In partnership with |
 |
|
|
| |
Hey Radarians, this week AI is getting bigger, smarter, and more practical. |
Billion-dollar AI infrastructure, space-based computing, automated follow-ups, writing research, and tools worth watching. |
Let’s dive in. |
|
|
|
|
|
|
|
|
|
| |
🤝 Broadcom offers Anthropic up to $42 billion for AI chips |
|
Broadcom agreed to lend Anthropic up to $42 billion to finance AI infrastructure, as the Claude maker expands its reliance on costly TPU computing capacity The financing could cover about one-third of Anthropic’s $125.2 billion five-year TPU lease commitment, while the debt could later convert into Anthropic shares Anthropic could become Broadcom’s largest compute customer in 2027, while its filing warns the financing and hardware relationship creates potential conflicts of interest |
🏗️ Amazon commits $1 billion to US data center |
|
AWS will invest more than $1 billion over five years in US communities hosting its data centers, funding education, job training, energy affordability, water conservation The move comes as AI data centers face growing backlash over electricity demand, utility costs, and water use, with more than 100 proposed data center moratoriums being considered Amazon says it invested $276 billion in data centers from 2011 to 2025 and is expanding across several states, while AWS says it is 75% toward becoming water positive by 2030 |
|
|
|
|
|
| |
Free Workshop: Turn a Tiny Audience Into $10k/month |
|
We’ve driven over $10M in revenue for our newsletter clients and are unpacking the plug-and-play system thousands of people are using to turn a small email list into $2k-$10k/month. |
(…without needing to quit your job, without needing to get on camera, and without needing prior experience). |
We’ll show you how to start a newsletter in <30 minutes, scale to 1,000 readers, and make your first $10,000 in revenue. |
Save Your Free Seat |
|
|
|
|
|
| |
More News on the Radar |
OpenAI fired three researchers after an internal probe found they mishandled confidential company data and broke policy |
Google launched an AI satellite with a TPU chip, moving closer to building AI data centers in space |
China is limiting AI companions for minors, adding safety rules to reduce addiction, dependency and self-harm risks |
|
|
|
|
|
| |
⚡ Steal This Workflow |
Never Forget a Follow-Up Again |
|
Stop relying on memory for “I’ll send this tomorrow,” “check back next month,” or “follow up after the launch.” Let AI surface the conversations that need your attention before they disappear. |
-
Connect Gmail to ChatGPT
Open ChatGPT → Settings → Plugins and connect Gmail.
This lets ChatGPT search your email history and understand the conversation around each commitment instead of working from isolated messages.
-
Run your first commitment scan
Ask ChatGPT to review your recent sent emails and look for things you promised to do later.
Focus on:
Files or information you promised to send People you said you would contact again Decisions you promised to make Updates you owe someone Conversations you said you would revisit Follow-ups without a reply yet
Have it return the person, commitment, expected date, and original conversation.
Review the list once and remove anything already completed.
-
Turn it into a recurring check
Open Scheduled in ChatGPT and create a weekday morning task.
Use:
“Check my recent Gmail conversations for commitments and follow-ups that are due, overdue, or coming up soon. Exclude anything later emails show has already been completed. Prioritize the remaining items and include the person, commitment, timing, and relevant thread.”
Set it to run every weekday morning and enable notifications.
You no longer need to remember to check.
-
Handle each follow-up from the original thread
When something appears in your daily list, open the relevant conversation.
Have ChatGPT read the latest messages before drafting anything so it knows:
What was originally promised Whether circumstances changed What the other person last said What still needs to happen
Then either complete the action yourself or have ChatGPT draft the follow-up using that context.
|
|
|
|
|
|
| |
🔎 AI RESEARCH: AI Writing Assistants Are Making Writing More Alike
|
|
What researchers studied: |
Researchers analyzed more than 880,000 texts across seven datasets to see how AI writing assistants affect the diversity of human language. A separate study examined 1.5 million Change .org petitions before and after an AI writing tool was introduced. |
What they found: |
AI-assisted writing kept much of the original meaning but made writing styles more similar. Across datasets and models, variation in writing complexity fell by 21–50%. |
The Change .org study found a similar pattern: AI changed how petitions were written and increased their linguistic similarity, but did not improve their chances of success. |
Why it matters: |
AI can make writing cleaner and easier to produce, but widespread reliance on the same models may gradually push millions of people toward similar language, structures, and ways of expressing ideas. |
Source: Nature Human Behaviour, August and September 2026 |
|
|
|
|
|
| |
🛠️ Tools Worth Watching |
Subscribr: AI workspace for YouTube research and production FLORA: Connects image, video, audio, and text models inside one visual creative workflow Hemory: Captures real-world conversations and turns them into searchable AI memory |
|
|
|
|
|
|
|
| |
📝 Prompt Radar: Turn Your Expertise Into a Framework
|
What it does: Extracts the repeatable thinking, methods, and decision rules hidden inside your experience. |
Work as a framework designer. Study the experience, notes, examples, processes, successes, failures, and opinions I provide: [PASTE]. Look beyond the obvious steps and identify the recurring principles, sequence, decision rules, signals, trade-offs, shortcuts, mistakes, exceptions, and judgment calls that explain how I actually get results. Separate fundamental principles from situational tactics, uncover patterns I may be using unconsciously, and organize them into a clear 3–7 part framework that another person could understand and apply. For each part, explain what matters, what decision needs to be made, and what commonly goes wrong. Do not invent missing expertise; flag gaps or contradictions. Finish with a memorable framework name and a concise explanation of how all the parts work together.
|
Best for: |
Consultants and experts Creators and educators Founders and professionals |
|
|
|
|
|
| |
💼 AI Side Hustle Idea: Build Digital Products With AI
|
|
People don’t need another pack of generic AI prompts. They will pay for digital products that solve a specific problem and save them time. |
The problem:
Creating a useful template, calculator, database, planner, research library, or mini-tool normally takes hours of research, writing, organizing, designing, and testing. |
Your AI service:
Pick a problem you understand and use AI to help research the market, structure the solution, create the content, build the first version, and improve it from customer feedback. |
Think Notion systems, spreadsheet calculators, industry databases, planning kits, research packs, educational resources, or simple niche tools. |
Simple workflow:
Choose a niche → find a repeated problem → validate demand → build with AI → test with real users → package it → sell online |
Best clients:
Creators, freelancers, students, job seekers, marketers, founders, small businesses, and niche professional communities. |
What you can charge for:
Individual digital products + bundles + premium versions + paid updates or memberships. |
|
|
|
|
|
| |
Learn AI with This Course |
|
|
What you’ll learn: |
Learn how Model Context Protocol (MCP) lets AI applications connect to external tools, data, and services, then build and deploy your own MCP-powered applications. |
Key topics: |
Understanding MCP architecture, clients, servers, tools, resources, and prompts Building an end-to-end MCP application using Python, TypeScript, Gradio, and Hugging Face Creating custom MCP servers that connect AI assistants with real tools, data, and workflows |
|
|
|
|
|
Until next week,
AI Radars |
|
