In partnership with |
 |
|
|
| |
Hey Radarians, this week AI is revealing where the money, attention, and real opportunities are moving. |
AI spending, coding shifts, smarter agents, creative research, automated workflows, useful tools, better prompts, and AI side hustles. |
Let’s dive in. |
|
|
|
|
|
|
|
|
|
| |
🧠 a16z reveals how people are spending money on AI |
|
Andreessen Horowitz’s seventh Top 100 Gen AI Consumer Apps report adds U.S. card-spending rankings, revealing how usage and revenue increasingly tell different stories 4.5% of eligible U.S. consumers paid for ChatGPT, Gemini, or Claude, while the top 1% of AI spenders averaged $903 monthly Subscriptions still dominate AI monetization, while power users favor creative and builder tools and personal agents emerge as the next major consumer battleground |
⚡ Meta and Microsoft cut back internal Claude use |
|
Meta’s internal Claude Code users reportedly dropped from about 60,000 earlier this year to roughly 30,000 as it promotes its own AI coding tools Microsoft also reduced projected internal Anthropic spending by more than a third while steering employees toward GitHub Copilot and OpenAI models Both companies still support Claude, but the shift shows Big Tech increasingly prioritizing its own AI products for internal work, cost control, and development |
|
|
|
|
|
| |
200+ Proven Ways to Make Money With AI in 2026 |
|
The next wave of millionaires will be people who figured out how to make AI work for them. |
The window to get ahead is still open. But not for long. |
Here are 200+ proven ways to make money with AI in 2026. |
Sign up for Superhuman AI, the free daily newsletter read by 1M+ professionals, and get instant access to all 200+ ways to profit from AI this year. |
Visit Superhuman Academy (original giveaway availability not verified) |
|
|
|
|
|
| |
More News on the Radar |
Instinct brings its AI agent to group chats, letting friends join and use it even without creating their own Instinct accounts |
Reflection unveiled Beam, an open-weight AI model it says rivals top Chinese models while using far less inference compute |
Petlibro’s Granary 2 feeder uses AI to recognize multiple cats, track eating habits, and manage personalized portions |
|
|
|
|
|
| |
⚡ Steal This Workflow |
Make ChatGPT Dot Run Your Morning Work Brief |
|
Stop starting your day by checking five different places. |
Give ChatGPT Dot the sources you already use and have it prepare one short work brief before you begin. |
-
Connect the sources Dot needs
Create your Dot in ChatGPT on desktop or web.
Connect the accounts you want it to use, such as your email and calendar. Then give it permission to the project files, notes, or other sources relevant to your work.
Keep the first version narrow. You can always add more context later.
-
Run the brief manually first
Before automating anything, ask Dot to prepare one morning brief.
Use:
“Review my calendar, relevant email, and approved project sources. Give me a morning work brief containing meetings, deadlines coming up in the next 7 days, important updates since yesterday, anything blocked or waiting on someone, decisions I need to make, and the three things that deserve my attention first. Keep it concise and tell me when something cannot be verified.”
Check the result for missing context, noisy updates, or priorities that do not actually matter.
-
Tell Dot what deserves your attention
Refine the instructions around how you work.
For example:
“Prioritize deadlines, client issues, blocked work, unanswered requests, upcoming meetings that need preparation, and decisions waiting on me. Ignore newsletters, routine notifications, FYI messages, and anything that requires no action.”
You can also tell it which projects, people, or calendars matter most.
This is what turns the brief from another summary into something useful.
-
Schedule it every morning
Once the one-off version works, ask:
Run this morning work brief every weekday at 8:30 AM.
Dot can run recurring work without waiting for another message and return the result directly to the conversation.
|
|
|
|
|
|
| |
🔎 AI RESEARCH: AI Teams Beat Human Teams at Creativity
|
|
What researchers studied: |
Researchers compared teams of AI agents with human teams across six creative problem-solving tasks. The study analyzed 4,541 ideas generated by multi-agent AI teams and 341 ideas from human teams to see which groups produced more creative solutions. |
What they found: |
The AI teams substantially outperformed the human teams on creativity. Their advantage came mainly from producing more novel ideas, while their ideas remained about as useful as those created by humans. |
The researchers also found that creativity improved when conversations explored a wider range of ideas instead of staying focused on one theme. AI teams benefited from covering more semantic ground efficiently, while human teams performed better when their discussions flowed smoothly and changed direction naturally. |
Why it matters: |
The findings suggest AI collaboration may be useful for more than automating routine work. Teams of AI agents could become powerful brainstorming partners for product ideas, research, strategy, design, and other work where generating genuinely different options matters. |
Source: Microsoft Research / arXiv, May 18, 2026 |
|
|
|
|
|
| |
🛠️ Tools Worth Watching |
Runway Ads: Creates, publishes and learns from paid ads automatically Holo4: Computer-use AI that clicks, codes and connects tools Google CC: Shared AI agent managing a family's everyday logistics |
|
|
|
|
|
| |
Free Workshop: Grow Your Newsletter to $10k/Month in 90 Days |
|
|
We’ve driven over $10M in revenue for our newsletter clients. For the first time, we’re unpacking the plug-and-play system thousands of people are using to turn a small email list into $2k-$10k/month. |
If you've tried to grow your email list’s revenue and hit a wall of vague advice like 'post more' or 'engage your audience,' this workshop is built to show you precisely what is working for the top newsletter operators in the world. |
We’ll show you how to pick or validate your niche, scale to 1,000 readers, and make your first $10,000 in revenue. |
No prior experience, no camera time, and no need to already have a big email list. |
Original workshop promotion — registration link unavailable |
|
|
|
|
|
| |
📝 Prompt Radar: Make AI Work Through a Task With You
|
What it does: Makes ChatGPT guide you through a task one step at a time instead of jumping straight to the answer. |
Work through this task with me instead of completing it. Start by asking the single most important question, then wait for my answer. Use each response to decide the next best question or action, and move forward one step at a time. If something is unclear, ask instead of assuming. When there are multiple options, briefly explain them and let me choose. Keep track of our decisions, and create the final result once we’ve worked through everything necessary. Task: [DESCRIBE YOUR TASK].
|
Best for: |
Difficult decisions Planning unfamiliar tasks Learning while doing |
|
|
|
|
|
| |
💼 AI Side Hustle Idea: AI Scope Creep Detection Service
|
|
Agencies don’t always lose profit through bad projects. Sometimes it disappears through dozens of “small” client requests that were never included in the original agreement. |
The problem:
Clients ask for extra revisions, additional pages, deliverables, or “quick changes” through email, Slack, and meetings. Teams often complete the work without checking the original scope, turning hours of extra work into unpaid work. |
Your AI service:
Build a system that compares incoming client requests with the approved proposal, contract, or statement of work. AI flags requests that may fall outside the agreed scope, shows what was originally included, and prepares a draft change request for the team to review. |
Simple workflow:
Client agreement → incoming requests → AI compares against scope → flag possible extra work → human review → approve work or send change request |
Best clients:
Marketing agencies, design studios, development agencies, consulting firms, video production teams, and other service businesses managing multiple client projects. |
What you can charge for:
One-time setup and workflow configuration + monthly scope-monitoring and maintenance retainer. |
|
|
|
|
|
| |
Learn AI with This Course |
|
|
What you’ll learn: |
Learn how to create AI-powered videos without filming, turn scripts into finished videos, customize AI voices, and produce content for platforms like YouTube, Instagram, and TikTok. |
Key topics: |
Creating and editing complete videos with AI using InVideo AI Generating scripts, setting up AI voices, and creating voice clones Producing YouTube Shorts and social media videos without filming |
|
|
|
|
|
Until next week,
AI Radars |
|
