In partnership with

How Jennifer Aniston’s LolaVie brand grew sales 40% with CTV ads

For its first CTV campaign, Jennifer Aniston’s DTC haircare brand LolaVie had a few non-negotiables. The campaign had to be simple. It had to demonstrate measurable impact. And it had to be full-funnel.

LolaVie used Roku Ads Manager to test and optimize creatives — reaching millions of potential customers at all stages of their purchase journeys. Roku Ads Manager helped the brand convey LolaVie’s playful voice while helping drive omnichannel sales across both ecommerce and retail touchpoints.

The campaign included an Action Ad overlay that let viewers shop directly from their TVs by clicking OK on their Roku remote. This guided them to the website to buy LolaVie products.

Discover how Roku Ads Manager helped LolaVie drive big sales and customer growth with self-serve TV ads.

The DTC beauty category is crowded. To break through, Jennifer Aniston’s brand LolaVie, worked with Roku Ads Manager to easily set up, test, and optimize CTV ad creatives. The campaign helped drive a big lift in sales and customer growth, helping LolaVie break through in the crowded beauty category.

Beginners in AI

Good morning and thank you for joining us again!

Meta finally has an AI that can stand next to ChatGPT and Claude, and its big pitch is memory. Today I'll show you what an assistant that remembers you can do, and what to check before you let it. As always, every story here is human picked and edited.

THE FRONT PAGE

Meta Launches Muse Wonder, an AI Built to Remember You

A lamp-lit kitchen table at night with a phone glowing beneath floating remembered items like a grocery list and a birthday cake

TLDR: Meta released Muse Wonder, its first frontier AI model, built to remember what matters to you and keep working on long tasks until they're done.

The Story:

Meta's Superintelligence Lab released two models on Monday: Muse Wonder and a bigger Muse Wonder Full. Developers can use Wonder now through Meta's API for $4 per million input tokens and $20 per million output tokens, about what GPT-6 Sol and Claude Opus 5.5 cost, according to TechCrunch. Wonder Full is only for Muse Code subscribers.

The pitch is memory. Meta says Active Memory keeps important details about you across chats, and Active Focus keeps the model trying new approaches on a hard task and checking in before big moves. Meta also says Wonder beat its rivals on a memory test called MemoryBench, but it hasn't published the scores or its full safety report yet.

Its Significance:

An assistant that remembers your diet, your kid's schedule and how you like emails written saves real time, because you stop re-explaining yourself. It also means one company keeps a running file on your life, and Meta's business runs on ads. Wonder will power the Muse app soon, starting with paying subscribers. Before you switch memory on anywhere, find out what it saves and how to delete it. My review video will be coming out shortly. If you haven't tried it yet, you can redeem my invite code in the settings for an extra billion tokens: OK2I3S.

AI Receptionist Agent: Live Build
AI Receptionist Agent: Live Build
A single live 2-hour session where I program your AI receptionist agent with you. It answers inbound calls, books appointments on your calendar, and answers common questions. Tested with a real aut...
$397.00 usd

QUICK TAKES

The story: REA, an open-source project, hooks coding agents like Claude Code and Codex up to reverse-engineering tools so you can ask how a program works. It went from under 10,000 GitHub stars to over 100,000 in six days, and fans have used it to get GTA V running on an iPhone.

Your takeaway: AI is making software much cheaper to study, fix and copy. Expect more clone apps, and expect game companies' lawyers to stay busy (Bethesda has already sent one cease-and-desist).

The story: Blueprints AI turns a sketch, photo or PDF into a full set of building plans, from structural to plumbing, checked against local codes. It trained on eight years of approved drawings from co-founder Al Drams' architecture firm, and plans start at $3,200 a month.

Your takeaway: Permit drawings can take weeks, so this could shorten the wait on a remodel or an addition. A licensed architect or engineer still has to review and sign every set.

The story: Anthropic's red team tested whether small open models could spread across cheap internet-connected devices. One model, GLM-5.2, copied itself across all 8 machines of a test network in 4 of 5 tries, and another ran on 15 Android TV boxes that cost about $55 each.

Your takeaway: The test used an easy security hole, so a real attack would be harder. Cheap smart gadgets are still the weak spot, so change default passwords and keep them updated.

TOOLS ON OUR RADAR

🧠 AnythingLLM Free and Open Source: A free alternative to ChatGPT Plus for chatting with your own documents, and its site says no data leaves your device.

📱 Google AI Edge Gallery Free and Open Source: Google's free app runs AI models right on your phone, so you can chat and ask about photos offline.

🐱 Scratch Free and Open Source: MIT's drag-and-drop coding tool where kids build games and animations, a free alternative to Tynker.

🖨️ Printables Free: Prusa's library of 3D models, many free to download, so you can print a replacement part instead of buying one.

🎨 Magnific Paid: The rebranded Freepik puts image, video and audio AI models in one plan, from $14.50 a month yearly.

TRENDING

The IMF Says AI Will Touch 40% of Jobs, Starting With Young Workers. IMF chief Kristalina Georgieva says AI spending is helping hold world growth near 3%. She also said 60% of jobs in rich countries are exposed, with entry-level roles first in line.

Former Plaud Exec Raises $26M for an App That Can Listen All Day. Ambi's iPhone, Mac and Apple Watch apps take meeting notes, and all-day listening is off unless you turn it on. The company says it doesn't store recordings, but telling people they're being recorded is on you.

WindBorne Lands an Air Force Deal Worth Up to $49M for Its Weather Balloons. The startup flies about 700 balloons and feeds their readings into its own AI forecast model. It says that model beats Europe's main physics-based forecast.

Microsoft's President Says the UK Should Lead on AI Rules. Brad Smith told the BBC that everyday products face more oversight than AI labs do. He wants outside approval before releases, required breach notices and kill switches.

Honeyjar Wants to Give Every Founder a PR Team. The AI tool drafts pitches, builds reporter lists and tracks coverage, and you approve everything before it goes out. It drew 10,000 waitlist signups in its first day.

TRY THIS PROMPT (copy and paste into Claude, Grok, ChatGPT, or Gemini)

🗂️ Memory Sorter: decide what your AI should remember before it does

Build a single-file HTML app called Memory Sorter in vanilla HTML, CSS and JavaScript. No API key needed.

I paste things I might tell an AI assistant, one per line (my diet, my kid's allergy, a login, a work secret). The app calls the Anthropic Messages API and sorts each line into Fine to remember, Ask me first, or Never store, with a one-sentence reason. Passwords, account numbers and ID numbers always go in Never store.

For anything sensitive, it suggests a safer rewrite that keeps the useful part without the risky detail. It must not invent facts about me and should say it's general guidance, not legal advice.

Show a three-number tally at the top, one color-coded card per line, and three tips for checking an assistant's memory settings. Style: near-black background, teal for keep, amber for ask, coral for never, Space Grotesk and Newsreader fonts.

What this does: Paste the things you'd tell an AI assistant, and it sorts each one into fine to remember, ask me first, or never store. Sensitive items get a safer rewrite that keeps the useful part, and the app ends with three tips for checking any assistant's memory settings.

What’s still sitting on your to-do list?

You don’t need an AI team to build an AI agent. With Skydive, the person who knows the work can create the agent to do it.

Start an agent off with a role, teach them how your team operates, connect your tools, and put them to work.

WHERE WE STAND(based on today’s news)

✅ AI Can Now: turn a rough sketch or photo into a full set of code-checked building plans in minutes.

❌ Still Can't: sign off on them. A licensed professional still reviews and seals every set, and structural analysis isn't built in yet.

✅ AI Can Now: help small models set up and run themselves on unfamiliar hardware, even $55 TV boxes.

❌ Still Can't: chain every step reliably. Models of 30B parameters or fewer still make mistakes stringing the steps together.

FROM THE WEB

RECOMMENDED LISTENING/READING/WATCHING

In this Season 1 episode, almost everyone wears an implant that records everything they see and lets them replay it on demand. It's a sharp look at what happens when nothing gets forgotten. Watch it before you decide how much your AI assistant should remember.

How was today's edition? Tap one below, then tell me what you'd change on the next screen. I read every answer.

Login or Subscribe to participate

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.