- Everyday AI
- Posts
- Ep 790: How To Make Slides with AI: Hands on Comparison With the 9 Most Popular Options
Ep 790: How To Make Slides with AI: Hands on Comparison With the 9 Most Popular Options
Microsoft unveiled a new AI-focused operating system for hardware, Meta selling business AI agents and Google tests AI search opt-out for publishers and more.
👉 Subscribe Here | 🗣 Hire Us To Speak | 🤝 Partner with Us | 🤖 Grow with GenAI
Sup y’all! 👋
Been kinda a busy week in terms of releases and features.
What should we cover tomorrow?
What should we cover on tomorrow's show?🗳️ Vote to see LIVE results 🗳️ |
I work for you.
✌️
Jordan
Outsmart The Future
Today in Everyday AI
8 minute read
🎙 Daily Podcast Episode: AI can build slide decks faster than ever, but speed doesn’t guarantee accuracy. We tested 10 popular AI presentation tools to see which ones create the best decks and where they still fall short. Give today’s show a watch/read/listen.
🕵️♂️ Fresh Finds: Lovable is partnering with Google Cloud to scale AI app development, Perplexity unveiled new AI agent capabilities for Windows and hybrid computing, and Martin Scorsese is helping shape AI tools for filmmakers. And more. Read on for Fresh Finds.
🗞 Byte Sized Daily AI News: Microsoft unveiled a new AI-focused operating system for hardware, Meta selling business AI agents and Google tests AI search opt-out for publishers and more. Read on for Byte Sized News.
💪 Leverage AI: AI can build a deck in minutes, but speed doesn’t guarantee accuracy. The companies that win will use AI to create drafts faster while keeping humans focused on the story and facts. Keep reading for that!
↩️ Don’t miss out: Miss our last newsletter? We covered: Microsoft debuts MAI-Thinking-1 and Autopilot 'Scout' Agent, OpenAI releases Enterprise Codex Plugins, White House signs AI order and more. Check it here!
Ep 790: How To Make Slides with AI: Hands on Comparison With the 9 Most Popular Options
No matter your role, experience or industry, we all (mostly) waste hours a week doing the same thing: manually creating slides.
😫
Most business leaders miss the obvious: there’s dozens of AI tools and LLMs that make slides and you probably didn’t know.
Case in point: Claude, Copilot, ChatGPT and Gemini can all create slides by default that you can download or edit with natural language.
And that’s just the start.
So why are you and Deborah meeting 3 times a month about the quarterly KPI again?
Also on the pod today:
• Claude Design's default slide look 😑
• Gemini finally creates real slides 🎉
• Canva's AI slides: major fail 🚫
Listen on our site:
Subscribe and listen on your favorite podcast platform
Listen on:
Here’s our favorite AI finds from across the web:
New AI Tool Spotlight – Hermes Agent is The Agent That Grows With You, Uselink turns any AI-generated doc into a permanent, commentable URL, Forward is an AI forward-deployed engineer that integrates your API directly into your customers' codebases
Lovable and Google Cloud — Lovable is teaming up with Google Cloud to let anyone build secure, full-stack apps with Gemini AI at massive scale.
Martin Scorsese and AI — Legendary director Martin Scorsese is now advising Black Forest Labs, helping shape their visual AI tools for filmmakers.
Perplexity Hybrid AI — Perplexity just launched a hybrid AI system that automatically splits tasks between your device and the cloud for smarter, more private, and energy-efficient answers.
Perplexity Computer on Windows — Perplexity is bringing its Personal Computer agent to Windows, letting AI manage your apps and files locally.
Sam Altman Attending G7 — Sam Altman is heading to the G7 in France, where AI and tech leaders are making big moves for the country’s future.
Agilent and OpenAI — Agilent is teaming up with OpenAI and BCG to supercharge its AI capabilities across products and operations.
Github and Copilot — GitHub Copilot app is now the control center for managing multiple AI agents in your dev workflow.
1. Microsoft Discovery launches for all organizations at Build 🔍
Microsoft used Build to make Microsoft Discovery generally available, turning its agent-based R&D platform from preview project into a product for companies trying to manage complex scientific and engineering work.
The pitch is simple: instead of asking one chatbot for answers, teams can coordinate specialized agents, internal data, research evidence, simulations, and review steps in a workflow that can be tracked and repeated.
2. Microsoft unveils Solara for AI-agent devices 🤖
At its Build conference in San Francisco, Microsoft introduced Solara, an Android-based operating system meant for a new class of cloud-managed devices built around AI agents instead of traditional apps, according to Ynet Global.
The big shift is simple: Microsoft wants devices that can respond to natural-language commands and reshape their screens automatically, rather than forcing users through fixed menus, icons and app layouts.
3. Microsoft says AI-redesigned Majorana 2 chip puts quantum goal on 2029 track ⚛️
According to Reuters, Microsoft on Tuesday unveiled Majorana 2, a quantum computing chip redesigned with AI-developed lead-based materials, and said it now expects commercially useful quantum machines by 2029.
That timing puts Microsoft in a tighter race with IBM, Google, Amazon, and Chinese teams chasing machines that could tackle problems far beyond today’s computers.
4. Meta rolls out its WhatsApp AI business agent globally 🤖
Meta is taking its customer support bot, now called Meta Business Agent, worldwide on WhatsApp after nearly two years of testing in markets including India and Mexico.
The agent can answer customer questions, recommend products, book appointments, qualify leads, and hand chats to a human, with support also coming to Instagram DMs.
5. Google tests publisher controls for AI Search in the UK 🔭
Google announced Wednesday in a Keyword blog post that it is testing new Search Console controls letting a small group of UK publishers decide whether their content can appear in AI Overviews and AI Mode.
The move comes as Google pushes search toward chatbot-style answers, raising publisher concerns that AI summaries could reduce visits to the original websites that supplied the information. Sites that opt out will lose traffic and impressions from Google’s generative AI features, but Google says the choice will not affect their ranking in regular search results.
AI can build the deck before your team agrees what’s true.
That’s the new slide problem.
The work that used to eat hours in PowerPoint, Canva or Google Slides can now become a polished draft in minutes, but a pretty AI deck can still skip a key story, merge two separate points or invent the wrong narrative with CEO-ready confidence.
On today’s Everyday AI, we stress-tested 10 beginner-friendly ways to generate slide decks with AI: Claude Design, Claude, Gamma, Gemini, Google Slides, NotebookLM, ChatGPT, Microsoft Copilot, Canva and PowerPoint with ChatGPT.
The bigger value is workflow leverage. If your competitors can spin up personalized, branded and source-aware decks on demand while your team is still copy-pasting screenshots, you’re not just slower.
You’re running strategy through yesterday’s workflow.
Here’s the business read: get to draft fast, audit the facts faster and standardize the stack before slide chaos gets a shiny AI wrapper.
1. Treat pretty decks as suspicious 🔥
AI deck tools are now good enough to pass the eye test, which is exactly why they’re dangerous.
Claude Design looked polished but missed one of the seven core stories. Gamma had clean visuals but drifted into hallucinated storylines, Gemini missed two stories, Google Slides produced cleaner-looking generic mush and Canva went wandering into fully made-up territory.
That’s the executive risk hiding inside the convenience. A deck that looks almost done can quietly move bad information into a sales meeting, board update or internal strategy review before anyone realizes the confidence came from typography, not truth.
Try This
Build a five-minute deck QA test before picking a company standard.
Feed each tool the same trusted memo, customer recap or market update, then grade it on source fidelity, missing points, unwanted merges, hallucinations and editability. Design only gets scored after the facts survive.
2. Prompt for control, not magic ⚡
The simple prompt exposed the real divide between tools.
Some systems asked useful follow-up questions. Some rushed into generation.
Some made assumptions, called the wrong capability or produced a deck that looked fine until the structure fell apart.
This is where leaders need to stop treating slide AI like a magic button. The value comes from turning the deck request into an operating brief, because the model can’t protect the business context your team never gave it.
Try This
Use a reusable deck brief every time.
Give the tool the audience, business decision, source material, slide count, required sections, forbidden combinations, brand rules, tone, output format and whether it should ask clarifying questions before building. That one extra minute can save the 45-minute cleanup nobody wants to admit is still happening.
3. Standardize where decks finish 🚀
The winner wasn’t a single tool. Annoying, but useful.
The top tier from this run was Claude Design, the two ChatGPT paths and Microsoft Copilot, but each earns its spot for different reasons.
Claude Design can produce polished work, though the default look is already entering AI-template jail. ChatGPT created strong visual storytelling and speaker notes, PowerPoint with ChatGPT gave a clean editable finish and Microsoft Copilot made in-app editing feel practical.
NotebookLM had the grounding advantage because it worked from a provided source, but the visuals were uneven. The real move is matching the tool to the stage of work: source-grounded draft, factual audit, design finish, team edit and final approval.
Try This
Pick one default slide stack this week.
For example, draft from a trusted source, run a factual QA pass, finish in PowerPoint or your brand system, then require a human owner to approve the story before sharing. AI can collapse






Reply