In this issue
Google just rolled out Gemini-powered Googlebooks and a major intelligence update for the Android ecosystem. Instead of forcing you into a standalone chatbot tab, these updates push agentic workflows straight into the operating system.
This shift toward ambient, background tools removes a huge layer of friction for teams. When your device can read context on its own to generate charts or compile notes from scattered apps, you spend less time prompting and more time executing.
Topics of the day:
Google's agentic Android OS and new Googlebooks
Google and SpaceX explore orbital AI data centers
Claude Code's new financial research plugins
Amazon launches autonomous Alexa for Shopping
Curated reads on agent readiness benchmarks, Claude Code over Telegram, and Cloudflare's new email service
The Shortlist: WhatsApp ships incognito Meta AI, Krea launches K2, and Perceptron debuts Mk1
Google brings Gemini intelligence to Android and introduces Googlebooks
What's happening: Google just introduced a lineup of new Googlebook laptops alongside a major Gemini Intelligence upgrade for the Android ecosystem. I love this approach because it moves AI out of a standalone chatbot window and turns it into a helpful background assistant running across your entire device.
In practice:
You can now automate multi-step workflows, like pulling scattered project notes from Gmail and turning them into a clean brief.
The new Magic Pointer cursor reads your screen's context, so you can point at a messy spreadsheet to instantly generate charts without writing a prompt.
Operators can quickly build custom widgets using plain-language instructions to track specific metrics or business updates right on the home screen.
Bottom line: Long prompts in a separate browser tab are on the way out. Wire these ambient tools into your daily routine now and you'll save real hours every week without changing how you already work.
Google and SpaceX explore orbital AI compute data centers
What's happening: Google is in talks with SpaceX to launch orbital data centers that bypass terrestrial power grid limits. The project aims to test space-based computing by 2027 to feed the massive energy appetite of future AI models.
In practice:
Solving the energy bottleneck means you should see API costs for enterprise AI deployments fall meaningfully over the next few years.
Cheaper compute directly enables complex automation pipelines, letting your team run hundreds of autonomous agents without blowing the budget.
With Anthropic also seeking orbital compute capacity, the infrastructure expansion ensures you'll have the horsepower needed to grow.
Bottom line: Build for where compute is heading, not where it sits today. You're not launching rockets, but this kind of buildout means raw power won't be your bottleneck when you start scaling agents into real workflows.
Turn Claude Code into a personal Wall Street analyst
What's happening: I've been experimenting with Anthropic's new plugin marketplace for Claude Code, and it essentially turns your terminal into a personal Wall Street analyst. You can install specialized agents that handle market research and equity analysis right out of the box.
In practice:
Automate deep research by running the Market Researcher agent to digest public filings and output formatted competitive landscapes.
Cut spreadsheet busywork with built-in financial analysis skills that generate valuation models in minutes, saving hours of manual data entry.
Prep for meetings faster by deploying the Meeting Prep agent to compile briefing packs and portfolio notes before your next client pitch.
Bottom line: Treat Claude Code like an analyst on call. Ask it to cite every source, double-check the math yourself, and you'll cut hours from every pitch deck without paying a junior banker's salary.
Amazon's Alexa for Shopping replaces Rufus
What's happening: I always thought Amazon would take AI shopping further, and they just replaced Rufus with Alexa for Shopping to do exactly that. The new agentic assistant tracks prices, automates your purchases, and even buys from competitor websites on your behalf.
In practice:
Automate your entire procurement process by setting recurring scheduled actions for essential office supplies and tech hardware.
The Buy for Me tool uses agentic AI to literally shop at other online stores on your behalf when Amazon falls short.
E-commerce founders now have to optimize product listings for autonomous shopping agents, not just traditional human search behaviors.
Bottom line: Agentic commerce took another stab on going mainstream and it'll change how you buy supplies and how you sell your own products. If you run an online brand, your most important target customer could be an AI assistant.
Read Later
ora Agent Readiness Ranking - A public leaderboard scoring how well different sites and apps work with autonomous agents, useful if you're optimizing for the Alexa-style buyer above.
MuleRun: the AI agent that gets work done - General-purpose work agent positioning itself as the do-it-all layer on top of your existing SaaS stack.
Where good ideas come from (for coding agents) - Sunil Pai's seven-ways framework on how to actually direct coding agents toward good decisions, not just any output.
Cloudflare Email Service public beta - Send and receive transactional email directly from Workers, a clean primitive for agent-built apps that need to talk to humans.
The Shortlist
WhatsApp introduced an incognito mode for Meta AI that processes data without saving logs, giving you a safer space to hash out sensitive business ideas.
Krea launched K2, an image model built for aesthetic control that lets design teams blend multiple reference styles into a single cohesive visual moodboard.
Arena updated its LLM rankings to crown Claude Opus 4.7 as the top overall model, while confirming Meta's Spark as the new leader for coding tasks.
Perceptron released Mk1, an interaction model that treats video as a continuous real-time stream rather than a pile of static screenshots.
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.
