Friday, October 9, 2026
Today I've got a search engine turning into your new office coworker, OpenAI's money numbers getting a reality check, and a $99 ring that lets you boss around an AI with one tap. Let's get into it.
Google Turns Gemini Into a Coworker With Its Own Email Address
Google just gave its AI assistant Gemini a new job description. Instead of just answering questions, Gemini can now plan projects, carry out multi-step tasks, and work across different business apps on its own, starting with companies first. It can even hand off pieces of work to smaller helper AIs, called subagents, and it gets its own workplace identity, complete with an email address, just like a new hire would. This is Google's answer to the wave of AI agents from OpenAI, Meta, and others that are moving from chatbots you talk to, into assistants that go do things for you.
What this means for you: If your company uses Google Workspace, expect an AI helper that can actually finish tasks for you soon, not just draft a reply.
What this means for your business: Every major tech company is now racing to put an AI worker inside your business software. Start thinking now about which tasks you would trust an agent to run end to end.
Source: TechCrunch
Radar 01
OpenAI's Revenue Is Reportedly $20 Billion Lower Than Thought
For months, reports pegged OpenAI's yearly revenue pace at around $70 billion. A new report says the real number is significantly lower, by about $20 billion. OpenAI has not confirmed the figure, but the gap matters because the company's eye-watering valuation and massive spending plans assume rapid revenue growth. If the real numbers are smaller, some of the AI industry's biggest financial bets may be resting on shakier ground than investors assumed.
What this means for you: Don't assume every AI company's growth numbers are as rock solid as the headlines suggest. Ask for real evidence before trusting big claims.
What this means for your business: If you're relying on an AI vendor's stability for a long-term bet, revisit your due diligence. Revenue gaps like this can affect pricing, support, and how long a product sticks around.
Source: TechCrunch
Radar 02
China's Manus Raises Over $500M After Splitting From Meta
Manus, a Chinese AI agent company, has raised more than $500 million in its first funding round since parting ways with Meta. Boyu Capital and IDG Capital led the round, with Tencent and other major investors also chipping in. The deal shows that huge sums of money are still flowing into AI agent startups outside the US, even as some American AI companies face tougher questions about their own finances.
What this means for you: AI competition is global. The best new tools for your business might come from companies you've never heard of yet.
What this means for your business: Keep an eye on international AI vendors, not just the big US names. Pricing, features, and speed of innovation can look very different outside Silicon Valley.
Source: TechCrunch
Radar 03
Anthropic Updates Its Rules: No Election Meddling, No Abusing Claude
Anthropic has rewritten the usage policy for its Claude chatbot for the first time in over a year. The new rules explicitly ban using Claude to interfere with elections, run deceptive campaigns, build weapons, or conduct surveillance. They also ban users from repeatedly, needlessly abusing the chatbot in extreme cases, though normal criticism or frustration is still fine. It's a sign that AI companies are tightening the rulebook as these tools get used for more sensitive, high-stakes work.
What this means for you: If your team uses Claude for anything touching elections, security, or surveillance, check the new policy before you build on it.
What this means for your business: AI usage policies are tightening across the industry. Build a habit of reviewing vendor terms of service regularly, not just once at signup.
Source: TechCrunch
Try This Today
Pick one repetitive task in your business, like sorting customer emails or drafting weekly reports, and ask: would I trust an AI agent to do this from start to finish without checking every step? Write down what would have to be true for the answer to be yes. That list becomes your roadmap for adopting agentic AI safely.
Quick Hits
- The company behind the popular LMArena AI leaderboard just raised $200 million, nearly doubling its valuation to $3.1 billion in only 10 months. It now also tests AI models on honesty, including whether they lie. [1]
- A startup called Natura launched a $99 smart ring that lets you summon an AI agent with a tap of your finger to capture notes, run tasks, or control devices, while also tracking your health. [2]
- French AI lab Mistral released a new open-weight model nicknamed Le Chonk, claiming it can go toe to toe with the best closed, paid AI models while staying free for anyone to download and modify. [3]
Enjoyed today's read? Help me get this in front of more people like you.
Start every morning a little sharper.
Free. Takes 5 minutes to read. No spam.
