|
In this newsletter:
- Anthropic releases Opus 4.6 that runs multiple AI agents simultaneously
- OpenAI launches Codex desktop for managing multiple AI coding agents
- Meta tests standalone Vibes app for AI-generated videos
Plus, you’ll find new AI tools and this week’s top AI news headlines!
🤖 Anthropic Releases Opus 4.6 That Runs Multiple AI Agents Simultaneously
Anthropic just made AI agents collaborate like a real team, letting you split complex projects across multiple specialists instead of waiting for one agent to finish everything sequentially.
What's new:
- "Agent teams" feature splits larger tasks into segmented jobs across multiple agents
- Instead of one agent working sequentially, multiple agents each own their piece and coordinate directly with others
- Scott White, Head of Product, says it's like having a talented team of humans working for you, coordinating in parallel to work faster
- Agent teams available in research preview for API users and subscribers
- Context window expanded to 1 million tokens (matching Sonnet 4 and 4.5)
- Longer context allows work with larger code bases and the processing of larger documents
- Claude is integrated directly into PowerPoint as an accessible side panel
- Previously, users told Claude to create a PowerPoint deck, then transferredthe file to edit, and now craft presentations within PowerPoint with direct Claude help
- Released just 3 months after Opus 4.5 launched in November
Who's using it:
- Non-professional software developers using Claude Code as a task engine
- Product managers, financial analysts, knowledge workers across industries beyond tech
- Scott White, Head of Product, says Opus evolved from a software development tool to "really useful for a broader set" of users
Why it matters:
Anthropic is racing to keep Opus competitive as OpenAI just launched Codex Desktop and Microsoft expands Copilot presence across enterprises.
Agent teams that work in parallel addresses major bottleneck of sequential AI task completion, positioning Claude as a platform for delegating work to multiple specialists rather than relying on a single assistant doing everything slowly.
👀 Read about Claude Opus 4.6 release!
💻 OpenAI Launches Codex Desktop for Managing Multiple AI Coding Agents
OpenAI released Codex Desktop for macOS as a native app, treating AI coding agents like teammates you can direct and review, arriving as Anthropic's Claude Code gains rapid adoption, and after a brief Claude outage, reminded developers that local tooling matters.
What's new:
- Codex Desktop runs on macOS as a command center for AI agents
- Create projects, spin up threads, run work in parallel with a live audit trail of diffs and comments before pushing code
- Integrates with Git, supports worktrees, and brings agent conversations into desktop view
- Jump between tasks without losing context
- Included in ChatGPT Plus, Pro, Business, Edu, and Enterprise plans
- Available to Free tier users for a limited time, no extra procurement needed
- Expanded token limits for all teams announced
- Unifies existing web, CLI, and IDE integrations intoa single desktop flow
- Supports multitasking and parallel execution to run several jobs at once
How it works:
- Agents handle specific tasks like contractors with specialized skills
- Assign jobs, watch progress, and sign off on changes before execution
- Can pause any task to ask why changes were made
- Designed for hours-long or days-long refactors instead of just quick completions
- Shifts from chat-based interactions to task-level delivery
- Users outline the project, define acceptance tests, and let agents work while reviewing
- Teams can run agents overnight to handle backlogs and testing chores
The competitive pressure:
- Anthropic surged with Claude Code and a desktop experience seen as fast and high value
- Claude Opus model is viewed as highly useful and powerful by many developers
- Anthropic's Cowork version is optimized for non-technical users with custom agent skills for everyday workflows
- Claude models experienced a brief outage today that rippled across teams using Claude Code
- Microsoft's Copilot sits inside nearly every major IDE with sophisticated chat, terminal support, agent mode, and an expanding "Workspace" concept
- OpenAI needs visibility to stay inthe enterprise conversation
Why it matters:
Desktop presence moves AI coding tools from "interesting add-on" to "daily seat" for enterprise deployment, but agent autonomy remains double-edged as too much freedom creates security and compliance issues while too tight control limits competitive advantage.
OpenAI's timing capitalizes on Claude's brief outage to remind developers that local tooling provides redundancy when web endpoints fail, positioning Codex as an insurance policy against cloud dependency.
👀 Read about OpenAI Codex Desktop App!
🎬 Meta Tests Standalone Vibes App for AI-Generated Videos
Meta is spinning out Vibes from its Meta AI app into a standalone platform competing directly with OpenAI's Sora, positioning AI-generated video as the new TikTok, where every clip you scroll throughis created by AI.
What's new:
- Vibes is launching as a standalone app after living inside the Meta AI app since its September launch
- Feed displays AI-generated videos like TikTok or Instagram Reels, but every video is AI-created
- Users can generate a video from scratch or remix videos they see on the feed
- Add new visuals, layer in music, and adjust styles before publishing
- Post directly to Vibes feed, DM to others, or cross-post to Instagram and Facebook Stories and Reels
- Meta says Vibes performed well, with Meta AI usage growing steadily since launch
- Collaboration and sharingare rising with many Vibes videos messaged to friends, mirroring how people use Reels
- A standalone app provides a more focused and immersive environment for creation and engagement
The subscription plan:
- Vibes has been free since launch, but Metais planning a freemium model
- Will offer free basic access with the option to subscribe for additional video creation opportunities each month
- Test subscriptions launching in the coming months
- Part of Meta's broader exploration of subscriptions for AI features across Facebook, Instagram, and WhatsApp
Why it matters:
Meta is betting that AI-generated video needs its own dedicated feed, separate from human-created content, directly challenging OpenAI's Sora by making Vibes a standalone destination rather than a buried feature.
The move to freemium subscriptions signals that Meta sees AI video creation as a premium feature users will pay for, testing whether people want a TikTok-style feed of AI clips enough to subscribe for unlimited creation.
👀 Read about Meta’s Standalone Vibes App!
My Latest LinkedIn & X/Twitter Posts:
- AI tools stack for founders and CEOs (view post)
- 12 AI terms you must know (view post)
- Top 11 free AI courses to learn core AI skills (view post)
- Top websites with real remote jobs that pay in USD (view post)
- 10 powerful ways to use ChatGPT deep research (view post)
- 12 ChatGPT prompts for critical thinking (view post)
- 9 ChatGPT prompts to extract strategies and find insights (view post)
In partnership with Kit:
As a creator, your time should be spent doing what you love, not juggling a dozen tools just to run your business.
Kit gives you everything you need in one place:
✅ Build and grow your email list (I use Kit for my newsletter)
✅ Easily monetize with paid newsletters and digital products
✅ Automate your emails with triggers and custom workflows
✅ Track what’s working and optimize with powerful insights
It’s not just email. It’s your entire creator business, simplified.
Join a thriving community of successful creators.
Use Kit for Free — and Start Building Smarter
New AI Tools to Boost Your Productivity:
- ElevenLabs: Generates text-to-speech and cloned voices.
- Riverflow: Creates campaign visuals from prompts.
- UGC Maker: Generates UGC-style video ads with avatars.
- Dokie: Creates presentation slides from text.
- Devoice: Transcribes audio and video into text.
- Hugo: Automates customer support chats and ticket handling.
- ShortFast: Creates short-form videos from long content.
- Yooz: Generate LinkedIn posts inspired by top creators & your brand voice.
- Team9: Builds custom AI agents for business workflows.
- Loamly: Track ChatGPT, Claude, Perplexity website traffic.
- Intuo: Monitors industry updates and delivers client-ready briefings.
- NextDocs: Turns ideas into stunning slides and documents.
- DartAd: Turn your product images into scroll-stopping ad videos.
- TryMusic: Generates music tracks from text prompts.
- Supernormal: Records meetings and generates notes and summaries.
This Week's Top AI News Headlines:
- Sapiom Raises $15M to Build Platform That Lets AI Agents Autonomously Buy and Manage Their Own Software Tools (View Article)
- Reddit Bets on AI Search as its Next Major Growth Engine to Turn Community Posts into Answer-Driven Discovery (View Article)
- AWS Revenue Continues to Soar as Cloud Demand Remains High, Driven by AI Infrastructure Growth, Enterprise Migration, and Expanding Customer Spend (View Article)
- Amazon and Google Lead the AI Capex Race as Spending on Data Centers and Chips Surges, Raising Questions About Long-Term Returns (View Article)
- Elon Musk Pushes Forward Plans for Orbital Data Centers to Power Space-Based AI Compute and Reduce Earth-Bound Infrastructure Limits (View Article)
- OpenAI Launches Enterprise Platform to Build, Deploy, and Govern AI Agents Across Business Workflows (View Article)
- ElevenLabs CEO Says Voice is Emerging as the Primary Interface for AI as Speech Technology Scales Across Consumer and Enterprise Products (View Article)
- Alphabet Declines to Discuss Google–Apple AI Deal Even With Investors, Fueling Questions About Strategy and Competitive Impact (View Article)
- Apple Integrates Anthropic’s Claude and OpenAI’s Codex into Xcode 26.3 to Bring AI-Powered Coding, Debugging, and Code Generation to Developers (View Article)
- Gizmo Launches a TikTok-Style Platform Where Users Create, Remix, and Share Interactive Mini Apps Using AI-Assisted “Vibe Coding” (View Article)
- Mistral Releases Voxtral Transcribe 2, an Open-Source Speech-to-Text Model that Runs on Consumer Hardware to Cut Cost and Latency (View Article)
- Amazon Begins Testing AI Tools for Film and TV Production to Speed Up Script Development, Editing, and Post-Production Workflows (View Article)
Work With Me:
If you enjoy this newsletter, please forward it to your friends and colleagues.
Follow me on LinkedIn and X/Twitter to see my latest posts.
Have a wonderful week!
Andrew Bolis
|