Welcome back! A new Gallup survey found that even Americans who use AI every day are worried about it. Which feels about right: we’re asking chatbots to write emails, plan meals, summarize PDFs, and occasionally explain tax law at midnight, while also quietly wondering whether any of this is a good idea, if we’ve lost the ability to think, and if any of us will have jobs in the future. It’s a confusing time!
Is This the End of AI Entanglement?
The AI web of companies just got a big hole punched in it.
Amazon has officially blocked Meta's Muse AI from shopping on its platform. Amazon says Meta didn't get permission and raised concerns about how the agent identifies itself and handles customer credentials.
We've talked before about the "AI web of companies”—where every big tech company seems to be investing in, partnering with, and competing against each other all at the same time. Google investing billions in Anthropic while competing directly with Claude is a pretty good example.
The Meta-Amazon dynamic is another one. Meta signed a multibillion-dollar deal in April to run Muse's agentic AI workloads on Amazon's Graviton chips. So Amazon is literally powering the same AI agent it just banned from shopping on its own site. Amazon products have also been purchasable inside Facebook and Instagram since 2023.
But this feels like big tech is starting to take its rivalries more seriously than before.
And it's part of a bigger pattern:
Amazon previously sued Perplexity over its Comet shopping agent and moved to block shopping agents from Google and OpenAI.
Anthropic cut off OpenAI's access to Claude last year, saying its use violated terms against building competing products.
Microsoft shut down its Bing Search APIs last year, cutting off a source of search results for smaller competitors.
So is this the end of AI entanglement? Or just the messy adolescent phase before it gets even more tangled?
Hit reply and let me know where you think this all lands.
— Matt
YouTube Adds AI Across Music and Creator Tools
Via Mashable
YouTube is rolling out a new wave of AI features across YouTube Studio and YouTube Music, giving creators and listeners more ways to use AI inside the platform.
Creator tools: In YouTube Studio, creators are getting more AI help with analyzing videos, improving drafts, and growing their audiences. YouTube is expanding Ask Studio, its AI-powered Q&A feature, to iOS and Android, and adding draft feedback that can review unpublished videos for pacing, structure, and storytelling. The app will also generate titles and thumbnails based on a video’s content and the creator’s style, then test multiple options with different audience segments.
Music and podcasts: In YouTube Music, a new Ask Music feature lets users describe what they want to hear in plain language instead of searching for specific songs or artists. Users can ask for custom queues, explore the influences behind songs, or dig into details about an artist or track. YouTube is also adding Your Podcast Lineup, an AI-generated audio guide that gives listeners short spoken previews of recommended shows.
The sensitive part: AI inside YouTube Music is where things get more complicated. Discovery tools and conversational search are useful. But music, creator identity, and AI-generated content are already touchy territory, especially when audiences are trying to understand what humans made, what AI assisted with, and what is essentially synthetic filler.
The bigger picture: Platforms want AI to feel native, not like something users leave the app to access. For YouTube, that means putting AI into creation, analytics, music discovery, podcast discovery, editing, and optimization. But content platforms have to walk a narrow line: creators want faster tools, platforms want more engagement, and viewers do not always want their feeds flooded with AI-shaped content.
Amazon Rolls Out Agentic AI for Sellers
Amazon introduced new agentic AI tools for third-party sellers, expanding Seller Assistant into a more proactive business tool inside and outside Seller Central.
How it works: Sellers use “workflows” that run in the background and act on prompts, like watching for pricing changes or surfacing risks before they become bigger problems. Amazon says it can monitor a seller’s business continuously, flag issues like falling ratings or slow-moving inventory, and recommend actions based on each seller’s profile and performance data.
AI integration: Amazon is also opening seller tools to outside AI assistants through plugins, starting with Amazon Quick and Anthropic’s Claude. That means sellers can connect their Amazon account to supported AI tools and ask questions or take actions without always logging into Seller Central.
Control and data: Amazon says the tool is free and optional, and sellers can decide how much data they share through the agentic service. That matters because the tradeoff is obvious: the more context the agent has, the more useful it can be. But for sellers, that also means giving Amazon even more visibility into pricing, inventory, operations, growth plans, and business strategy.
The bigger picture: Agents are becoming the next layer of software. For sellers, embedding agents directly could be genuinely useful if they save time, catch problems early, and turn scattered dashboards into decisions. But it also tightens Amazon’s position as both a marketplace and an operating system.
Your New AI Teammates Get Work Done. No Prompting Needed
Imagine opening your laptop Monday morning and finding that your SEO report, competitor updates, meeting prep, and next campaign recommendations are already there, waiting for you. With Optimizely’s new Virtual Teammates, you don’t have to imagine.
Join Optimizely Sept. 29 to see how you can:
Add extra capacity to your team by hiring Virtual Teammates with role-based responsibilities
Use them to keep work moving and collaborate on execution, always completing work to your standards
Get a full audit trail of everything they do, and configure when human reviews and approvals are required
Want to meet your Virtual Teammates? Save your webinar spot today.
Build API Workflows Visually
Via Orchesty
Orchesty is an open integration platform for designing, scheduling, and orchestrating asynchronous workflows across connected services and APIs.
How you can use it
Use open-source connectors or create your own
Generate connectors from API specs with AI
Deploy in managed cloud or self-hosted environments
Add retries, rate limits, persistence, and access controls
Pricing: Paid
Unify Customer Conversations
Via Mixdesk
Mixdesk is an AI-powered customer engagement platform that brings channels like WhatsApp, Instagram, Facebook, Telegram, LINE, LiveChat, and email into one shared workspace.
How you can use it
Manage customer messages from one inbox
Automate routine support conversations
Qualify leads with AI
Route customers to the right agent or team
Run targeted campaigns and proactive outreach
Pricing: Free and paid
Jobs, announcements, and big ideas
Meta gives its Muse AI agent a cute, fuzzy mascot.
Microsoft releases a Copilot super app for business users.
OpenAI upgrades prompt caching for GPT-6 with higher hit rates and new diagnostics tools.
Tech bosses discuss AI over dinner with Trump and Xi.
TypeSafe AI is hiring a Developer Advocate to build AI demos, write clear docs, grow their developer community, and turn feedback into product improvements.
One day, two models. Here’s why I think Opus 5.5 is crazy good.
That’s a wrap! See you next week for more.
