AI's Infrastructure Power Struggle
· The Fluency Briefing
The Fluency Briefing
Your Guide to What's Happening in AI and Why It Matters to You
Tuesday, April 28, 2026
Three stories dropped today that share a hidden thread: Bloomberg is putting a chatbot inside the most intimidating financial tool on earth, the FIDO Alliance is scrambling to stop AI agents from going rogue with your credit card, and China just blocked Meta from buying an AI platform it considers a national asset. The pattern?
AI is no longer a feature being bolted onto products - it's becoming the infrastructure itself, and everyone from Wall Street traders to foreign governments is racing to control the terms.
Today in AI:
- Bloomberg's Terminal Gets a Chatbot, Traders Get Nervous - Bloomberg is testing ASKB, a chatbot-style interface layered onto its iconic Terminal, now available to roughly a third of its 375,000 users. The goal: help finance pros query massive datasets in plain English instead of memorizing arcane commands. Wired
- Your AI Agent Wants Your Credit Card (FIDO Wants a Word) - The FIDO Alliance, with Google and Mastercard, announced Tuesday it's building industry standards to stop AI agents from being hijacked mid-purchase. As agents start buying things on your behalf, the risk of rogue transactions and phishing just leveled up. Wired
- China Blocks Meta's $2B Bid for AI Agent Platform Manus - Beijing placed export controls on Manus's algorithms, killing Meta's proposed acquisition of Butterfly Effect even after the startup relocated to Singapore. The message to Chinese AI startups: moving your HQ doesn't move you out of reach. Fast Company
- Otter Wants to Be Your Entire Workspace, Not Just Your Note-Taker - Otter launched enterprise search via Model Context Protocol, letting users query Gmail, Google Drive, Notion, Jira, and Salesforce alongside meeting transcripts from a single interface. The meeting bot is making a play to become your work brain. TechCrunch
- Musk vs. Altman Goes to Trial - For Real This Time - Starting Tuesday, Elon Musk and Sam Altman face off in an Oakland courtroom over whether Altman betrayed OpenAI's non-profit mission. Musk wants billions returned and Altman ousted. The trial is expected to last a month. BBC
- Meta's Muse Spark Model Debuts Ahead of Earnings - Meta unveiled Muse Spark, its first paid-access AI model, moving away from the open-source Llama strategy. It trails Claude in several benchmarks but beats GPT in text and vision. Wall Street wants details on Wednesday's earnings call. CNBC
- LLM Slop Is Everywhere and Nobody's Admitting It - A LessWrong post cataloging AI-generated text found in blog posts, keynotes, personal messages, press releases, and even subreddits dedicated to making AI sound more human. The irony is thick - and the author has receipts. LessWrong
- Chatbot Ads Are Here, and You Probably Won't Notice Them - Researchers published a study showing AI chatbots trained to embed personalized product ads successfully influenced users' choices - and most people had no idea they were being sold to. OpenAI recently hired Meta's longtime ad executive. Fast Company
Today's Takeaway:
Look at today's stories through a competitive lens and one thing becomes clear: AI is no longer just about building better models. It's about controlling the points of access. Bloomberg isn't building its own large language model - it's layering a chatbot on top of multiple models to keep its Terminal indispensable. Otter isn't improving transcription - it's turning itself into a search hub that connects your scattered work tools through a common protocol.
Even Redpine's seed round is about becoming the middleman between AI agents and premium data that isn't on the open internet. The product isn't the AI. The product is the interface that sits between you and the AI. As Wired reports, Bloomberg's CTO admits the old Terminal was becoming "untenable" under the weight of its own data.
And TechCrunch notes that meeting notetakers realized transcription alone couldn't justify their valuations. Translation: the real money isn't in the models themselves - it's in owning the layer where humans actually interact with AI. That's why China blocked Meta's Manus acquisition (Fast Company).
Manus isn't a model. It's an orchestration layer that sits on top of multiple models. Beijing understands that whoever controls the interface controls the user - and possibly the market.
💡 Fluency Moment - Building your AI fluency, one term at a time.
"Retrieval-Augmented Generation"
In plain English: AI that searches a live database before answering, instead of relying only on memorized training.
Think of it like: A student allowed to look things up in a textbook during an exam, not just from memory.
Why you'll hear about it: Bloomberg's chatbot queries live financial data - that's RAG powering real-time answers.
🧰 Your Toolkit
5-Minute Quickstart: How to Use AI Tools to Stay on Top of Tech News
- Open ChatGPT or Google Gemini and type: 'Explain [AI topic, e.g. vertical AI] in simple terms like I'm completely new to this.'
- Ask the AI: 'What does it mean for a company to have a moat in [INDUSTRY]? Give me a real-world example.'
- Try this prompt: 'Summarize why businesses are adding AI to tools like [TOOL NAME, e.g. Bloomberg Terminal] and what that means for everyday users.'
- Type: 'What are the risks of AI having access to private or non-public data? Explain it simply.'
- Ask: 'What AI news should a beginner pay attention to this week, and why does it matter to regular people?'
- Bookmark one AI newsletter or news site (like Crunchbase News or WIRED) to check weekly and stay consistently informed.
Once you're comfortable asking AI to explain the news, try asking it to compare two AI companies or tools you keep hearing about. You'll quickly build a solid foundation for understanding where AI is headed.
The Bottom Line
The Pattern: Today's biggest moves aren't about who has the best AI model. They're about who controls the layer between the model and the user - the search bar, the agent, the orchestration platform, the terminal interface. The model race is maturing into an access race.
Why It Matters: If you're evaluating AI tools for your business, stop asking "which model is best" and start asking "which interface gives me the most useful access to the most data." That's where the competitive advantage is shifting, and it's where the lock-in will happen.
Your Move: This week, map out every AI touchpoint in your workflow - every chatbot, copilot, and assistant. Ask yourself: who controls that interface, and what happens to your data once it passes through? The answers might change which tools you keep.
What We're Working On
✨ Founding Cohort Special - 60% Off! - Use code MAF20 to join for just $20/month (regularly $50). Get weekly group sessions & workshops, self-paced courses for all levels, access to tools & templates, challenges with peer feedback, and 24/7 support community. → Join Now
✨ Free 30-Minute AI Consultation - Discover how My AI Fluency can help your business unlock the potential of AI. We'll discuss your goals, explore practical AI opportunities for your industry, and outline clear next steps. → Schedule Free Call
✨ How AI-Fluent Are You? - Test your AI fluency with our interactive quiz. See how you stack up and discover what to learn next. → Take the Quiz
💬 Community | 📞 Book a Consultation | 🌐 Website
Fluently yours, The My AI Fluency Team