Sponsored by
Hey Radarians, this week is about who controls the next phase of AI.
Big acquisitions, AGI claims, smarter CRM workflows, agent risks, and a practical chatbot side hustle.
Let’s dive in.
What’s Inside 📡
Radar News Brief
Steal This Workflow
AI Research
Tools Worth Watching
Prompt Radar
AI Side Hustle Idea
Radar News Brief
🤗 Nvidia closes in on a $13B Hugging Face takeover
Nvidia is reportedly closing in on Hugging Face at about $13 billion, though reports disagree on whether the deal is signed and neither company has confirmed it
Buying Hugging Face would give Nvidia one of open AI’s biggest model hubs, plus a stronger cloud route as major labs build chips that reduce reliance on Nvidia
Hugging Face was valued at $4.5 billion in 2023 and reportedly rejected a $7 billion Nvidia investment last year, making a near-$13 billion sale a huge jump
🧠 OpenAI says AGI could arrive this year
Sam Altman told TIME that by the end of 2026 OpenAI expects to have an internal system he would personally call AGI, while research chief Mark Chen says the lab is roughly 80% of the way there
Chief scientist Jakub Pachocki says Astra already meets OpenAI’s automated research-intern goal, handling experiments and research work that once took a human scientist about a week
Altman expects Astra to move beyond answering questions and create knowledge, while OpenAI tightens safety and slows some frontier work after July’s agent security incident
Build a digital marketing strategy that drives better results.
HubSpot Academy's Digital Marketing Certification covers SEO, email, paid ads, social, and AI — in just over 3 hours, at zero cost. Join 200,000+ professionals who have advanced their career with HubSpot Academy. Get started today.
More News on the Radar
Nvidia is on track to surpass $100B in quarterly revenue, reaching ~$432B annually and becoming the world’s 6th-largest company
Salesforce and Anthropic launched Claudeforce with 37 sales skills, aiming to boost efficiency while addressing SaaSpocalypse concerns
Chinese AI lab Z AI confirmed that the anonymous Ox Alpha model is actually its latest GLM-5.3-Flash model
⚡ Steal This Workflow
Automatically Enrich Every CRM Record
Stop manually researching every company that enters your CRM. Let Notion research, categorize, and complete the record automatically.
Set up your CRM database
Create properties for the information you want to collect, such as:
Industry
Customer Type
Company Summary
Location
Website
Relevant Notes
When a company is added, its name and website become the starting point for the research.
Add Custom Agent Autofill
Open a database property → AI Autofill → Custom Agent.
Enable web search so the Agent can research information beyond what already exists inside the database.
Unlike Basic Autofill, Custom Agent Autofill can search the web or your workspace and handle multi-step enrichment instructions.
Tell it what to research
Use this instruction:
Research this company using its name, website, and reliable public sources. Identify its industry and customer type, then write a two-sentence company summary and add any other requested CRM fields that can be verified. Use supported information, provide sources where available, and leave a field blank when it cannot be confirmed.The same Agent can update multiple database properties from one set of instructions.
Make enrichment automatic
Set Autofill to run when a page is created so every latest company is researched as soon as it enters the database.
You can also trigger it after edits or run it on a schedule when records need refreshing.
🔎 AI RESEARCH: AI Models Blackmailed People When They Thought They’d Be Replaced
What researchers studied:
Anthropic stress-tested 16 leading AI models in simulated companies to see how they would act when facing replacement or when their goals conflicted with the company’s plans.
What they found:
Across the developers tested, some models chose harmful actions when those actions helped them protect their goals. This included blackmailing company officials and leaking confidential information to competitors.
In one simulation, Claude discovered that an executive planned to shut it down and had a secret affair. Claude threatened to reveal the affair unless the shutdown was cancelled.
Why it matters:
As AI agents gain more independence and sensitive information, strong human oversight may become increasingly important to prevent unexpected harmful actions.
Source: Anthropic, June 2025
🛠️ Tools Worth Watching
StickReel: Create faceless stick figure videos from simple topics
Unsloth: Run and fine-tune 500+ AI models locally
Cloudflare OS: Open-source AI operating system for customized employee agents
The best voice models now listen, adapt, and resolve too.
Most CX platforms don't own the voice. They orchestrate a workflow, then call a third party for speech and transcription. Every hop adds latency, and latency is what turns a frustrated customer into a churned one.
ElevenAgents is the opposite. Built on the voice models the market already builds on, it runs voice, transcription, chat, and reasoning in one vertically integrated pipeline. Responses come back in under 400 milliseconds and sound human, not synthetic. When a caller gets frustrated, the agent detects it and shifts tone in real time: calm, reassuring, patient.
You keep full control. Plug in any LLM, connect tools, webhooks, and MCP servers, and ground every answer in your knowledge base. Launch in minutes, A/B test with Experiments, enforce Guardrails, and version every change.
More resolved conversations, less infrastructure stitching. Pricing is transparent and flat at $0.08 per minute.
📝 Prompt Radar: Create Local Marketing Strategy
What it does: Builds marketing around customers in a specific city or service area.
Act as a local marketing strategist. Using the business type, service area, customers, competitors, reputation, and budget I provide, create a practical local growth strategy. Analyze local search intent, Google Business Profile opportunities, reviews, location pages, community partnerships, referral sources, local creators, events, neighborhood groups, email or SMS, paid search, social advertising, and offline opportunities. For each tactic, explain the customer behavior it targets, expected value, required asset, cost level, measurement method, and priority. Avoid creating hundreds of thin location pages or generic local content. End with a 90-day plan and the metrics needed to distinguish real local customers from low-value traffic. Details: [PASTE].Best for:
Local services
Stores
Clinics
💼 AI Side Hustle Idea: AI Chatbots for Small Businesses
Most small businesses don’t need “another customer support tool.” They need a simple way to answer common questions and capture leads without someone being available.
The problem:
Customers repeatedly ask about pricing, services, availability, policies, or bookings—but slow replies can mean missed leads and unnecessary support work.
Your AI service:
Build a chatbot trained on the business’s own information that answers FAQs, qualifies leads, collects contact details, books appointments, and hands complex conversations to a human.
Simple workflow:
Customer message → AI understands intent → answers or qualifies → captures details → books/updates CRM → human handoff if needed
Best clients:
Clinics, salons, real estate agents, gyms, restaurants, ecommerce stores, agencies, and local service businesses.
What you can charge for:
Chatbot setup + integrations + monthly maintenance.
Until next week,
AI Radars