Google Can't Turn an Outline Into Slides. That Tells You Everything.
Google owns the AI and the tool. The most obvious integration still doesn't work.
I had a presentation outline. Bullet points, section headers, speaker notes. The kind of thing you’d hand a designer and say “make this into slides.” I pasted it into Gemini. Google’s AI. It couldn’t do it. Gemini can generate one slide at a time through a chat sidebar. One. You type a prompt, it gives you a single slide with 4-5 bullet points and a stock image. Then you prompt again for the next one. And again. And again.
The company that owns both the AI and the presentation tool cannot turn a simple outline into a slide deck?.
The One-Slide-at-a-Time Problem
Here’s what Gemini actually offers in Google Slides: a chat panel on the right side of your screen. You type what you want. It generates a single slide using your current theme. That’s it. It cannot generate a full presentation. It cannot update existing text. It cannot add transitions between slides. It cannot read your outline and produce a structured deck. You need a $20/month Google AI Pro subscription just to access this limited functionality. The $7/month business plan gives you 5 Gemini prompts per day and no Slides access at all.
What I expected: paste outline, get deck. What I got: a chatbot that makes one slide per prompt and requires me to do the assembly work manually.
Everyone Else Already Built This
The absurd part: third-party tools already solve this exact problem. SlidesWizard lets you paste an outline and generates a complete presentation. Presentations.ai takes your outline and produces a finished deck in under a minute. Gamma lets you type a topic, generates a full outline, and builds the entire presentation. These are startups. Small teams. Limited resources. They built the obvious feature that Google, with its $1.9 trillion market cap and the team that built both Gemini and Slides, still hasn’t shipped.
The presentation software market is growing at 17.8% annually, on track to reach $15.76 billion by 2029. Over 60% of enterprises already use cloud-based presentation platforms. Over 500 million people use PowerPoint globally. This is not a niche use case. This is one of the most common knowledge-work tasks on the planet. And Google’s AI integration is a one-slide-at-a-time chatbot.
Why This Keeps Happening
This isn’t a Gemini problem. It’s an AI product problem. Every major platform is shipping AI as a chat sidebar. A text box that sits next to your work and waits for you to figure out what to ask it. Google Slides gets a chat panel. Google Docs gets a chat panel. Gmail gets a chat panel. The AI is always adjacent to the work, never embedded in it. The result: users still do the work. They just have a chatbot nearby while they do it.
The best AI products hide the AI entirely. Users don’t interact with a model. They paste an outline and get slides. They upload a photo and get the background removed. They type an address and get directions. The AI is invisible. The value is immediate. What Google shipped in Slides is the opposite of this. It’s an AI feature, not a solution to a user problem. The team built “Gemini in Slides” when they should have built “paste your outline, get your deck.” Those sound similar but they’re fundamentally different product decisions. One starts with the technology: “We have Gemini, let’s put it in Slides.” The other starts with the user: “People have outlines and need decks. Let’s make that instant.”
What “Just Works” Should Look Like
The interaction should be this simple: Open Google Slides. Click “New Presentation from Outline.” Paste your outline. Get a complete deck with your organization’s theme applied, one key idea per slide, speaker notes generated from your bullet points. No chat sidebar. No prompting. No one-slide-at-a-time assembly. No $20/month subscription for the privilege of talking to a chatbot that can’t do the obvious thing.
The AI reads your outline structure (headers become section dividers, bullets become slide content, sub-bullets become supporting points). It applies your existing theme. It produces a complete, editable deck. You refine from there. This is not technically difficult. Third-party tools with a fraction of Google’s resources already do it. The models are capable. The APIs exist. The only thing missing is the product decision to solve the user’s actual problem instead of shipping an “AI feature.”
The Product Thinking Gap
When the company that owns both the AI and the tool can’t do the obvious thing, the problem isn’t technical capacity. It’s product thinking. Google has the model. Google has the presentation tool. Google has the user’s files, themes, and organizational context. Google has every advantage. And yet a startup you’ve never heard of does a better job turning your outline into slides. This is what happens when AI teams ship features instead of solving problems. When the goal is “integrate Gemini into Workspace” instead of “make presentations instant.” When the metric is “AI feature launched” instead of “user problem eliminated.” The outline-to-deck problem will get solved. Probably by Google, eventually.
But the fact that it wasn’t the first thing they built tells you everything about how AI products are being prioritized right now.
Technology first. - Users second.
It should be the other way around.
---
Andy Dahley has 20+ years of product design experience at Google, Meta, and
other places. He writes about AI product design at dahley.com.
---