Death by PowerPoint: Why I Now Let AI Draw My Placemats
Published on
The Advisory Artifact
If you've spent any time in the corporate trenches, you know the weapon of choice: the PowerPoint presentation.
In my early days of Management Consulting, the default output for any complex problem was a slide deck so dense it possessed its own gravitational pull. We called it "death-by-PowerPoint." You would walk into a steering committee, fire up the projector, and watch the exact moment the light left your stakeholders' eyes as they tried to parse bullet point number fourteen on a slide detailing a target operating model.
But to actually get people to think - to truly understand a complex technical architecture or a strategic pivot - we had to ditch the projector entirely. We would drag people to whiteboards for interactive sessions, or we would spend days agonizing over what we called "placemats."
A placemat was a single, highly distilled A3 page of pure visual summary. It was an infographic designed to sit on the desk in front of a client, anchoring the conversation. And it worked miracles.
The "Thud Factor" and the Infinite AI Loop
This obsession with volume highlights a scenario from my early consulting days that I never quite understood. There was a prevailing, unspoken belief among many senior leaders that value was directly proportional to the physical weight of the deliverable. They wanted a massive, heavy document they could physically drop onto a client's desk - a phenomenon affectionately known as the "thud factor" - to prove just how much arduous work we had billed them for.
It was entirely counterproductive. A consulting document is almost always a temporary vehicle designed to help a client implement a change. By burying the actual, actionable insights inside hundreds of pages of indecipherable corporate filler, they completely destroyed its only reason for existing. It takes genuine, grueling effort to distill complexity. As the French mathematician Blaise Pascal famously apologised in a 1657 letter: "I have made this longer than usual because I have not had time to make it shorter."
It is highly amusing to consider how this plays out today. Those same leaders can now use an LLM to effortlessly bloat a simple strategy into a 500-page monolithic report in seconds. The profound irony? The client on the receiving end is immediately going to feed that massive document straight into their own LLM to summarise it back down into the five bullet points it should have been in the first place.
I do wonder how the "big heavy document" strategy is working out for them now... anyway.
The Psychology of the Placemat
There is a very literal science to why a single visual placemat succeeds where a 40-slide deck or 500-page document fails, and it comes down to how our brains are wired.
In the 1980s, educational psychologist John Sweller developed Cognitive Load Theory, which essentially proves that our working memory is a terrifyingly narrow bottleneck. When you hit someone with a wall of text while talking over it, their working memory simply overflows. The data is dropped before it ever reaches long-term storage.
Contrast this with Dual Coding Theory (Allan Paivio, 1971), which suggests that our brains process visual and verbal information through two separate, parallel channels. When you pair a simple, well-chosen visual analogy with a concise verbal explanation, you aren't just making it look pretty; you are effectively doubling the bandwidth of the brain's intake. Imagery creates spatial relationships and neural links instantly. It doesn't force the brain to parse complex syntax; it allows it to simply map the concept.
The Creative Bottleneck
So, if infographics and placemats are objectively better for human learning, why did we still default to the 40-slide deck?
Because making a good infographic is brutally hard
(It also lacks the aforementioned 'thud factor'... but let's set that point aside for now.)
Distilling complexity into simplicity takes vast amounts of time. In the old days, it meant taking a rough, ugly sketch, handing it over to a dedicated creative team, and waiting a week. You’d get back something that looked beautiful but almost certainly missed the nuanced technical reality of the project, leading to endless, frustrating iterations.
As the creator of this blog's wildly unhinged orange starfish, I have already thoroughly documented my own inability to bridge the gap between technical concept and visual execution, so you will understand my pain.
The Multi-Modal Shortcut
This is where the latest leap in AI image generation changes the game entirely.
We are no longer just asking models to "draw a cyberpunk cat." With modern multi-modal tools like Nano Banana 2, we can feed an AI a dense, 40-slide strategy deck or a sprawling 80-page requirements document, and ask it to synthesize the core narrative into a single visual infographic.
The AI utilizes its underlying language reasoning to understand the meaning of the text, and its image-generation capabilities to render the visual analogies required to explain it.
I’ve taken to generating these infographics myself for a variety of use cases now. What used to require a week of back-and-forth with a design agency can now be prototyped, generated, and iterated upon in a matter of minutes. I can ask the model to adjust the metaphor, simplify the flow, or change the visual hierarchy, and it updates instantly. If I want to switch from a cartoon to a sketch, or change a cooking analogy to a house building one, I can do so in moments. It is quite incredible.
Long documents and detailed presentations absolutely still have their place - they are the necessary reference material, the deep dive, the architectural blueprint. We can, and do, still reference them for that next level of information. But for that crucial initial cognitive download to get someone interested? Creating a visual anchor is no longer a luxury reserved for massive budgets and long timelines. It’s just another tool in the stack.