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!
Welcome to this daily edition of Beginners in AI, where we explore the latest trends, tools, and news in the world of AI and the tech that surrounds it. Like all editions, this is human curated and edited, and published with the intention of making AI news and technology more accessible to everyone.
THE FRONT PAGE
Alibaba's AI Caught 15 Cancers Doctors Missed in 10,000 Scans
TLDR: An AI built to do one job, read liver scans, caught 15 cancers that doctors had missed across more than 10,000 patients.
The Story:
Alibaba's DAMO Academy built a model called DAMO LiON that reads liver CT scans and nothing else. Working with Shengjing Hospital of China Medical University and hospitals tied to Zhejiang University, researchers ran it next to real radiologists for two months on scans from more than 10,000 patients. LiON flagged 51 spots the first reports had skipped. Fifteen were cancer. Thirty-seven reports got rewritten, and 22 cases went to a full tumor board. Doctors using it read scans 27% faster and caught 11.5% more cancerous growths, and junior doctors with LiON performed about as well as senior ones. The results ran in Nature Medicine on August 19.
Its Significance:
Most AI news is about models trying to do everything at once. This one can't write your email or plan a trip. It only reads liver scans, and that's why it's so good at reading liver scans. Alibaba already has the same kind of single-job model for pancreatic cancer, stomach cancer, and colon cancer. The next few years of useful AI may look less like one giant brain and more like this: small models trained hard on one problem, cheap enough to run inside a hospital or a shop. LiON is a research system tested at a few sites. You can't ask your doctor for it yet.
QUICK TAKES
The story: Gadot told the Happy Sad Confused podcast her lawyers spent six months on her contract for "Bitcoin: Killing Satoshi," a $70 million thriller shot in an empty warehouse with AI-built sets and lighting. She said the AI handled backgrounds, not her acting, and that SAG called her lawyers asking how to update its own contracts.
Your takeaway: When a union starts borrowing language from one actor's deal, that language shows up in everyone else's within a year or two. If your work involves being recorded, filmed, or voiced, this is likely going to be standard language in your contracts. AI is just going to become another special effect type software over time, like CGI.
The story: Anthropic released Claude Fable 5.1 and Claude Mythos 5.1 on September 1. On a science task benchmark, Fable 5.1 scored 52.6% against the older Fable 5's 24.7%, and re-reading saved text now costs 75% less.
Your takeaway: Those cheaper reads trim about 25% off a typical job and up to 45% off long ones. Base prices didn't move, still $10 per million words in and $50 out. One wrinkle worth knowing before you upgrade: Fable 5.1 isn't included in Pro plans, so Pro users pay for it from usage credits. The testing firm Artificial Analysis also says that at max effort the model burns about 1.7 times more output, which wipes out the savings.
The story: Deere introduced JD, a chatbot inside its Operations Center app that answers plain questions about a farm's own records: fuel use, planting rates, yields, machine settings, when to start harvest. It's in early access with some US customers, going wider on web, mobile, and in-cab screens later this year.
Your takeaway: It's free. Operations Center has never cost anything and JD doesn't add a fee. Deere also put out a farmer data commitment promising it won't sell farm data and that growers control who else sees it, which is a notable move from a company that spent years fighting its own customers over repair rights.
TOOLS ON OUR RADAR
✅ Vikunja Free and Open Source: Organize tasks with list, kanban, gantt, and table views, self-host for free forever, an open alternative to Todoist.
🧠 Anytype Freemium: Build a private, offline-first knowledge base with notes, databases, and tasks in one object-based workspace, your data stays local and encrypted.
🔗 IFTTT Freemium: Connect apps and smart home devices with simple if-this-then-that automations, no coding required.
📈 Plausible Paid: See who visits your website without cookies, consent banners, or complicated dashboards, lightweight and privacy-first.
TRENDING
A New Kind of AI That Thinks Without Words - Pathway's BDH-CQ has 150 million settings, a tiny fraction of what big models carry, and it works problems out internally instead of typing its reasoning. It hit 29.5% on the ARC-AGI-1 puzzle test at about seven hundredths of a cent per task, roughly 11 times cheaper than GPT-5.6 Luna, which scored 34.2%.
ChatGPT Health Connects to Epic - Doctors can now pull notes, lab results, and medication lists from Epic, which holds records for over 325 million patients, into ChatGPT for a summary. Access is read-only, so nothing gets written back into your chart. OpenAI says doctors rated 99.1% of 4,300 test answers as safe.
OpenAI Says Astra Hit Its Highest Cyber Risk Tier - Astra is the first OpenAI model rated "Critical" for cybersecurity, meaning it can find unknown flaws in well-defended systems and build working attacks without a person steering each step. It scored 100% on one exploit benchmark and turned up two brand-new flaws during testing. Advanced cyber features go to a small group of testers first.
World Labs Introduces Atlas - Fei-Fei Li's company released a model that turns a handful of photos into a 3D space you can fly a camera through, up to a full minute of video at 1440p. It rebuilds real rooms from as few as two or three pictures and can generate what a robot's camera would see walking through them. Early access only for now.
Google Pics Arrives in Workspace - Google's image tool, built on its Nano Banana model, is rolling out to AI Pro and Ultra subscribers and most Workspace business accounts. Pull one object out and change it without touching the rest, edit or translate words printed inside a picture, or edit the same image with a coworker. Live in Docs and Slides now, Drive in a few weeks.
OpenClaw 2.0 Is Here - The open-source agent project shipped version 2026.8.1 with over 16,000 code changes from 933 contributors, about half of everything it's ever merged. Setup now finds your existing ChatGPT or Claude subscription by itself, and teammates can join a running session.
TRY THIS PROMPT (copy and paste into Claude, ChatGPT, or Gemini)
🕰️ Describe what you gravitate toward. Find the real historical era that already dressed like you.
Build a single-file HTML app with vanilla HTML/CSS/JS. The Historical Style Twin — describe your personal taste, get matched to a real historical fashion era via live web search. No API key needed.
Aesthetic: dark burgundy (#1a0f12), gold (#c8a458) primary with a warm glow top-left, dusty rose (#c8888c) secondary for element accents with a glow bottom-right. Playfair Display serif for headings/era name/why-it-fits, Inter for body, JetBrains Mono for labels.
Form: single text input describing personal style/taste.
Technical: call the Anthropic Messages API WITH tools:[{type:'web_search_20250305', name:'web_search'}] (works automatically), system prompt as a fashion historian who matches the described taste to ONE real, specific historical era or named movement (not a vague century), grounded in real well-documented fashion history — real silhouettes, fabrics, colors, construction, and real associated designers/icons where relevant — never inventing details. Return raw JSON: era_name, time_period, why_it_fits (2-3 sentences), defining_elements (3-4: element category + real detail), notable_figures (optional 1-2 sentences), modern_ways_to_channel_it (3-4 practical tips).
Render: a gradient "your style twin" hero card with the era name, period, and why-it-fits paragraph in italic serif. A "what defined it" list of rose-bordered element cards. An optional notable-figures card. A gold "ways to channel it today" card with practical tips.What this does: Describe your taste, minimal and neutral, romantic and flowy, whatever comes to mind, and it searches the live web to match you to one real, specific historical fashion era or named movement, not a vague century but something with real character, 1920s Flapper, Edwardian tailoring, 1970s Bohemian.
What this looks like:
AI Insights That Turn Your Data Into Bigger Profits
Your store is full of hidden growth opportunities. StoreClaw analyzes your Shopify and Amazon data, identifies the highest-impact actions, and helps you grow revenue while improving margins. Start free today with 10,000 bonus tokens. No credit card required.
WHERE WE STAND(based on today’s news)
✅ AI Can Now: Read a liver CT scan and point out tumors a radiologist's first pass missed, cutting reading time by 27%.
❌ Still Can't: Fill in a part of a 3D scene it never saw. When Atlas doesn't have enough photos, it invents the missing area, and the invention looks just as real as the real part.
✅ AI Can Now: Solve visual reasoning puzzles without writing out a single step of its thinking, for about seven hundredths of a cent a puzzle.
❌ Still Can't: Tell a defender's security work apart from an attacker's. OpenAI says Astra's safety checks will sometimes slow or stop legitimate jobs, including defensive security work, and they haven't found a way around that yet.
FROM THE WEB
RECOMMENDED LISTENING/READING/WATCHING
The Lawnmower Man (1992) - Movie
Pierce Brosnan plays a scientist running virtual reality intelligence-enhancement experiments on a simple-minded gardener, whose mind rapidly outpaces the equipment designed to expand it. Loosely inspired by (and later legally separated from) a Stephen King short story, the film became the first major studio movie to build its whole premise around virtual reality and rapid cognitive enhancement. Dated effects, but the core idea, that boosting intelligence with technology has consequences nobody can fully predict, holds up better than the CGI.
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.
