened, and it’s also where you rehearsed, without realising it, for the moment someone in the room would push back.

AI has taken that struggle away without taking away the moment someone pushes back.

I think that’s the real story hiding underneath the AI productivity coverage of the last two years. We’ve become very good, very quickly, at teaching people to use AI to produce polished work. Almost no one is teaching them what to do once they have to stand up, put the laptop away, and answer for it in real time.

The stakes of that gap are rising because the work itself looks so good now. A rough, obviously human-made deck used to come with built-in permission to stumble in the room. A flawless one sets a different expectation: if the thinking looks this sharp on the slide, people assume it’s this sharp in your head. When it isn’t, that gap shows more than it would have five years ago.

I hear a version of this from clients constantly now, and it usually shows up in three places: story, sameness and brand. AI-generated decks are good at producing individually reasonable slides and much worse at making them add up to anything. Point follows point with no real throughline, evidence gets dropped in without tying back to the argument it’s meant to support, and by the end there’s rarely a clear sense of what the audience is meant to conclude. A deck becomes a story when each slide earns the one after it, and that’s exactly the part a tool built for speed tends to skip.

Then there’s sameness. People can spot an AI-built deck surprisingly quickly, not because it looks bad, but because it often looks like every other AI-built deck: the same visual rhythm, the same handful of layouts, the same relentless level of polish. Brand has a similar problem. Colours and fonts are easy enough to get right, whilst tone of voice, imagery and the way a company actually makes an argument are much harder. The result can be technically on-brand and still not feel like the company that made it, and that becomes much harder to hide once someone starts asking questions.

Muscle memory

Audiences feel the difference immediately, even the ones who couldn’t name a single one of the three.

So, I’d push back on the current framing of the AI skills gap. Most professionals don’t lack technical fluency with these tools. Plenty of people can now produce something that looks impressive on a screen. What’s atrophied is the muscle for thinking on your feet, defending a position under real scrutiny, and earning trust with a room full of sceptical people, while everyone was busy getting better at producing the artefact.

That muscle doesn’t build itself, and it doesn’t build inside a chat window. It’s built through practice: being asked hard questions by people who know what a hard question looks like, and learning what it feels like to hold your ground, or change your mind, in front of an audience that’s watching closely.

It’s one of the reasons I’m involved in bringing Present to Succeed to London for the first time this November. The conference brings together people who spend their careers thinking about how ideas are communicated, stories are built and audiences are persuaded. What interests me most about bringing that community together now is how much the context has changed. The challenge is no longer simply how to make a better presentation. Increasingly, it’s how to make sure the person behind it is as convincing as the material on the screen.

I believe there are real, practical answers to a lot of what I’ve just described. Ways of setting up how a company plans, structures and reviews its presentations that would close much of this gap on the production side alone, if done properly. That’s active work I’m involved in right now, and I’ll have more to say about it before long. But even solved perfectly, it would only close half the problem. The harder half happens after the slides are finished, in the room, with nothing left to help you but your own thinking.

AI has raised the floor on what “good work” looks like on a page, but it hasn’t done anything for anyone’s ability to stand behind that work when it matters most. If anything, that ability matters more than it has in years, because it’s now one of the only things left that reliably separates people in the room.

The organisations that figure this out early, that invest as deliberately in judgement and presence as they have in AI tooling, will have a real advantage over those that assumed better decks meant better decisions. For consulting and professional services firms in particular, that matters: when everyone has access to tools capable of producing polished analysis, the ability to interrogate that analysis, defend it and earn a client’s confidence becomes part of the value you’re actually selling. The next skills gap will be closed by people who’ve practised standing up and defending their thinking, out loud, in front of someone who disagrees with them, not by anyone chasing a better prompt.