|
Google Cloud launched the Gemini agent, a single universal agent for enterprise work.
It routes tasks across Gemini and Anthropic’s Claude models behind the scenes. Who gets blamed when the router picks wrong?
Today in the “Incredible AI” world:
- Google Cloud Unveils Gemini Work Agent
- Anthropic Offers OSS Scanner For Free
- Thiel Ties AI Regulation To Antichrist
- How to Have Crystal Clear Calls Using AI
- 5 Super-Useful AI Tools You Should Try
- Top AI-Generated Image of the Day
Read Time: 4 minutes
|
|
|
Google Cloud Builds One Gemini Agent:
Google Cloud just launched the Gemini agent, one universal agent for enterprise work. It answers questions, writes code, creates media, and routes tasks between Gemini and Anthropic’s Claude. One box rules.
Things to Know:
- Coworker Agents: Workspace coworker agents receive their own email, calendar, and Drive accounts. Colleagues @mention them, and edits appear under their names.
- Memory And Stamina: The agent maintains four distinct memory types and writes reusable skills itself. Assignments continue running for days after laptops close.
- Spend Caps: Administrators set hard per-project spending limits inside Cloud Billing Console. Once triggered, the agent freezes until somebody resumes it.
- The Missing Pieces: Google shared zero benchmarks, pricing, or availability dates. Buyers get customer anecdotes, like Bloomberg Media’s 63% SQL accuracy improvement.
Enterprise competition is shifting from whose model is smartest to whose agent runs the office. Google just admitted the brain can come from a rival lab. The memory, permissions, and invoice stay home.
|
|
|
Anthropic Launches Free Security Scanner:
Anthropic just launched OSS Scanner, a free opt-in service that hunts security bugs in open source projects. Its strongest models, Claude Mythos included, run every scan. No human double-checks the reports.
Things to Know:
- How It Works: Opted-in projects get thorough, periodic security scans from Anthropic’s strongest models. Anthropic covers the entire bill, so maintainers pay nothing.
- The Catch: Reports skip human review and triage entirely, which speeds scanning considerably. Anthropic openly admits findings could be incorrect or invalid.
- Proven Track Record: AI tools already uncovered Copy Fail, a vulnerability affecting nearly every Linux distribution in May. Machine-found bugs are definitely real.
- The Inbox Problem: Maintainers, including Linus Torvalds, are already struggling with floods of AI-generated bug reports. Even Google reportedly buckled under the pressure.
Security work is turning into a volume game, and the bottleneck is now human attention. Finding bugs got cheap overnight. Deciding which reports deserve a busy maintainer’s evening is still painfully expensive.
|
|
|
Peter Thiel Links AI Regulation To Antichrist:
Peter Thiel is now telling paying audiences that government attempts to regulate AI are evil. His Nashville lecture series casts AI safety advocates as enablers of the antichrist. Tickets cost $100 each.
Things to Know:
- The Tour: Thiel’s antichrist lectures began in San Francisco, then toured Paris and Rome. Nashville currently hosts a four-part series, tickets $100.
- The Targets: Thiel targeted Barack Obama, Pope Leo XIV, and effective altruism. He claimed effective altruists would create one-world governance blocking AI.
- The Money: Gizmodo argues Thiel’s fortune depends heavily on AI. Holdings include Palantir, Founders Fund startups, and Thiel Macro’s $418 million portfolio.
- The Backlash: Vatican advisor Paolo Benanti’s essay openly asked whether Thiel deserves burning at the stake for heresy. Medieval energy, modern billionaire.
The fight over AI regulation is no longer fought only with policy papers and lobbying budgets. Some of its loudest voices now argue in theological terms. That makes compromise considerably harder to find.
|
|
|
Let AI Fix Your Audio Before Your Next Meeting:
You are in the middle of an important client call and your neighbor decides that is the perfect time to start drilling. Or the dog barks. Or someone in the background is having their own full conversation. We have all been there, and it is not a great look.
​Krisp is a Voice AI tool that removes background noise in real time, takes notes during your meetings, and converts accents for clearer communication. It works with every calling app you already use, Zoom, Google Meet, Teams, Slack, and anything else.
Here’s how to use it:
- Download and install Krisp: Sign up for free and install the desktop app on Mac or Windows. There is also a Chrome extension if you primarily use Google Meet in the browser. Setup takes a couple of minutes and no technical knowledge is needed.
- Select Krisp as your microphone and speaker: Once installed, go to your calling app’s audio settings and select Krisp as your microphone and speaker. That is all the configuration required. Every call you make from that point runs through Krisp automatically.
- Let noise cancellation handle the rest: Krisp filters out background noise, echo, and cross-talk in real time on both ends of the call. The person you are speaking to sounds cleaner too, not just you. You can toggle it on and off mid-call if you ever need to.
- Use Accent Conversion for clearer communication: If you or someone on your team speaks with a strong accent that sometimes causes misunderstandings, Krisp’s Accent AI can convert your voice to a clearer, more neutral accent in real time without changing who you are or how you naturally speak.
- Get AI meeting notes automatically: Krisp’s built-in AI note taker records, transcribes, and summarizes your calls as they happen. After the meeting, you get a clean summary with action items ready to share or push to Slack, HubSpot, or Salesforce.
If you spend a good chunk of your week on calls and still deal with audio problems or scrambled notes afterward, Krisp takes both of those off your plate.
|
|
|
- daily poll and result - |
|
|
Today's Poll:
| Q) Will AI reduce the need for personal brands? |
|
|
|
Vote and find out the result tomorrow.
Yesterday's Result:
Q) Is model centralization dangerous long-term?
- A) Yes, power concentrates fast - 91% đź‘‘
- B) No, competition balances it - 9%
|
|
|
​Abridge: Convert complex conversations into structured summaries and key takeaways in real time.
​Vana: Create a personalized AI trained on your own data and digital footprint.
​Cogram: Automatically generate meeting notes and draft follow-up emails for technical teams.
​Palette.fm: Colorize black-and-white photos realistically using AI.
​Krisp Call Summary: Generate structured call summaries and action items automatically.
|
|
|
Neo Defies Physics on a Skyscraper Rooftop:
Prompt:
​
"Semi-realistic anime illustration of Neo frozen in a bullet-time dodge on an open high-rise rooftop, upper body arching backward toward the camera, knees bent, feet planted behind him. Long black leather coat, narrow Matrix sunglasses. Two or three crisp bullets with thin trails fly from a distant suited gunman. Overcast daylight, hazy city skyline, premium key-art detail and refined linework."
|
|
|