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
Google's New AI World Model Can Now Build Full Videos From a Single Photo in Seconds
TLDR: Google launched Gemini Omni, a new AI model that turns photos, audio, or plain words into video and tries to copy how the real world actually works.
The Story:
Google introduced Gemini Omni at its I/O 2026 event this week. It's a new AI model that takes almost any input, a photo, a few words, an audio clip, and turns it into video. The model mixes Google's Gemini brain with its older tools like Veo for video, Nano Banana for image edits, and Project Genie for world simulation. Google DeepMind CEO Demis Hassabis called it a "step toward artificial general intelligence" and said the goal is for AI to understand how things in the real world actually behave, like gravity, motion, and physical objects bumping into each other. The first version, Gemini Omni Flash, is rolling out now inside the Gemini app, Google Flow, and free inside YouTube Shorts. Users can also make videos of their own digital avatar, but they have to record themselves reading out numbers first so the system blocks deepfakes.
Its Significance:
This isn't just another video generator. Most tools like Sora or Runway only know how to make pretty clips. Gemini Omni is trying to learn how the actual world works, so a marble rolling down a track stays consistent with how a real marble would move. For regular people, that means making short videos for social media, school projects, or family memories gets a lot faster and a lot cheaper. For pro creators, it raises the bar on what's possible without a film crew.
QUICK TAKES
The story: James Daunt, CEO of Barnes & Noble, told the Today show his stores will carry AI-written books as long as they're clearly labeled and don't pretend to be human-made. He added that of the 300,000 titles already on his shelves, some are probably AI-written without anyone knowing.
Your takeaway: Book lovers online are not happy. Several said they'd stop shopping there. But Daunt's point is sharp: AI books are already in stores, the question is just whether to label them. If you write or buy books, expect "Written by AI" stickers to become as normal as "Made in China" tags on toys.
The story: Google announced Gemini Spark at I/O 2026, an AI agent that handles tasks across your apps even when your phone is closed. It can scan your credit card statements for hidden subscriptions, pull meeting notes from Gmail and Docs into clean write-ups, and watch your kid's school emails for important dates.
Your takeaway: This is Google's answer to Anthropic's Claude Cowork and OpenAI's ChatGPT Agent. The advantage Google's got is 900 million people already using Gemini. No new app to learn. Spark goes live for U.S. Google AI Ultra users next week, and Google just cut that plan from $250 to $200 a month.
The story: MIT associate professor Connor Coley is teaching AI models to follow the same rules a real chemist uses, like conservation of mass and reaction mechanisms. Most AI tools just guess what molecule comes next. Coley's models try to reason through the chemistry behind it.
Your takeaway: This matters for drug discovery. Scientists estimate there are between 10²⁰ and 10⁶⁰ possible molecules that could be useful as medicine. AI that gets the chemistry right could cut years off finding the next cancer drug or antibiotic. The early version is free on GitHub.
TOOLS ON OUR RADAR
📝 Standard Notes Free and Open Source: A highly secure digital notepad that fully encrypts your personal journals and daily thoughts so absolutely nobody else can read them.
🌍 QGIS Free and Open Source: A tremendously powerful mapping tool used by professionals and hobbyists to create beautiful detailed geographical maps and visualize spatial data.
🦊 Waterfox Free and Open Source: A privacy focused web browser that removes tracking telemetry and gives you complete control over your internet surfing experience.
🗜️ Keka Free and Open Source: A brilliantly simple file archiver for Mac computers that easily compresses and extracts large folders without any complicated menus.
TRENDING
People lie to AI more than they lie to humans, new study finds - Researchers at Sun Yat-sen University found people are more likely to exaggerate results, claim fake discounts, or exploit price errors when dealing with a chatbot instead of a person. The reason: no fear of being judged.
OpenAI joins C2PA and adds Google's SynthID watermark to its AI images - OpenAI is now embedding invisible watermarks in every image made by ChatGPT, Codex, and its API, plus signing them with C2PA metadata. A public tool at openai.com/verify lets anyone check if an image came from OpenAI.
Google launches Gemini 3.5 Flash, says it beats Gemini 3.1 Pro on coding - Google's new default model runs four times faster than rival frontier models at the same tier and is now powering Search's AI Mode, the Gemini app, and Gemini Spark. Gemini 3.5 Pro is coming next month.
Google rebuilt the search box for the first time in 25 years - The new "intelligent search box" expands as you type, accepts text, images, video, files, and open Chrome tabs, and runs on Gemini 3.5 Flash. AI Mode now has more than 1 billion monthly users.
How Indianapolis newsrooms are using AI right now - Most local newsrooms use AI for transcription only, with reporters double-checking quotes by hand. Axios Indianapolis has a formal ChatGPT partnership for data analysis. Indiana Capital Chronicle, Mirror Indy, Chalkbeat, and WFYI all keep humans in the loop.
Study finds AI use rises as people stop noticing it - PYMNTS Intelligence found that 56% of U.S. adults, about 146 million people, now use AI for at least one personal task. The pattern looks like mobile banking adoption: not a hype cycle, just a habit forming.
Shield AI's Hivemind picked to power Pentagon's low-cost attack drones - The Office of the Under Secretary of War selected Shield AI to put its Hivemind autonomy software on LUCAS, a new fleet of cheap one-way attack drones built to fly in coordinated groups. Humans still make strike decisions; the AI handles navigation and timing.
TRY THIS PROMPT (copy and paste into Claude, ChatGPT, or Gemini)
🌳 Name something you want to learn. Get the dependency tree of exactly what to learn in what order.
Build a single-file HTML app with vanilla HTML/CSS/JS. The Skill Tree — dependency map for learning anything. Persist to localStorage key 'skill_tree_v1'.
Aesthetic: deep forest (#0d1410), faint hex pattern, green/gold mist gradients. Cinzel for headings (all-caps, letter-spaced), Lora serif for body, JetBrains Mono for labels. Tier accent colors progress from earth brown to green to teal to amber to deep amber. Tree-themed emoji icons per tier (🌰🌱🌿🌳🌲).
Form: learning goal, current level dropdown, time-per-week dropdown, motivation input.
System instructions to the model: design dependency trees where each tier depends on the previous one. Skills must be specific and named, not generic ("Read 5 Chekhov stories" not "read good fiction"). Resources must be specific too. Return raw JSON: goal_clean, total_time_estimate, tiers array (4-5 items: name, time_estimate, purpose, skills array with name/why/resource, milestone test), first_step (one specific action to take in 30 minutes).
Render: a centered goal header bar, tier cards stacked with colored left-border stripes and emoji icons, each tier showing purpose, numbered skill items (with name/why italic/resource line), and a green-tinted milestone test box at the bottom. A 3-stat journey card (tiers/skills/total time). An amber-bordered first-step card at the bottom. Archive in localStorage.What this does: Type your goal, your level, and how much time per week. Get a 4–5 tier dependency tree showing the foundational skills underneath the goal, in order. Each tier has 3–4 named skills (with a one-line why and a specific resource), a milestone test to know when you've cleared it, and a time estimate. Plus one concrete action to take in the next 30 minutes. Saves to localStorage.
What this looks like:
WHERE WE STAND(based on today’s news)
✅ AI Can Now: Turn a single photo, voice clip, or sentence into a video that mostly follows how real-world physics work (Google Gemini Omni)
❌ Still Can't: Reliably tell the difference between an AI-made image and a real one across every platform, since watermarks like SynthID only cover images from companies that opt in
✅ AI Can Now: Run quietly in the cloud for hours, sorting your email, summarizing meetings, and tracking your kid's school deadlines without you doing anything (Google Gemini Spark)
❌ Still Can't: Be trusted with full autonomy on important decisions. People lie to AI more than to humans, and AI still gets chemistry, math, and judgment calls wrong often enough to need a human checking the work
FROM THE WEB
RECOMMENDED LISTENING/READING/WATCHING
A detective in a near-future Britain where every citizen is constantly surveilled by a benevolent AI called the Witness investigates the death of a woman during a routine "mind audit," and what she finds inside the dead woman's memories collapses into four nested lives across different eras. Harkaway built this as a Borgesian puzzle box about algorithmic governance, and it predicts the texture of an AI-monitored society.
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.
$992 Billion in Art Could Change Hands. Why Are These 71,105 Investors Paying Close Attention?
Deloitte ran the numbers. They project UHNW art and collectibles wealth -- already at $2.5 trillion -- to hit $3.47 trillion by 2030.
The institutional world has been quietly preparing for this. Back in 2011, 25% of wealth managers surveyed offered art-related services. In 2024, 51%. Family offices now average a 13.4% allocation to art and collectibles. And it’s not just because they love art. It’s because they like the math.
These positions were built over decades through private dealer relationships most investors never had. The access just wasn't there.
Masterworks is changing that:
71,000+ investors
$1.3B deployed across 525+ artworks
29 closed sales
Net annualized returns like 16.5%, 17.6%, and 17.8%, not including those unsold.
Investing involves risk. Past performance is not indicative of future returns. See important disclosures at masterworks.com/cd.
