The question that matters
There’s a version of this conversation that gets written constantly: AI will generate your deck, chat interfaces will replace slides, PowerPoint is dying.
We don’t think that framing is useful. It’s too coarse, and it’s probably wrong.
The question we actually want to answer is more specific: at what capability threshold does AI start changing how professional teams work with slides? What does the workflow look like at 70% AI capability? At 85%? At 95%?
We’ve spent the past months watching consultants, analysts, and strategy teams work - and building tools for them. We have a view on where this goes. Here’s how we think it unfolds.
How teams work today
In most strategy teams, there’s a significant gap between analytical work and slide output.
The thinking happens in models, documents, and conversations. The slides are a translation of that thinking into something the client can engage with in a 30-minute meeting. That translation is currently done by humans - often the most junior people on the team - and it consumes a disproportionate amount of time.
The numbers vary by context. A due diligence deck might take 2-3 days to build and format. A weekly update deck: 4-6 hours. A proposal: a full day’s work for a team of two, much of it spent on formatting and layout rather than argument.
The senior person has the judgment. The junior person has the file open. The feedback loop between them is slow and expensive: a structural change in thinking means going back into the production layer, reformatting, realigning, rebuilding charts.
This is the problem AI is now positioned to solve. But not all at once.
What changes at each threshold
At 70% capability: production becomes automatable for most slides.
A junior analyst currently spends a meaningful fraction of their time on tasks that require no judgment: applying brand colors to charts, adjusting text box sizes, ensuring layout consistency across a 50-slide deck. At this threshold, an AI handles most of that - reliably, on demand, in seconds.
This doesn’t make the analyst redundant. It changes the nature of their work. Less formatting, more reviewing. Less assembly, more judgment. The slide becomes the output of a conversation rather than a manual production process.
The skill premium shifts too. Being fast in PowerPoint matters less. Being able to judge what a slide should say - what to put on the page, what to cut, what hierarchy to use - matters more. That’s the skill junior consultants develop faster when they’re not spending half their time adjusting margins.
At 85% capability: the feedback loop collapses.
At this threshold, complex elements - charts, tables, multi-column layouts - become reliably editable. The gap between analysis and slide output narrows to near-zero on most slide types.
This changes the relationship between senior and junior team members. Currently, a partner who wants to restructure a narrative - “move section three to the front, add a bridging slide, rebuild the waterfall with updated assumptions” - is asking for hours of rework. At 85% capability, that feedback costs minutes. More rounds of iteration become possible in the same time.
The output gets better as a result. Not because the AI is smarter than the analyst, but because the constraint was never analytical - it was mechanical. Removing it lets the team cycle faster and raise the quality ceiling.
It also changes what associates and engagement managers focus on. With production out of the way, the question becomes: is the argument right? Is the story tight? Does slide 7 actually answer the question the client will ask? That’s where experienced professionals should be spending their time anyway.
At 95%+: the format itself may evolve.
This is more speculative. But at near-perfect capability, the deck becomes dynamically generatable from structured inputs: a narrative outline, a data model, a set of brand constraints. The static file might give way to a presentation layer that regenerates from source data rather than being assembled manually each time.
We’re uncertain this is the right direction. The static deck exists for a reason - it’s a shareable artifact, a record of a specific argument at a specific moment in a client engagement. That’s valuable. But the format was partly shaped by the constraints of manual production. When those constraints disappear, the format is free to evolve.
Whether it does will depend on what clients and partners actually want from a deliverable. Our instinct is that the deck persists - but the process of producing it becomes unrecognizable.
What doesn’t change
We want to be precise about what AI doesn’t touch in this picture.
The analytical work doesn’t change. Knowing which metrics matter for a specific due diligence, how to frame a recommendation for a client with a particular risk profile, which slide to cut when you’re over time - none of this gets easier because production gets faster.
The client relationship doesn’t change. A consulting partner’s value was never in their formatting ability. It’s in their judgment, their experience, and the trust they’ve built over years of engagements.
What changes is the ratio of time spent thinking versus producing. For many teams today, that ratio is badly skewed. AI pushes it back toward where it should be.
Our bet
We’re building Verso on a specific hypothesis: the tools that become embedded in professional workflows will be the ones that work within existing files, not around them.
Every major firm has a slide template that encodes years of brand investment and client expectations. An AI that ignores it generates output that gets thrown out. An AI that understands it becomes part of the workflow.
The slide deck isn’t going away. The question is how it gets produced. We think the answer - within a few years - is mostly AI for production, and entirely humans for the thinking behind it.
That gap is closing faster than most people expect.
