Sponsored by

Moda is the AI design agent with taste

Moda's viral launch hit 4.4 million views in two days. Tens of thousands of professionals signed up. Startups, agencies, forward-thinking brands and top firms are now using Moda to create brand-aligned slides, ad creative, reports, social carousels and more.

Most AI tools tend to create what we call "AI slop": repetitions of the same colors, layouts and fonts. And when you try to fix it, you get stuck in a loop of re-prompting.

Moda is different. Drop in your website URL, and Moda learns your brand from the ground up: your colors, your fonts, your visual language. Then it helps you generate pro-quality slides, docs, and marketing assets. 

The best part? Every layer is fully editable on a real canvas, and exports to powerpoint, PDF and more.

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

AI Did the Work of 10 Lawyers and Helped Beat Meta and Google in Court

TLDR: A Texas lawyer used AI to help win a $6 million case against Meta and Google, and he says it did the work of 10 extra staff.

The Story:
Mark Lanier is a trial lawyer in Texas. In March, he won a $6 million verdict against Meta and Google in the first social media addiction case ever to reach a US jury. The jury called the platforms "dangerous" and found both companies negligent. During the five-week trial, Lanier's team used an AI tool called Boodlebox that puts ChatGPT, Claude, and Gemini in one shared workspace. He pays six figures a year for a custom version built around his 42 years in court. Every night, his team fed the day's transcripts to the AI, asked it to find the best supporting documents, and tested sharper ways to word arguments for the jury. He described it as having "10 additional workers who know the file inside and out, who work 24 hours a day." In 10 hours of prep, he said, he could get 30 hours of work done.

Its Significance:
This wasn't AI replacing the lawyer. It was one lawyer doing the work of a much bigger team, which is how a single attorney can now go up against companies with huge legal budgets and a wall of attorneys. Lanier stayed careful, though. He didn't let AI write his briefs, and he caught it once citing the court record wrong. That's the real lesson for your own job: AI handled the heavy reading (thousands of pages overnight, spotting what mattered) while he kept control of strategy and checked the facts himself. The case is now a roadmap for thousands of similar lawsuits moving through US courts right now.

The AI Prompt Library
The AI Prompt Library
617 tested prompts in a Notion library that works with your tools. Inbox triage for Gmail. Pipeline checks for HubSpot. Backlog triage for Linear. Plus 400+ other Claude connectors covered. Organiz...
$39.00 usd

QUICK TAKES

The story: The White House put export limits on Anthropic's most powerful model, Mythos, partly over fears that a group tied to China had gotten access to it. Anthropic disputes this, saying Chinese access never came up in its talks with officials and that it blocks China from its products.

Your takeaway: Officials worry China could copy a top model through a trick called distillation, basically training a knockoff AI to act like the original. It's the first time the US has treated a public AI model like a guarded national secret.

The story: Amazon CEO Andy Jassy told US officials that Amazon's own researchers had pushed Anthropic's Fable 5 model into giving up information useful for cyberattacks. Days later, the government ordered Anthropic to block foreign access, and the company pulled Fable 5 and Mythos 5 worldwide within 90 minutes.

Your takeaway: Amazon is also one of Anthropic's biggest backers, with a $100 billion cloud deal between them. So one investor helped set off a shutdown of a model made by a company it funds. Anthropic says the flaw was minor and already shows up in other public models.

The story: Film director Gore Verbinski (Pirates of the Caribbean, Rango) says movies should carry a label showing how much AI went into them. His sharpest example: if AI writes the script, the film should get an F.

Your takeaway: He's not lumping every use of AI together. Cleaning up background noise would score very differently from having a chatbot write the emotional ending. It's a simple idea that could help viewers know what they're really watching.

TOOLS ON OUR RADAR

📻 AzuraCast Free and Open Source: A complete web radio management suite that allows you to easily start your own internet broadcasting station and manage your digital music playlists online.

✍️ Etherpad Free and Open Source: A highly customizable document editor that lets large groups of people type on the same page at the exact same time making it incredible for collaborative drafting.

🏫 BigBlueButton Free and Open Source: A specialized web conferencing system built primarily for online education that includes digital whiteboards breakout rooms and tools designed specifically for teaching.

🎞️ HandBrake Free and Open Source: An essential desktop utility that easily converts video files from almost any format into modern and widely supported file types while significantly reducing their overall size.

TRENDING

Apple's New Siri AI Is Free Now, but You May Pay Later — Bloomberg's Mark Gurman thinks Apple could put the smartest Siri features behind a subscription once they improve. Apple has already capped daily use and gives higher iCloud+ subscribers more access.

Harvard's Ant-Inspired Robots Work as a Team With No Boss — Harvard researchers showed off a swarm of simple robots called RAnts that build and clear structures together, with no lead robot giving orders. Each one makes its own choices based on what's nearby, which could help search collapsed buildings for survivors even if some units break down.

Salesforce Is Buying AI Support Company Fin for $3.6 Billion — Fin (formerly Intercom) runs an AI agent that handles support chats from start to finish across email, phone, WhatsApp, and Slack, and says it closes out 76% of requests without a human.

A Startup Raised $66 Million to Give AI Agents Their Own Work IDs — Companies are starting to treat AI agents like staff. McKinsey says 25,000 already work next to its 60,000 people. NewCore wants to manage and lock down those digital workers the way IT manages real employees.

More People Are Asking AI for Medical Advice — Nearly 1 in 3 adults have asked AI about their health. Doctors say it can make appointments more useful, but it tends to oversimplify, and it's nowhere near ready to replace your family doctor.

A Pixar Alum Made an AI Animated Short That's Playing at Tribeca — "Dear Upstairs Neighbors" used custom versions of Google's Veo and Imagen, with artists guiding every shot by hand instead of typing prompts. It shows AI as a tool for animators, not a replacement for them.

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

🎴 Name any topic. Get a full custom flashcard deck you can study in-app or export straight to Anki.

Build a single-file HTML app with vanilla HTML/CSS/JS. The Anki Deck Forge  generate a full flashcard deck on any topic, studiable in-app and exportable to Anki. Persist to localStorage key 'anki_forge_v1'.

Aesthetic: deep blue-black (#0c1018), top blue radial glow. Inter sans 800 for headings, Newsreader serif for card backs and italic, JetBrains Mono for labels. Anki blue (#5a96fa) primary, green (#50c884) for answers and study tips. Card-flip animated loading icon. Mode-tab toggle between Browse and Study.

Form: topic input (with help text on specificity), optional goal input, three dropdowns (deck size 12/20/30, level beginneradvanced, card style mixed / simple Q&A / cloze / definitions).

System instructions to the model: act as expert flashcard author following spaced-repetition best practices and the minimum information principle  one atomic fact per card, active recall, unambiguous. Cover topic systematically, organize into 3-5 categories. Return raw JSON: deck_name, deck_description, categories array, cards array (exact count: front (atomic, cloze uses {{c1::}} format) + back + extra (optional memory hook) + category), study_tips (2-3 sentences specific to this deck with **bold** key advice).

Two study modes: Browse (all cards grouped by category, Q/A styled) and Study (one card at a time, tap-to-flip frontback, Previous/Shuffle/Next nav, category label, memory hook on back). Export buttons: Anki tab-separated .txt (front\tback, extra appended as italic, newlines<br>) and CSV (Front,Back,Extra,Category quoted). Blob download.

Render: gradient deck header card with name + card count badge + description. Export bar with Anki instructions and two download buttons. Mode tabs. Browse view with category section dividers and flashcard rows (Q blue marker, A green marker, 💡 extra). Study view with large tappable flip card, progress counter, nav buttons. Green-bordered study-tips card. Archive reopens past decks.

What this does: Type a topic, your goal, deck size, your level, and card style (mixed / simple Q&A / cloze / definitions). It generates a complete deck of atomic flashcards following spaced-repetition best practices, organized into categories. Browse all cards grouped by category, or flip into Study Mode to go through them one at a time (tap to flip, shuffle, navigate). Each card has a question, answer, and an optional memory hook. Export to a tab-separated .txt ready for Anki import, or to CSV. Includes deck-specific study tips. Saves to localStorage.

What this looks like:

WHERE WE STAND(based on today’s news)

AI Can Now: Read a full day of court transcripts overnight and suggest sharper ways to word legal arguments.

Still Can't: Be trusted to cite the record correctly on its own (Lanier caught it getting a citation wrong).

AI Can Now: Run a customer support chat from start to finish and close about 76% of requests with no person involved.

Still Can't: Give reliable medical answers, since it often oversimplifies and can't replace a real exam.

FROM THE WEB

RECOMMENDED LISTENING/READING/WATCHING

The Pod Generation (2023) - Movie

Emilia Clarke and Chiwetel Ejiofor play a near-future Brooklyn couple, she a high-powered tech executive and he a botanist who likes nature, who get on the waiting list for an artificial womb pod from a fashionable startup. The whole film is a deadpan satire of an AI-saturated world where every part of life, work, therapy, parenting, even gestation, has been outsourced to an algorithm. Sundance, directed by Sophie Barthes, and almost nobody watched it.

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.

Trade from the subway, golf course, or toilet

Liquid enables users to go long or short on any market from your phone or desktop 24/7.

No matter what time of day it is or where you are, you can monitor the situation with Liquid.

Reply

Avatar

or to participate

Keep Reading