Hey!
Google just rolled out an update that turns any pair of headphones into a real-time, two-way translator using Gemini. While it sounds like a travel feature, this is a major step in making global business collaboration a lot more practical.
This moves live translation from a high-cost service to a standard feature anyone can use. For operators, it lowers the barrier for international sales calls, partner meetings, and even customer support without needing a human interpreter.
Topics of the day:
Google’s Gemini brings live translation to any headphones
The shift from AI ‘intelligence’ to ‘usefulness’
How Zoom’s multi-AI approach beat bigger models
The Shortlist: Meta builds its new Avocado model on Alibaba’s Qwen, Cohere upgrades enterprise search with Rerank 4, and World expands its “proof of human” app with secure chat and global payments.
What’s happening: Google just rolled out a major update to its Translate app, using Gemini to stream real-time, two-way audio translations to any headphones connected to your Android phone.
In practice:
This makes international sales calls or partner meetings feel more natural, helping you understand tone and pacing without the usual lag of an interpreter.
You can reduce overhead for live events or training sessions by replacing the need for human interpreters in many informal scenarios.
Small teams can now confidently engage with new markets, making it easier to expand globally without hiring a dedicated multilingual support staff.
Bottom line: This moves real-time translation from a niche travel gadget to a practical business tool. It's a huge step toward making global collaboration accessible to any team, anywhere.
Why AI's 'IQ' is peaking, and 'usefulness' is next
What’s happening: Top tech investor Gavin Baker argues the AI industry is hitting diminishing returns on raw intelligence. He says the next frontier isn’t a higher IQ, but genuine “usefulness”, making AI reliable, context-aware, and capable of handling complex tasks.
In practice:
Target automations with verifiable outcomes, like balancing a ledger or qualifying a sales lead, as these are the first tasks AI will master with high reliability.
Start treating your company’s data as the ultimate competitive advantage, because an AI with deep context on your emails and projects is far more valuable than a generic one.
Shift your thinking from single tasks to complex workflows by identifying processes that take hours of coordination, not just minutes of work, as the next big automation win.
Bottom line: The race is shifting from who has the smartest model to who provides the most tangible ROI. This means focusing on AI tools that solve real business problems, not just ones that win benchmark tests.
Zoom's AI shocks by topping reasoning benchmark
What’s happening: Zoom’s AI research team just achieved a new state-of-the-art result on a tough AI reasoning benchmark, outperforming dedicated labs. They did it not by building one massive model, but by orchestrating multiple AIs in a “federated” approach that picks the best one for each task.
In practice:
Instead of relying on one "do-it-all" AI, you can build cheaper and more effective automations by routing tasks to the best model for the job, like using a specialized model for data entry and another for creative writing.
Expect tools like Zoom AI Companion to get much smarter, delivering more accurate meeting summaries and automating complex business workflows directly from your calls.
This proves you don’t need to build a foundation model from scratch; breakthroughs can come from cleverly combining existing tools, a strategy any business can adopt.
Bottom line: This shows the AI race isn't just about who builds the biggest model. Smart orchestration can beat raw power, creating a massive opening for companies that are expert integrators, not just builders.
The Shortlist
Meta trained its new Avocado model using Alibaba's Qwen, showing how the open-source feedback loop now has major competitors building on each other's work.
Cohere launched Rerank 4, a model that upgrades enterprise search by allowing AI agents to process entire documents at once, reducing errors and improving context.
World announced major updates to its Sam Altman-backed “proof of human” super app, adding secure chat and global payments as part of its push to build a global digital identity network.
This newsletter is where I (Kwadwo) share products, articles, and links that I find useful and interesting, mostly around AI. I focus on tools and solutions that bring real value to people in everyday jobs, not just tech insiders.
Please share any feedback you have either in an answer or through the poll below 🙏🏽
