In partnership with
It's time to set up your holiday streaming campaign
Are you ready to reach holiday shoppers on CTV? For many streamers, the season is already underway.
Last year, according to Roku internal data, holiday content searches on Roku saw an 86% increase during October. And in November, those searches rise an average of 39% weekly. This surging holiday spirit is mirrored by consumer activity, as about half of Roku streamers shopped Black Friday and Cyber Monday deals in 2025.
Meeting those shoppers is easier than you might think. Begin with the seasonal outcome you hope to drive: awareness, online sales, leads, store visits, or another measurable goal. From there, define your audience, upload your creative, set the budget and campaign dates, and establish measurement before launching. You can even create a shoppable ad unit that lets streamers buy with a click of their remote.
Our holiday playbook offers a month-by-month guide to reaching shoppers during this all-important time of year.
First-time spenders get a $5,000 match on the first $5,000 spent with code BEEHIIV5K. Terms apply.
Beginners in AI
Good morning and thank you for joining us again!
Signal's encryption guards private chats for a lot of people, and Google's AI just spent four weeks proving key parts of it hold up. Today's lead explains what that proof covers, what it doesn't, and what it means for the apps you text with. As always, every story here is human picked and edited.
THE FRONT PAGE
Google's AI Wrote 6 Million Lines of Proof to Check Signal's Encryption
TLDR: Google DeepMind says its new AlphaProof Code system proved key security promises in the code behind Signal's encryption in about four weeks, a job it says used to take expert teams years.
The Story:
Signal's encryption protects private chats in the Signal app and in Google Messages. DeepMind pointed a team of AI agents at the real code that runs it, not just the design on paper. Human experts chose what had to be proven, like whether a stolen key can unlock old messages. Then the agents wrote the proofs in Lean, a math language a computer can check line by line.
The result, per DeepMind: over six million lines of machine-checked proof, with about 1% of the code flagged for a person to review. As a test, the team slipped two old, already-fixed bugs back into a copy of the code. Both times, the system proved that version was unsafe. Signal CTO Ehren Kret called turning live code into checkable theorems "a major breakthrough."
Its Significance:
Most software gets tested by trying lots of inputs and hoping nothing slips through. A proof covers every case the experts wrote down, and a computer double-checks the math, which is why it's usually saved for small, critical code. Keep the limits in view: it covers selected security properties, not the whole Signal app, it assumes the underlying cryptography is sound, and it found no new bugs. DeepMind is open-sourcing the proofs so outside experts can check its work.
JOIN THE COMMUNITY
Learn AI With Us
Most people try AI once, get a so-so answer, and quit. Our 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 want to 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. We just released one for older adults, with a gentler progression in the lessons.
QUICK TAKES
The story: Google says Gemini Live is rolling out worldwide to iPhone users on CarPlay, after earlier launches on Android Auto. Tap the Gemini icon on your car's screen and talk: ask what your day looks like, have it sum up an email thread, or find lunch nearby.
Your takeaway: Update the Gemini app, allow the microphone, and use Gemini Live once on your phone before you drive. Chats sync back to your iPhone so you can reread them after you park, though Daily Brief and some personal features aren't offered to under-18s or in the EEA.
The story: Bloomberg reports, citing unnamed sources, that OpenAI's third-quarter revenue topped $10 billion, up about 50% from $6.7 billion the quarter before. It's still losing money, though its adjusted operating margin improved from -150% to -109%.
Your takeaway: That growth is coming from pricier plans, business tools and AI agents, so expect more paid tiers and upsells inside ChatGPT. For scale, Anthropic, which makes Claude, reported $17 billion for the same quarter, but the two companies count revenue differently.
The story: The new model reads a million letters of DNA and predicts how one small change affects a specific type of cell, after training on over 100 million cell profiles. In DeepMind's test across 28 kinds of immune cells, it pinned down the likely culprit behind a gene change 30% of the time, against 16% for the best earlier tools.
Your takeaway: It's a research tool, not a test your doctor can order, and DeepMind says it isn't approved for diagnosis. Still, helping scientists sort which of thousands of DNA differences drive diseases like inflammatory bowel disease means fewer dead ends in the lab.
TOOLS ON OUR RADAR
🔐 Signal Free and Open Source: Private texts and calls with end-to-end encryption, the same code today's lead says Google's AI checked.
🕵️ Have I Been Pwned Free: Type in your email to see which of 1,051 hacked websites leaked it, then get alerts.
🌿 Seek by iNaturalist Free: Point your phone at a plant, bug or bird and it names it, with no account needed.
📄 PDF24 Tools Free: Merges, shrinks, converts and edits PDFs in your browser with no sign-up and no watermarks.
🗣️ ElevenLabs Freemium: Turns your text into lifelike speech, with 10,000 free credits a month and paid plans from $6.
✍️ Typefully Freemium: Writes and schedules posts for X, LinkedIn, Threads, Bluesky and Mastodon from one screen, with a free plan.
TRENDING
Nvidia's Board Is Reportedly Weighing Whether to Oust CEO Jensen Huang. The Wall Street Journal reported Wednesday that the board was preparing to decide the future of Huang, who has run the AI chip giant for 33 years, after he admitted an affair with a woman U.S. intelligence reportedly believes is a Chinese spy. Nvidia's stock fell about 8% after hours, and the company has no obvious successor.
ChatGPT Retires Its Old Voice Modes on Dec. 11. Advanced and Standard Voice go away and everyone moves to Live voice, which OpenAI says has nearly doubled voice use since it launched in July. Old voice chats stay in your history, and video and screen sharing keep running on Advanced Voice until Live supports them.
Claude Adds Shared Sheets and Whiteboards on All Paid Plans. Anthropic says Claude can now build a spreadsheet with tabs and formulas or turn a plan into a whiteboard your team edits together, both in beta on web, desktop and mobile. If you've lost track of everything Claude added this year, our month-by-month timeline of what's new in Claude lays it all out.
Spotify's AI DJ Gets Nine New Voices and a Quieter Mode. Premium listeners can pick from seven English voices and two Spanish ones, or switch off the chatter between songs. Spotify says over 94 million Premium listeners use the DJ; search "DJ" in the app and tap the three dots to change it.
Tether's Tiny AI Agent Beat a Model 2.5 Times Its Size at Phone Tasks. The 800-million-parameter model, called QVAC BotPsy, scored highest on five of six tool-use tests in Tether's own testing and is free to download. A model this small can run on the phone itself, so requests don't have to leave your device, though no consumer app uses it yet.
Apple Is Testing Harbor, an App That Turns Macs Into a Private AI Server. The early-access app from Apple's Claris unit downloads open AI models and runs them on your own Mac, and can pool memory across Macs linked by Thunderbolt 5. It isn't public yet and pricing is unknown; free apps like LM Studio already do the one-Mac version.
TRY THIS PROMPT (copy and paste into Claude, Grok, ChatGPT, or Gemini)
🔏 Who Can Read It?: Type in any app you trust with private stuff and see who else can see it
Build a single-file HTML app called Who Can Read It? in vanilla HTML, CSS and JavaScript. No API key needed.
I type an app I trust with private stuff (like WhatsApp or Google Photos) and pick what I keep in it: messages, photos, notes, AI chats or backups.
The app calls the Anthropic Messages API with the web_search tool to check the app's official privacy and security pages plus independent reporting. It must never invent facts, audits or links, and says "unknown" when sources don't confirm something.
Show a one-line verdict, whether it's end-to-end encrypted by default, a "who can see it" list (the company, police with a court order, AI training, your backups) with clear badges, any independent security checks, three settings to change today, and real source links.
Style: near-black background, lime and lilac accents, Bricolage Grotesque, Literata and JetBrains Mono fonts.What this does: Type in an app you use for private things, like the messaging app in today's lead. It searches the company's own privacy pages and outside reporting, then tells you whether your stuff is end-to-end encrypted, who else can see it, and three settings worth changing, with links to every source.
Sales Owns Retention, Not Just Closed Deals
Pebl’s Chief Business Officer, Dutta Satadip, joins The GTM Garage to show how Pebl merged Operations and GTM under one leader, and how it aligned goals and incentives across the customer lifecycle.
You’ll also see how Pebl uses AI across that lifecycle and the stack behind prospecting, pipeline, and handoff to onboarding.
15 min conversation. 15 min open Q&A. For Seed to Series A teams.
WHERE WE STAND(based on today's news)
✅ AI Can Now: Write millions of lines of computer-checked proof that the code behind Signal's encryption keeps key security promises, in about four weeks.
❌ Still Can't: Decide what's worth proving. People wrote the rules being checked, and a rule written too narrowly can let a bug slip past.
✅ AI Can Now: Predict how a single DNA change affects one specific type of immune cell, after training on over 100 million cell profiles.
❌ Still Can't: Get it right most of the time. In DeepMind's own test it found the likely cause 30% of the time, and it isn't approved for diagnosis.
FROM THE WEB
RECOMMENDED LISTENING/READING/WATCHING
📚 The Code Book by Simon Singh - Book
Singh's 1999 history of secret codes runs from Mary, Queen of Scots' doomed coded letters to the public-key math that protects your texts today. He explains it all with no math background needed, including why a code is only as safe as the proof that nobody can break it. Read it after today's lead and the Signal story will make a lot more sense.
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.
