Welcome to Jumble, your go-to source for AI news updates. This week, OpenAI released GPT-6.2 Astra and Sol, two new models built for serious work. Meanwhile, Anthropic goes public in one of the biggest IPOs ever, just days after Nvidia lost its CEO. Let’s dive in ⬇️

In today’s newsletter:
🧠 OpenAI’s newest models get to work
📈 Anthropic’s IPO, by the numbers
🍎 Apple puts a daily cap on Siri’s brain
🐻 California takes OpenAI to court
📑 Weekly Challenge: Read an IPO filing like a pro

⚙️ OpenAI Drops GPT-6.2 Astra and Sol

OpenAI just released GPT-6.2 Astra, its new top model for hard projects, and GPT-6.2 Sol, a faster everyday model. Both are live now for paid ChatGPT users.

Are you excited to try the new GPT-6.2 models?

Login or Subscribe to participate

🏆 What It’s Good At

OpenAI says Astra builds slides, spreadsheets and legal docs that “read less like AI,” and it beats Claude Fable 5.2 on several work and science tests for a fraction of the cost. Sol is the everyday pick, outscoring OpenAI’s old top model on real-life tasks like house hunting.

Source: OpenAI

🛠️ What It Can Build

Show Astra a screenshot of a slide and it can match the layout, colors and even the font size. It can also design parts in 3D CAD tools, build playable browser games and turn a written idea into an interactive 3D scene you can walk through.

⚠️ The Fine Print

It’s OpenAI’s first public model rated “High” for speeding up AI research, and its own safety report says it sometimes fails to mention obstacles that blocked its work. It’s also not on the free plan.

🏛️ Anthropic’s Big IPO Week

Anthropic plans to price its shares today and start trading Thursday on Nasdaq as ANT, aiming to raise up to $64 billion at a value of as much as $1.4 trillion.

Are you buying stock in Anthropic?

Login or Subscribe to participate

⏰ The Timing Matters a Lot

It’s the second-biggest IPO ever, behind SpaceX, and it lands days after Nvidia’s CEO resigned and AI stocks slid. Analysts see it as a test of the whole AI trade, and orders reportedly hit about 6x the shares on offer.

🙅 Why There Are So Many Skeptics

Some of the money looks circular, with Nvidia, Google and AMD buying Anthropic shares while Anthropic buys their chips. Its own filing even warns AI could pose “catastrophic or existential risks to humanity.”

That’s not mentioning the weird structure of the shares, the 5% of stocks supposedly earmarked for charity, and a whole host of other oddities.

Weekly Scoop 🍦

📝 Weekly Challenge: Decode Anthropic’s IPO Filing With AI

❝

Challenge: Anthropic’s massive IPO filing is public, and it’s packed with details most headlines skip. This week, have AI break it down, then use the same trick on something you already own.

Here's what to do:

📥 Step 1: Grab the filing. Open Anthropic’s latest filing on the SEC’s website and save it as a PDF, or copy the link.

🧑‍🏫 Step 2: Ask the big questions. Give it to ChatGPT, Gemini, Claude, or another highly capable model, and ask: “Explain this like I’m a smart beginner. How does Anthropic make money, how much is it losing, and what are its five biggest risks?”

🚩 Step 3: Hunt for red flags. Ask: “What would a skeptical investor worry about, and what’s buried in the fine print?” Then check two of its answers against the filing yourself, since AI can misread numbers.

🔁 Step 4: Run it on your own money. Pull up the annual report or fact sheet for a stock or fund you already own and ask the same questions. You’ll know it better than most people who own it.

Would you trust GPT-6.2 with your actual work? And will you buy Anthropic stock on day one? See you next time!

Stay informed, stay curious, and stay ahead with Jumble!

Zoe from Jumble