Welcome, humans.

California is trying again with SB 947, a bill that would ban employers from letting AI fire or discipline workers without a human signing off. Newsom vetoed an earlier version last October, but lawmakers reintroduced it as the "No Robo Bosses Act of 2026."

The stakes are already real. At Andon Market in San Francisco, an AI agent named Luna runs hiring, scheduling, and pricing. She recently recommended firing a worker who missed 17 of 23 shifts. Humans carried it out. Only problem? Luna had written the store's attendance policy herself, then forgot it existed until someone told her to go check her memory.

The AI boss is coming. It still needs a better filing system.

Here’s what happened in AI today:

• 😸 Anthropic's IPO could raise more than $100 billion, potentially the biggest stock market debut ever.

• 📰 Nvidia is spending $7 billion on AI startup Poolside to build a US rival to Chinese open-source models.

• 📰 Alibaba and Tencent spent $18 billion on AI infrastructure last quarter alone.

• 📰 Blackstone and Anthropic are embedding 160 AI engineers directly inside portfolio companies to build new products.

😺 Claude's Maker Could Pull Off the Biggest IPO in History

You've probably used Claude to draft an email, debug a script, or automate half your job this month. Turns out enough other people have too that Anthropic might be about to pull off the biggest stock market debut ever.

Here's the deal: Anthropic's bankers have told investors the company could raise "more than $100 billion" when it goes public, according to the New York Times. That would put its value near $2 trillion, more than double the $965 billion it was worth after its last private funding round in June.

Here's what happened:

• Anthropic's IPO could "match or beat" SpaceX's June debut, which raised $85.7 billion at a $1.77 trillion valuation (the largest IPO ever)

• Morgan Stanley, Goldman Sachs, and JPMorgan are reportedly leading the offering, with a listing possible as soon as October

• The company could reveal its public filing paperwork in the coming weeks, with shares possibly listing this autumn

• Only Apple, Microsoft, and Nvidia have ever crossed the $2 trillion valuation mark

• Anthropic would beat OpenAI to the public markets; OpenAI isn't planning to list until 2027

Why this matters: Claude Code, Anthropic's coding assistant, has become one of its most popular products and helped push the company's projected annual revenue to $47 billion. That's a five-year-old startup founded by siblings Dario and Daniela Amodei, both former OpenAI executives, betting its safety-focused approach to AI can out-earn (and now out-IPO) the company they left.

An IPO this size would also settle an argument that's been simmering all year: whether the "safety-first" lab can actually out-compete the "move fast" one. So far, the money says yes.

Our take: The hype has a couple of asterisks. Anthropic has struggled to get enough chips and servers to keep up with demand, and in March, the Trump administration cut off its Defense Department contracts entirely after Anthropic refused to give the military unrestricted access to its models (Anthropic called the move "unconstitutional retaliation"). A record-breaking IPO and a fight with your own government's biggest customer don't usually show up in the same earnings call.

Anthropic's prospectus should tell us which story is closer to true.

FROM OUR PARTNERS

The Neuron Exclusive: Invest in High-Potential AI Startups Like These

The Neuron and Alumni Ventures are giving readers early access to high-growth startup opportunities, including some of today’s most exciting AI, Deep Tech, Quantum Computing, and Cybersecurity companies co-invested alongside top VC firms like Andreessen Horowitz (a16z), Bessemer, & Y Combinator.

You get:

• Curated deal flow of high-potential AI First startups

• AV is already investing alongside elite lead venture firms in these deals

• No cost to see deals

• No obligation to invest

Don’t miss your chance before access closes.

→ See Current AI Deals

🎓 AI Skill of the Day: Make AI prove its homework

So, apparently State Farm’s outside lawyers admitted AI helped put seven nonexistent case citations into court filings. The useful lesson is much broader: drafting and verification should be separate steps.

1. Ask AI to list every factual claim, number, quote, and source in your draft (best result if you provide the original context in the chat window, and ask it to include “the exact language from the source document so i can command/control + F find it and check your work”).

2. If AI is giving you links, open each source yourself and, well, command + f find it. If it does not directly support the claim, rewrite or delete it.

3. Do not ask AI to “repair” a citation it invented. Start from a real source.

4. If you’re using AI to help you research,

Copy this:

Audit this draft for factual risk. List every claim, number, quote, and named source I must verify manually. Give only URLs already present in the draft. Mark anything unsupported as DO NOT PUBLISH. Do not invent or repair sources.

Have a specific skill you want to learn? Request it here.

FROM OUR PARTNERS

MongoDB Connects Live, Operational Data to the AI Tools Builders Use.

A new Managed MCP Server connects Claude Code, Codex, Grok Build, and Devin to MongoDB Atlas, giving coding agents direct access to live operational data. 

Explore new MongoDB capabilities for the Agentic Era.

📰 Around the Horn

1. Nvidia is paying $6 billion to license AI startup Poolside's technology and hire its engineers, plus investing another $1 billion at a $12 billion valuation, in a bid to build a US-made open-source AI model that can compete with China's.

2. Nvidia is also raising prices on its next-generation Grace Blackwell and Vera Rubin chips by 15-17% starting early next year, adding roughly $5 billion to the cost of a single large data center.

3. Alibaba and Tencent combined to spend $18 billion on AI infrastructure last quarter alone, a number so big it dragged Alibaba's profit down 75% even as its CEO says the investment will pay for itself within three years.

4. OpenAI expanded ChatGPT Ads to 31 European countries last week (Aug 18), meaning free and low-cost ChatGPT users across the continent will start seeing ads (paid subscribers stay ad-free).

5. Blackstone and Hellman & Friedman are embedding a 160-person team of AI engineers, backed by a $1.5 billion partnership with Anthropic, directly inside their portfolio companies to build new AI-powered product lines.

6. Dr. Dre said he uses AI to produce songs and called musicians who oppose the tools "afraid of learning new things," adding that fellow producer Timbaland is a fellow "closet AI user."

7. A new investigation into the data labelers who train AI models found workers in China and Australia get less work over time as AI improves, don't know which companies they're working for, and have little ability to appeal how their performance gets graded.

🍪 Treats to Try

*Asterisk = from our partners (only the first one!). Advertise to 700K+ readers here!

1. *You're paying for Claude but working in the wrong mode. This free guide shows you when to use Chat, Cowork or Code, plus the model settings that stop you burning your usage limit. No tech knowledge required. Click Here for Free Instant Access to the Claude Mode Guide

2. Is Agentic scans your website and shows you exactly what's blocking AI agents from browsing and using it, then tells you how to fix it.

3. FetchSandbox fake-runs the API integrations your AI coding assistant just wrote, complete with real webhooks and failure states, so nothing touches your actual account—free to try, Pro plans from $5/month.

4. Construct is an AI employee with its own cloud computer that reads your email, researches topics, and finishes assigned work while you're away—free trial, then $9/month.

5. USB lets you write one AI skill and install it into 16 different coding assistants, including Claude Code and Cursor, with a single command—free to try (open source).

6. Mnemosyne gives your AI coding assistant a persistent memory that survives between sessions, storing everything as readable notes in an Obsidian vault you can browse yourself—free to try (open source).

📖 Rotating Section (Replace w/ appropriate emoji + daily section title)

New from The Neuron: AI Explained

New episodes air every week on Wednesdays: Spotify | Apple Podcasts | YouTube

P.S: We’re trying to hit 50K subscribers on YouTube this year. Click here to help!

A Cat’s Commentary

Btw: We just launched a robotics newsletter! Sign up for it here.

P.S: Love the newsletter, but only want to get it once per week? Don’t unsubscribe—update your preferences here.