📊 ChatGPT Slides Adds Real-Time Collaboration

Good Morning! Here's what I have for you in today's newsletter:

OpenAI updated ChatGPT Slides so teammates can edit a deck together and leave comments without leaving ChatGPT. It’s available on Pro, Business, and Enterprise. To start, ask it to “create ChatGPT slides” using a conversation or a file you’ve uploaded.

Fewer “final” decks floating around, fewer comments lost in a separate chat. Your team can work in one copy and ask ChatGPT to handle the feedback. You can still make small adjustments yourself, so moving a logo doesn’t turn into a conversation.

MiniMax released M3.1 Flash, a model for agent tasks that runs three times faster than M3 at one-third of the cost. The company says an earlier version was Space Bunny, which topped global usage leaderboards with more than 200 trillion tokens.

One task can send an agent back to its model dozens of times. Every wait and token charge adds up. M3.1 Flash promises to make those repeated calls quicker and cheaper. If you’re using M3, try both on the same jobs to see whether the savings hold up without sacrificing results.

xAI added an iPhone Action Button shortcut that opens a voice chat with Grok Bot in one press. You’ll need a supported iPhone 15 Pro or later model and Grok Bot version 1.16.

Have a thought while walking out the door? Press the button and say it before you forget. There’s no need to open the app and find a chat, and unfinished tasks can continue there after you hang up. Check your phone’s compatibility and app version before setting up the shortcut.

Let Jev organize your notes so Opus 5.5 can find answers in them.

Jev classifies your notes by type, project, and priority, and Opus 5.5 reads the relevant files and answers with source filenames. You'll need Claude Code, Opus 5.5, and a TypeSafe API key.

Install the TypeSafe skill with claude plugin marketplace add typesafe-ai/skills, then claude plugin install typesafe@typesafe-ai.

Create your API key and set the environment variable expected by the skill. Keep it out of chat. Start Claude Code in ~/second-brain with inbox and notes folders, select Opus 5.5, and run a test call.

Ask for a labels.md containing note types, three to eight projects, and five clearly defined priority levels. Include a “none or unclear” project option.

Use Choice for categories, Score for priority, and Noul for long-term usefulness.

Have Claude build a filer that sends inbox notes to Jev, records the labels, and proposes destinations under notes/[type].

Start with a dry run. Review the choices before moving files, leave uncertain notes in the inbox, and keep archive candidates for approval.

Ask which open tasks belong to a project, sorted by priority. Tell Opus 5.5 to read the relevant notes and cite each file.

Open a couple of sources to check the answer.

Reassess older tasks using today's date, deadlines, and recorded status. Don't archive tasks just because they're old.

Ask for weekly.md covering new notes, the top five open tasks, inbox items, and proposed changes. Review it before approving any archive moves.

P.S. Want me to coach you 1-on-1 to build your own AI Second Brain that knows your business, projects, goals, and expertise, so you can automate hours of work, and have a personal AI strategist helping you grow your business or accelerate your career? Reply “SECOND BRAIN” to this email.

OpenAI improved its models for building and migrating software written in COBOL, Fortran, and Visual Basic. These languages still power banking, government, and business systems. The company says understanding decades of business logic is the priority, and it’s asking teams which languages and workflows need attention next.

Inception released Mercury 3, a diffusion language model that generates more than 1,000 tokens per second and supports a 1 million token context window. The company says it beats GPT-6 Luna and Claude Haiku 5.5 on several benchmarks. Access it through the Inception API, Baseten, or OpenRouter.

Google added co-creation for Slide Decks and Video Overviews in Gemini Notebook for Pro and Ultra subscribers. Choose the topic, format, length, language, and visual style, then adjust the slide outline or video producer notes before generation starts.

🎬 FLUX 3 Max: Launched on fal as a faster version of Black Forest Labs’ FLUX 3, producing a 10-second video in about 12 seconds.

🧠 EverOS Cloud: Gives agents memory across sessions through an API, turning previous tasks into Cases and Skills they can use again.

🧪 Together Post-Training: Customize open models you own through supervised fine-tuning or reinforcement learning, using your data, rewards, and training loop.

📣 Okara: Released CMO v4 with agents that learn your writing style, LinkedIn influencer campaigns, and MCP connections to services including Stripe and PostHog. 

Thanks for making it to the end! I put my heart into every email I send, I hope you are enjoying it. Let me know your thoughts so I can make the next one even better!

See you tomorrow :)

- Dr. Alvaro Cintas

✓ Full archive of premium guides with ready-to-use prompts

✓ Structured AI courses (step-by-step, start-to-finish)

✓ Every upcoming premium tutorial

✓ $500+ worth of AI tools