Build Decks at the Speed of Thought with AI (Go From Notes to Narrative in 30 Minutes)

the problem: Decks are corporate horcruxes. You open PowerPoint to get a first draft going, and suddenly it’s 1:17 AM. The hours just vanished into endless formatting. Text boxes won’t align, everything keeps jumping around, and somehow there are 40 invisible layers you never created. 

Even if the content is solid, the formatting becomes a whole second project—one that quietly burns time. And time is money.

Yet companies still run everything through decks: sales, marketing strategies, and fundraising. And the same few people always get stuck rebuilding them from scratch.

These people don’t need inspiration. They need a deck that looks really f****** good without costing hours. 

the solution: Jon Noronha (Gamma co-founder + CPO) has watched 70M people create 400M pieces of content—decks, docs, microsites—at speeds that make traditional workflows look prehistoric. His thesis is simple:

Stop designing decks. Start orchestrating them.

the loop: Let your LLM build a robust outline based on your research and company docs. Then, hand it to Gamma to take over the heavy lifting on structure, layout, and iteration.

Here’s the play in four steps:

1. kill your anxiety: Starting a deck from zero is like licking an iceberg, hoping it turns into a sculpture. The blank page isn’t the enemy—inertia is.

Your fastest escape hatch is stupidly simple: prompt your LLM to build out an outline. Feed it relevant markdown files like: brand positioning, customer quotes, transcripts, internal docs, financials, and even screenshots. And make it ask clarifying questions before it builds anything to help curb hallucinations. Steal this:

2. feed Gamma: After beating up your outline, create a first pass of a deck. Click through these pages:

Open Gamma → Create New → Paste in Text

Paste in your outline and use the Special Instructions feature to force Gamma to preserve any text and data you don’t want estimated or changed. The program will build a fully fledged deck with charts, images, and all the context you gave it.

3. the last mile: Is it Fortune-50-board ready? Not yet.

Gamma’s Agent makes it pretty easy to update your entire deck at once. Just prompt it to either tighten the language, generate or cut out graphics, or use your brand fonts and colors. You can also prompt it to steal or mimic aspects directly from the template your org already has. Lastly, go through the deck with a fine-toothed comb and fix any AI-sounding text.

4. content farm: Once it’s approved + battle tested, you can scale and automate with Gamma’s API. Jon wired Gamma to HubSpot/Salesforce, and now his proposals build themselves.

On the user experience: Google Slides and Canva have nicer UX than PowerPoint, which hasn’t been updated since the Bush administration. But none of them reliably hit all three: the ability to use AI to generate structure, protect your exact wording, and update slides globally. Gamma is the leader here.

➿ Grab the full playbook for every prompt, trick, and automation that turns 12-hour deck pain into a 30-minute machine.

p.s. We compare Gamma to the brand-new NotebookLM Slides feature below. 👇

Runway dropped Gen-4.5 yesterday. The company says the model “achieves an unprecedented level of physical accuracy and visual precision… Objects move with realistic weight and momentum… Surfaces behave the way they would in the real world. Laws of physics can be both observed or ignored, depending on your vision.”

They also acknowledged its ceiling: failures in causal reasoning and object permanence—gaps they’re actively working to close.

Gen-4.5 is currently leading independent benchmarks ahead of Veo 3, Kling 2.5, and Sora 2 Pro. But it feels like every day there’s a new leader in the space.

We just dropped a playbook on turning text into viral videos with PJ Ace, arguably the best marketer using this tech today. You can slot Whisper Thunder in for Veo 3.1 and run his exact workflow.

eli5: Josh Kushner is thriving. Thrive Holdings is buying a portfolio of service companies—starting with accounting and IT services—and stacking them under one roof.

OpenAI is taking a small ownership stake to embed engineers within those companies and deploy frontier models directly into their day-to-day operations.

Thrive gets eng support; OpenAI gets long-term upside and real deployment feedback on how its models perform inside live workflows.

If OpenAI proves it can lift margins inside operationally heavy companies, every industry with repeatable workflows becomes an AI-upgrade market.

Alright, let’s address the elephant in the room.

NotebookLM rolled out its slide-deck tool last week. If you haven’t used NotebookLM, it’s Google’s workspace for research and synthesis: you load in sources—docs, PDFs, articles, YouTube links—and it reads everything. Then, it lets you chat against those sources. Its claim to fame is turning those sources into podcasts with virtual hosts.

Slides now extends that loop, using Gemini 3 Pro to generate graphic-heavy cards—each slide a single, fully designed visual with the copy locked in.

Google vs. Gamma: Gamma gives you full control over every word, box, and layout. Its agent can update the entire deck in one move, and its modular structure makes iteration easy.

If you want something visual, narrative-first, and closer to an “editorial” aesthetic, NotebookLM hits that lane.

We’re bullish on NotebookLM, not as a Gamma replacement, but as another storytelling tool. Gamma still leads for copy-heavy decks and fast, global iteration.

On the eve of ChatGPT’s third birthday, I saw a conversation on X. It sparked this thought: 

When output becomes infinite, average work loses its value. Once production is free, the real competition shifts from making things to choosing well. Judgment becomes an even scarcer resource.

ChatGPT (+ every LLM since its launch) gives everyone infinite output. Infinite blog posts. Infinite designs. Infinite code. Infinite decks. Infinite everything. But when creation becomes free and abundant, the cost “good enough” is worth exactly $0.

A 10x marketing professional recently reminded me about this…

the Soundcloud rapper effect: If you were online between 2016 and 2018, you remember the SoundCloud flood. Anyone with a laptop could upload music. The result wasn’t a creative renaissance—it was sameness at scale (unfortunately for my mom’s sanity, mostly mumble rap).

While the market flooded with mumble, outliers like Lil Uzi Vert and Juice WRLD blew up. Why? They had something no other SoundCloud rapper had: fresh taste + an undeniable point of view.

AI is recreating that pattern across the business world:

apply it: Become the curator-in-chief. Build teams around people with a point of view and good judgment—and bake that into your hiring process.

Prioritize asking talent to complete micro-projects instead of sending over their portfolio. With portfolios, half the time you’re seeing the polish of an entire team, and you don’t get to see how your prospect truly thinks and if they have undeniable taste.

As Syndrome in The Incredibles put it, “If we’re all super, no one will be.” When the tools make everyone “super,” the differentiator is the person who knows what actually deserves a spotlight, what looks different than everything else, what works best, and what will hit the hardest to audiences.