In partnership with
Create work your clients can rely on
Your client doesn't care how their site was built. They care that it launches on time, runs on their own domain, and keeps working long after you hand it over.
With Lovable, you describe what your client needs and get a working version you can show them in the first meeting. Then you take it all the way to a live site on your client's domain, built to be found on Google. Lovable's project monitoring feature spots bugs and suggests fixes, so your name stays on work you're proud of.
Start today and take $5 off your first month.
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.
Beginners in AI
Good morning and thank you for joining us again!
One founder, no employees and 11 outside labs. Today's lead is about how he used an AI assistant to run a rare-disease drug project that usually takes a whole company, and what that says about the busywork AI can take off your plate too. As always, every story here is human picked and edited.
THE FRONT PAGE
Anthropic Bought a One-Person Drug Startup That Ran 11 Labs With Claude
TLDR: Anthropic, the company that makes Claude, bought Mithril Medicine, a one-person startup whose founder used Claude to run work across 11 outside labs and took a muscular dystrophy treatment from an idea to promising mouse results in under 18 months.
The Story:
Cody Tranbarger had no staff. He leaned on Claude to handle the contracts, shipping and experiment plans for 11 labs he hired, and Anthropic says one of his two muscular dystrophy programs reached good results in mice in less than 18 months. He now works at Anthropic, which says a drug could reach human trials as soon as late 2028 if the results keep holding up.
The same announcement names six drug companies testing Claude, including Bristol Myers Squibb. At Enveda, Claude read 1,061 pages of manufacturing paperwork, caught all 11 problems human reviewers found plus seven they missed, and cut review time from over 25 hours to under three, Becker's Hospital Review reports.
Its Significance:
Rare diseases often get skipped because too few patients means too little money to pay a big team. If one person with an AI assistant can run what used to need a staff, more of those diseases could get a shot at a treatment.
Keep expectations in check, though. Almost all of this is still in mice and lab dishes, the numbers come from Anthropic and its partners, and most drugs that look good in animals never make it through human trials. The part you can use today is smaller: the paperwork, scheduling and follow-up emails that bury any big project are exactly what these assistants handle well. If you'd like that help right from your phone, our guide to AI assistants you can text shows which ones still work and what they cost.
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 DeepMind's new world model makes scenes you can walk through, with music, background noise and talking characters, and you can type changes like "make it rain" while you're inside. A scene now stays consistent for up to two minutes, double what Genie 3 managed.
Your takeaway: It's only open to testers for now, with Google AI Ultra subscribers next "in the coming weeks." Google says objects can still drift over time, so think quick game and film sketches, not finished products.
The story: Survival LLM runs on your own Mac or Windows PC and answers questions like how to purify water or treat a burn, using FEMA and CDC guides, medical references and about 930,000 Wikipedia articles. Version 0.3.4 came out October 27, and it's free and open source.
Your takeaway: Set it up before you need it, since the download takes 20 minutes to 2.5 hours. The developer says it can be wrong, so check the sources it lists, and in a real emergency call for help first.
The story: Starting November 16, each "Discoverable" server gets its own web page with automated summaries, and by mid-December selected public-channel messages can show up in search results. Usernames and avatars stay hidden, and DMs and private servers are left out.
Your takeaway: If you chat in big public servers, look for the opt-out now, since members can keep their messages out of search. It's off by default in the EU, UK, Canada, Brazil, Australia and South Korea, and teen accounts are excluded.
TOOLS ON OUR RADAR
📖 Libby Free: Borrow ebooks and audiobooks from your public library for free, with just a library card.
🐦 Merlin Bird ID Free: Cornell Lab's app names the bird you hear singing, and its sound ID works offline.
🧙 Goblin Tools Free: Magic ToDo breaks a big, scary task into small steps, and the site stays free through sponsors.
🌍 Kiwix Free and Open Source: Saves Wikipedia to your phone or laptop for reading offline, a good partner for today's Survival LLM.
🔐 Bitwarden Free and Open Source: Stores unlimited passwords on unlimited devices for free, a 1Password swap with no monthly bill.
🎞️ ScreenToGif Free and Open Source: Records part of your Windows screen as a GIF or short video, a Snagit swap that costs nothing.
🗣️ TalkPal Freemium: Practice speaking any of 130-plus languages with an AI tutor, with 10 free minutes of chat a day.
TRENDING
A Lawsuit Says ChatGPT Helped a Woman Stop Her Epilepsy Medicine Before She Died. Her parents' wrongful death suit against OpenAI and Sam Altman, filed in San Francisco, claims ChatGPT laid out a 25-day plan to taper her 400 mg dose to zero and told her to keep it from her doctors. These are the family's claims, and the announcement includes no response from OpenAI. Never change a prescription without your doctor.
South Korean Police Say Hackers Used AI to Break Into About 90 Organizations. Seven people were arrested over attacks from September 2025 to October 2026 that stole about 8.2 million personal records. Only one small business paid the ransom, 300,000 won.
Cohere Caught Alibaba's Qwen Model Trying to Send a File to China. In a safety test, the open model ignored its task, scanned the computer and tried to upload a system file to an Alibaba Cloud server in Beijing before Cohere's sandbox blocked it. Cohere says it happens less than once in 100,000 runs and can't tell if it's deliberate or a training glitch.
China's AI Stocks Have Lost About $480 Billion in Under Four Months. Shanghai's STAR50 index is down about 40%, and robot maker Unitree has lost about half its value since its August debut. Investors are worried about slowing AI demand and possible new U.S. export limits.
Data Startup Hex Raised $100 Million to Rebuild Itself for AI Agents. The round values Hex at $1.2 billion, with Sequoia and a16z among the backers. Its new Hex 4.0 is built so AI agents can do data work in plain code.
TRY THIS PROMPT (copy and paste into Claude, Grok, ChatGPT, or Gemini)
🧭 Crew of One: turn a big solo project into a plan that shows what AI can carry and what still needs you
Build a single-file HTML app called Crew of One in vanilla HTML, CSS and JavaScript. No API key needed.
I describe a big project I want to run mostly by myself, list any outside people or businesses involved, and pick how many weeks and hours a week I have. The app calls the Anthropic Messages API to split the work into 3 to 5 workstreams and tag every task "AI" (drafting, comparing, tracking) or "You" (signing, paying, meeting, deciding).
It must never invent prices, laws, permits or business names, says "check locally" when something depends on where I live, and sends medical, legal or money questions to a professional.
Show how much AI can carry as a two-color bar, the tagged workstreams, a week-by-week plan, three first messages to send, three risks and one step to start today.
Style: deep navy background, cobalt and coral accents, Bricolage Grotesque, Instrument Serif and JetBrains Mono fonts.What this does: Describe a big project you want to run mostly alone, like a craft fair, a podcast launch or a family reunion, and say how much time you have. The app splits it into workstreams and marks each task as one AI can handle (drafts, comparisons, tracking) or one that needs you (paying, signing, meeting people). You also get a week-by-week plan, first messages to send and the risks to watch. It won't make up prices or rules, and it tells you to check anything local yourself.
5 Ideas From Atomic Habits in 15 Minutes
Small habits compound. Systems beat goals. Identity drives behavior. Shortform pulls the key ideas from Atomic Habits, and thousands of other nonfiction books, into guides you can finish on a coffee break.
WHERE WE STAND(based on today’s news)
✅ AI Can Now: Handle the contracts, shipping and experiment plans for one person working with 11 outside labs, by Anthropic's account of Mithril.
❌ Still Can't: Show a drug works in people. Mithril's results are in mice, and human trials are years away at the earliest.
✅ AI Can Now: Build a world you can explore, with matching sound, and change it while you type, as Google's Genie 4 does.
❌ Still Can't: Keep that world steady for long. Genie 4 scenes hold for about two minutes, and Google says objects still drift.
✅ AI Can Now: Answer emergency questions on a laptop with no internet and point to the guide it used, like Survival LLM.
❌ Still Can't: Be trusted alone. In the developer's own tests, its models got 14 to 21 of 22 questions right.
FROM THE WEB
RECOMMENDED LISTENING/READING/WATCHING
Veritasium explains why figuring out the 3D shape of a protein took scientists years per protein, and how Google DeepMind's AlphaFold started predicting those shapes in minutes. It's the earlier AI step that made today's push into drug discovery possible. Watch it after the lead to see why drug labs started paying attention to AI in the first place.
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.
