Sponsored by
Trading options is complicated. Liquid simplifies your trading strategies.
Trading options can be highly profitable; it can also be highly unprofitable. The problem is that trading options is tough:
Option decay is real
Picking expiration dates is crucial
Market makers scalp retail traders
Liquid allows anyone to trade perpetual futures, or “perps”. Perps are options without an expiration date.
This means you don’t need to worry about Greeks, simply
Pick an asset
Pick long or short
Determine your collateral
Determine your leverage
Place your trade
Getting started is simple too; start to finish in less than 10 minutes:
Login with google
No-KYC
Deposit with Apple Pay or your bank account
Trade
Trading volatility doesn’t need to be harder than absolutely necessary.
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
Midjourney AI Image Company Just Built a Body Scanner. Yes, Really
TLDR: Midjourney, famous for AI images, just revealed its first piece of hardware, a 60-second full-body scanner it wants to put in spas and scale to 50,000 machines by 2031.
The Story:
Midjourney built its name turning text into pictures, and it's still one of the best image tools around. On June 17, founder David Holz stood on a stage in San Francisco and showed off something nobody saw coming: a full-body medical scanner. The machine, called the Midjourney Scanner, lowers you into a pool of water where a ring of tiny sensors sends sound waves through your body. Holz says one scan takes about 60 seconds, while a full-body MRI can take an hour or more, and he claims the results beat an MRI in some ways, with no radiation and no giant magnets. The plan is to put these scanners inside "Midjourney Spas" with hot tubs and saunas, the first opening in San Francisco around the end of 2027. The company says it wants 50,000 scanners running worldwide by 2031.
Its Significance:
What makes this interesting: Midjourney makes pictures, and now it wants to make medical hardware, which is a much harder job with far higher stakes. The move makes sense when you look at its main business. AI image tools got crowded fast, with strong options from Google, OpenAI, Grok, and others chipping away at the lead Midjourney once held, so trying something brand new is one way to stand out. Holz admits the scanner barely uses AI yet, it has no FDA approval yet, but no outside investors are paying for the hardware either.
QUICK TAKES
The story: Snap opened preorders for Specs, standalone augmented reality glasses that drop digital images into the real world, shipping this fall for $2,195. People online roasted the thick, goggle-like frames over their size, the price drew plenty of eye-rolls, and Snap's stock fell more than 4% the day of the reveal.
Your takeaway: Snap beat Meta, Apple, and Google to a real consumer AR product, but "first" doesn't mean "ready." At more than 15 times the cost of its old camera glasses, and with a chunky look, these feel built for developers and die-hards, not your average shopper.
The story: Researchers at Boston Children's Hospital used OpenAI's o3 model to comb through the genomes of children whose rare illnesses had gone undiagnosed for years, and it pointed to the likely cause in 18 cases. The work, published in NEJM AI, found new answers in about 5% of genomes that human experts had already studied many times.
Your takeaway: 5% sounds small until you remember each one is a family that finally got an answer. This isn't AI replacing geneticists. It's AI catching patterns in messy data that tired human eyes can miss after the tenth read.
The story: Police and shelters are warning about a scam where crooks grab photos from "lost pet" posts, use AI to make fake images of the animal hurt or on an operating table, then demand cash for "emergency surgery." Victims have been pushed to send hundreds or thousands of dollars through hard-to-trace apps before they realize the pet was never found.
Your takeaway: The fix is low-tech. Ask anyone who claims to have your pet to send a live video, or to hold up a paper with today's date and a random word on it. A fake image can't do that on the spot.
TOOLS ON OUR RADAR
🎙️ Mumble Free and Open Source: A private and low latency voice chat application designed for gamers and large groups of friends to talk to each other clearly over the internet.
🧸 GCompris Free and Open Source: A wonderful educational software suite packed with hundreds of fun activities and mini games designed specifically to help young children learn math geography and computer skills.
💡 LightBulb Free and Open Source: A clever application that protects your eyes by automatically adjusting the color temperature of your computer screen as the sun sets in your local time zone.
🖨️ Cura Free and Open Source: A remarkably powerful slicing software used by hobbyists around the world to prepare their three dimensional digital models for actual physical production on a desktop 3D printer.
TRENDING
Microsoft Patched a Copilot Flaw That Leaked 2FA Codes With One Click - Researchers at Varonis found a bug in Microsoft 365 Copilot Enterprise that let an attacker steal 2FA codes, emails, and files when a victim clicked one link that looked like a normal Microsoft web address. Microsoft rated it max severity and has since fixed it on its end, so users don't need to do anything.
Most Psychologists Say Their Patients Are Bringing AI Into Therapy - A new survey of more than 1,200 U.S. psychologists found 77% have patients using AI for emotional support, self-diagnosis, or companionship, and 35% said patients treat a chatbot like a second therapist. More than a third noticed patients leaning on the bots a bit too much, and 15% saw distorted thinking tied to the chats.
OpenAI's New Science Test Stumps Even the Best AI Models - OpenAI released LifeSciBench, 750 tasks written by 173 working scientists to check how well AI handles real drug-discovery and biology research. The strongest model passed only about 36%, a reminder that today's AI is still far from doing a scientist's full job.
Character.AI Rolls Out New Tools to Help Its Creators Grow - The chatbot platform added a Creator Guide, performance insights, a "following" tab, and leaderboards so the people who build its characters can see what's working and find an audience. It's a push to reward the creators behind millions of characters, though there's still no direct way to cash out.
Anthropic and Google DeepMind Push for a US-Led AI Coalition at the G7 - Anthropic's Dario Amodei and Google DeepMind's Demis Hassabis used a closed-door G7 lunch to ask world leaders, including President Trump, to back the U.S. leading global AI rules. The pitch covered shared access to top models and chip trade that leaves out China, and Canada's prime minister reportedly agreed the U.S. could lead.
The Hidden Risk of "Orphaned" AI Agents Inside Companies - When the person who built an internal AI tool leaves a company, the tool often keeps running with full access to sensitive databases and code long after their login is shut off. Security pros say these forgotten agents are a quiet back door for attackers, and most companies can't even name who owns them.
Ukrainian AI Drones Reportedly Picked Their Own Targets and Killed - A Ukrainian drone maker says 10 fully autonomous "Terminator" drones flew to the front line about two years ago, searched for targets on their own with no human in the loop, and killed Russian soldiers. If confirmed, it may be the first time AI alone chose to take human lives, and Ukraine's own rules still require a person to approve the final strike.
TRY THIS PROMPT (copy and paste into Claude, ChatGPT, or Gemini)
📐 Name a growth stock you're eyeing and your levels. Get a structured swing-trade plan with the position-sizing math done right and clear if-then rules.
Build a single-file HTML app with vanilla HTML/CSS/JS. The Trade Plan Forge — a swing-trade planner for growth stocks that computes position sizing locally and uses one AI call for the rules layer. Persist to localStorage key 'trade_plan_forge_v1'.
Important framing: this is educational, not financial advice, and has NO live market data. Show a prominent amber disclaimer. The user supplies all prices; the app does not pretend to know current prices.
Aesthetic: trading-terminal dark (#0a0e13), top green radial glow. Inter sans 800 headings, JetBrains Mono for all numbers/prices, Newsreader serif italic for notes. Confident green (#2ecc8f) primary, red (#e5484d/#f0686c) for stop/risk, blue (#6aa8f0) for entry, amber (#e5a048) for cautions. Price-ladder visual (target/entry/stop stacked). Dollar-prefix inputs.
Form sections: The stock (ticker, current price, optional thesis textarea), The levels (entry/stop/target price inputs), Your account & risk (account size, risk-per-trade dropdown 0.5-3%, timeframe dropdown).
Compute ALL math client-side from inputs (validate: stop<entry, target>entry, account>0): risk per share, reward per share, R-multiple, dollar risk (account×risk%), shares (floor of dollarRisk/riskPerShare), position value, actual dollar risk, % of account, % to stop, % to target, potential profit. Display a price ladder and a position-sizing stat grid.
ONE AI call adds the rules layer (pass it the computed math, tell it not to recompute, and that it has no live data so it should comment on PLAN structure not stock quality). System instructions: disciplined swing coach, rules-based, honest, flag weak R:R (<2:1) or oversized positions, never guarantee profit. Return raw JSON: thesis_framed (2-3 sentences), entry_rules (3 if-then rules referencing their entry), stop_rules (3), exit_rules (3 referencing target), checklist (4 pre-trade items), invalidators (3 conditions that kill the thesis), honest_read (2-3 sentences on structural quality only).
Render: ticker header with color-coded R:R badge (green ≥2, amber ≥1.5, red below). Price-ladder card (target ▲ / entry / stop ▼ with %s and per-share amounts). Position-sizing math grid (shares, position size, $ at risk in red, profit at target in green, R:R). Then AI sections: framed thesis, blue entry rules, red stop discipline, green exit plan, green-bordered checklist (✓), amber-bordered invalidators (⚠), honest-read card. Copy plan button appends "Educational only, not financial advice." Archive shows R:R and risk per saved plan.What this does: Enter the ticker, your thesis, your entry/stop/target prices, account size, and risk-per-trade. The app computes the position-sizing math locally so it's exact (shares to buy, dollar position, actual dollars at risk, reward-to-risk ratio, percent moves to stop and target, potential profit), then displays a clean price ladder. The AI layer adds the structure: the setup framed, if-then entry rules, stop discipline, an exit and scaling plan, a pre-trade checklist, what would invalidate the trade, and an honest read on the plan's structural quality (it flags weak R:R or oversized positions). Saves plans to localStorage. It's an educational discipline tool, with prominent reminders that it has no live data and isn't financial advice.
What this looks like:
WHERE WE STAND(based on today’s news)
✅ AI Can Now: Read a child's full genome and point doctors to the likely cause of a rare disease (Boston Children's found 18 new answers).
❌ Still Can't: Pass most real-world life science research tasks (the best model scored about 36% on OpenAI's new test).
✅ AI Can Now: Search for and strike a target on a battlefield with no human giving the order (Ukraine's Terminator drones).
❌ Still Can't: Stop itself from being tricked by one crafted link into leaking your 2FA codes (Microsoft Copilot's SearchLeak bug).
RECOMMENDED LISTENING/READING/WATCHING
The Thirteenth Floor (1999) - Movie
A 1999 sci-fi noir about a tech company whose engineers have built a fully populated simulation of 1937 Los Angeles, where each AI inhabitant believes they are real, until one of the company's founders is murdered and his colleague starts investigating clues left inside the simulation. Came out the same month as The Matrix, got completely buried, and is in some ways the headier of the two films. One of the best simulation-hypothesis movies ever made and almost nobody has seen 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.
Moda is the AI design agent with taste.
Your designer is busy. Agencies are expensive. And the deck needs to go out today.
Moda is your AI design partner for the moments when you need great work fast.
Drop in your brand, tell Moda what you need, and get polished, on-brand slides, docs, or ad creative in minutes.