|
|
|
|
This week, Google launched Gemini Spark, a personal AI agent that handles tasks for you in the background around the clock. Microsoft countered with seven new in-house AI models, led by its first flagship reasoning model.
We also have a free playbook on putting AI agents to work, six new tools worth a look, and three deep dives to round things out. Let's dive in.
In This Issue:
- 💥 Big News: Google launched Gemini Spark, a 24/7 personal AI agent, and Microsoft unveiled seven new MAI models
- ✨ Weekly Spotlight: A free playbook on putting AI agents to work as amplifiers of your team, not replacements
- 🔥 Trending Tools: Six new tools for fundraising, meetings, social, and shipping software faster
- 🛠️ Deep Dives: A closer look at Loom, Duolingo, and Stratup.ai
|
💥 Big News
Google just launched Gemini Spark, an AI agent that runs in the background and handles tasks for you around the clock, even when your phone and laptop are turned off. You give it a job and it works on its own, but always under your direction. You decide when to turn it on, and it checks with you before making any big moves.
Once you connect your apps, Spark can act across your digital life. Think weekly inbox recaps with a prioritized to-do list, organizing your Google Drive into a labeled spreadsheet, tracking down stray invoices and receipts, or doing the research to compare options and help you book a trip.
If you are on Google's AI Ultra plan, you can use it right now. 
Why it matters:
- Works in the background around the clock, even when your phone and laptop are off
- Runs tasks on its own but stays under your direction and checks before big actions
- Connects to your apps to handle email, files, bookings, and more
Read more →
Microsoft just announced seven new in-house AI models in its MAI family, unveiled at its Build 2026 conference. They span reasoning, coding, image generation, transcription, and voice, and they are built to work together across the tools you already use.
The headliner is MAI-Thinking-1, Microsoft's first flagship reasoning model, which works through harder problems step by step. Microsoft says it holds its own against top models on software engineering and math benchmarks, and that it was trained from scratch on clean, licensed data rather than copying from other companies' models.
It is a clear move toward Microsoft building its own AI in-house, and it comes with a new option called Frontier Tuning that lets businesses train a private version of a model on their own data. 
Why it matters:
- Seven new models spanning reasoning, coding, image, transcription, and voice
- MAI-Thinking-1 is Microsoft's first flagship reasoning model, trained from scratch
- Frontier Tuning lets businesses train a private version on their own data
Read more →
|
✨ Weekly Spotlight
AI agents amplify your team. They don't replace it.
This playbook from Mindstream and Futurepedia cuts through the hype and shows you how to put AI agents to work as amplifiers of your team, not replacements.
You'll get practical frameworks for 2026, from automating content workflows to qualifying leads and streamlining operations.
Key Takeaways:
- Start small where 90% accuracy is acceptable: automate low-stakes processes first, then expand as you build confidence
- Real use cases that drive value today: content production, lead qualification, audience research, campaign optimization, and knowledge management
- Build partnership, not replacement: let agents handle execution while humans provide direction and judgment
- Measure efficiency and quality: track dual metrics so you know what's actually working
- Treat agent literacy as an edge: teams that learn to direct, evaluate, and orchestrate agents pull ahead of those who wait
|
|
|
🛠️ Deep Dives
1. Loom: Record quick videos and share them easily to communicate better.
Why It's Worth Your Time:
- Clear Communication: Explain complex ideas visually instead of with endless text.
- Save Time: Record once and share with many, cutting down on meetings and emails.
- Easy Feedback: Get quick visual feedback on designs or documents by talking through them.
Pricing: Free version for basic recordings, paid versions ($12.50/month) add longer videos, editing features, and advanced analytics for teams.
2. Duolingo: Learn new languages with fun, bite-sized lessons that feel like a game.
Why It's Worth Your Time:
- Gamified Learning: Stay motivated with points, streaks, and fun challenges.
- Personalized Pace: The AI adapts lessons to your progress, making learning efficient.
- Free Access: Get started learning many languages without any cost.
Pricing: Free to learn with ads, paid Duolingo Plus ($6.99/month) removes ads and allows offline lessons.
3. Stratup.ai: Generate innovative startup ideas quickly using AI and current market trends.
Why It's Worth Your Time:
- Idea Generation: Quickly get fresh, market-savvy startup ideas.
- Trend Analysis: Ideas are based on current market trends, increasing relevance.
- Personalized Suggestions: Get ideas tailored to your interests and input.
Pricing: Free to generate a certain number of ideas, paid plans unlock more advanced features and generations.
|
⏭️ What's Next
Agents were the throughline this week, from Gemini Spark handling tasks while you sleep to Microsoft's new models built to work together. Pick the one job you'd most like to stop doing, and hand it to an agent. See you next week.
|
How did we do? |
|
|
|
