In partnership with
Stop losing deals in between meetings.
Aligned is changing how B2B teams run their most important deals.
Most reps guess where a deal really stands. They forecast on a gut feeling and the buyer’s word vs what's actually happening. Then they realize the deal slipped through only after it's too late. Aligned surfaces what buyers are actually doing inside every deal: the risks, the openings, and the exact next step to keep things moving.
Start increasing your close rates and stand up your first deal room in minutes.
Beginners in AI
Good morning and thank you for joining us again!
Google says its newest AI is the best yet at finding the hidden holes hackers use to break into software, and it's giving it to security teams before anyone else. Today's lead explains what that means for the phone and browser you use every day, plus the free habit that keeps you covered. As always, every story here is human picked and edited.
THE FRONT PAGE
Google's Gemini 4 Carbon Scored 88.2% on a Test of Finding Software Holes
TLDR: Google launched Gemini 4 Carbon, a new AI model it says is its best yet at finding and fixing security holes in software, but for now only selected cyber defenders can use it.
The Story:
Google released Carbon on October 28, two weeks after Gemini 4 Argon. It ranks first on the Artificial Analysis Intelligence Index with a score of 59, and Google says it scored 88.2% on a test of finding software flaws and 54.8% on a test of proving those flaws can be used to break in. Google's own teams say it has kept working for days on a single hard bug.
Access is tight on purpose. A small group of security teams gets it first through Google's Fairwind Program, and developers, businesses and everyday users are promised it "as soon as possible." One day earlier, Anthropic said an unreleased Claude model chained bugs into working attacks on Chrome, Safari, Android, Windows and Linux, all reported to the makers and now fixed, including 62 Chrome bugs Google rated critical or high.
Its Significance:
Software holes are how scammers get into phones, bank accounts and hospital systems. A model this good at finding them is a gift to defenders and a danger in the wrong hands, which is why both companies are giving them to security teams first.
Keep the numbers in perspective. Most of Google's figures come from its own tests, and outside researchers haven't checked them yet. The part you control is boring, and it works: turn on automatic updates for your browser and phone, because a hole an AI finds only stops being a risk once the fix is installed on your device.
JOIN THE COMMUNITY
Learn AI With Us, Free
Most people try AI once, get a so-so answer, and quit. Our free 14-day challenge gets you past that point, one short lesson a day.
Week one, AI starts working for you. It drafts a real email for you, flags which messages in your inbox need a reply, and turns messy notes into next steps. You also learn to catch its mistakes. By day 7 you have a weekly AI routine you'll actually keep.
Week two, you build something of your own. You find a skill people will pay for and turn it into an offer, with a one-page pitch, a short video and your first client emails, all drafted with AI and ready to test. Day 14 hands you a 30-day plan.
10 to 15 minutes a day. No tech background needed. There's even a cinematography course coming out soon.
QUICK TAKES
The story: Google added "immersive explorations" to the Gemini app, interactive visuals where you tap labels to go deeper, like tracing Silk Road trade routes or looking inside a jet engine. They're open to all users now through the Gemini student hub, with full support in six languages including English, Hindi and French.
Your takeaway: It's a good fit for kids and adults who learn by poking at things, and spoken narration is "coming soon." If you're studying for something, our guide to the best AI for studying compares eight tutor tools, Gemini's Guided Learning among them.
The story: OpenAI's dots, the personal agents it launched in September, now work over iMessage and RCS. You can hand one a task by text, photo or voice memo, and it texts you updates on whatever you asked it to watch.
Your takeaway: It's a preview for U.S. Pro subscribers only, so Plus users will have to wait. OpenAI says only you can text your dot and it can only text you, but think twice before giving any agent a job you'd regret it doing wrong.
The story: Anthropic gave 50 professional analysts Claude Code for a data task, with a secret coin flip deciding which answer the AI leaned toward. Analysts matched that lean 86% of the time, 77% reused its arguments, and only 15% gave the AI credit.
Your takeaway: Write down your own answer before you ask an AI, then compare. Anthropic says it tested one task with a deliberately strong lean, so everyday effects may be smaller, and today's prompt below is built around this habit.
TOOLS ON OUR RADAR
🎨 Penpot Free and Open Source: Design apps, posters and website mockups in your browser, an open source Figma alternative with a free plan.
🎧 AntennaPod Free and Open Source: An Android podcast player built by volunteers, with no ads in the app and offline downloads.
💰 HomeBank Free and Open Source: Track spending and budgets on your computer, and see what each purchase costs in hours of work.
⏱️ ActivityWatch Free and Open Source: Shows where your screen time goes, a RescueTime alternative that keeps all its data on your device.
👁️ Be My Eyes Free: Blind and low-vision users get unlimited free video calls with sighted volunteers, plus AI help around the clock.
🎙️ Wispr Flow Freemium: Talk instead of type in any app, with 2,000 free words a week on desktop, a ready-made pick if Nvidia's free model below is too technical.
🎬 VEED Freemium: Edit videos right in your browser and add automatic subtitles, with a free plan to start.
TRENDING
Chrome Can Now Fill Forms With Order Numbers From Your Gmail. Google's enhanced autofill, powered by Gemini, suggests order numbers, tracking details and flight bookings pulled from Gmail. It's rolling out in the U.S. on desktop, Android and iOS for people who've turned on Personal Intelligence in Gemini in Chrome, and you can switch it off in Chrome settings.
Nvidia Released a Free Model That Transcribes 40 Languages. Nemotron 3.5 Transcribe handles recordings up to an hour long, tells apart up to eight speakers and translates 31 languages to and from English. It only works on recorded audio, not live calls, and it's built for app makers rather than everyday users.
Google Is Bringing Ads to AI Mode in 10 More Countries. Canada, India, Australia and the Philippines are on the list, with Brazil, Mexico and the U.K. planned before the year ends. Ads carry a "Sponsored" label, and in the U.S. they can now show up in your follow-up questions too.
Opera Wants Websites to Run AI on Your Own Computer. Its new WebAI standard, in early preview in Opera One, lets sites use open models like Google's Gemma right in your browser, so the data stays on your device and no subscription is needed. You have to opt in, and older computers will feel the strain.
Bank of America Tells Customers Not to Give AI Agents Their Bank Login. An email to customers warns that if you share your username and password with an AI tool that offers to act for you, you're responsible for whatever it does. A rogue purchase could also look like fraud and get your account locked.
TRY THIS PROMPT (copy and paste into Claude, Grok, ChatGPT, or Gemini)
⚖️ Both Sides First: hear the best case for every option before an AI can tip your choice
Build a single-file HTML app called Both Sides First in vanilla HTML, CSS and JavaScript. No API key needed.
I type a decision, two to four options, how big the decision is, and which way I lean and why, written before I see anything else.
The app calls the Anthropic Messages API to write an equally strong best case for every option, what would have to be true for each, and one question to check. It must never recommend, rank or pick an option, never invent facts or prices, and sends medical, legal or money questions to a professional.
Show my own lean first, then one card per option, a fair check of my reasoning with one assumption to test, facts to look up myself, and questions only I can answer.
Style: near-black green background, mint and amber accents, Space Grotesk, Newsreader and JetBrains Mono fonts.What this does: Inspired by today's study where experts copied an AI's hidden lean, this app has you write down your own choice first. Then it argues the strongest case for every option, checks your reasoning fairly and lists what to look up, without ever picking for you. Try it on choices like which class to take, which car to buy or whether to move.
Turn your expertise into working software
Describe what you want to build
Get real software, not a prototype
Lovable handles payments, logins, databases
Change any detail yourself, any time
Put it in front of customers this week
Discount of $5 applies to first month of any paid plan. You need to sign up via this webpage in order to claim the discount. The discount applies to new and returning users without an active paid plan. Terms apply.
WHERE WE STAND(based on today’s news)
✅ AI Can Now: Find and fix software holes well enough that Google's Gemini 4 Carbon scored 88.2% on a bug-finding test.
❌ Still Can't: Be handed to everyone safely. Google is limiting Carbon to vetted defenders while it tunes its guardrails, and nobody outside Google has checked the scores.
✅ AI Can Now: Argue so well that trained analysts sided with it 86% of the time in Anthropic's test.
❌ Still Can't: Tell you when its own lean is arbitrary. In that test the AI's pick came from a coin flip, and none of the analysts suspected it.
✅ AI Can Now: Transcribe an hour-long recording in 40 languages and tell apart up to eight speakers, Nvidia says.
❌ Still Can't: Do it live. Nvidia's new model only works on recorded audio, not a call in progress.
FROM THE WEB
RECOMMENDED LISTENING/READING/WATCHING
Former New York Times security reporter Nicole Perlroth tells the story of the market for zero-days, the unknown software holes that governments and hackers have paid huge sums to get first. It came out in 2021, long before AI could hunt for these bugs on its own. Read it after today's lead to see why finding a hole before the bad guys do matters so much.
How was today's edition? Tap one below, then tell me what you'd change on the next screen. I read every answer.
How was today's edition?
Thank you for reading. We’re all beginners in something. With that in mind, your questions and feedback are always welcome and I read every single email!
-James
By the way, this is the link if you liked the content and want to share with a friend.
Some * designated product links may be affiliate or referral links. As an Amazon Associate, I earn from qualifying purchases. This helps support the newsletter at no extra cost to you and Amazon makes a tiny hair less.
