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

ChatGPT's CEO Says AI Will Kill the Payroll Tax. Here's Its Plan to Replace It

TLDR: OpenAI published a major policy paper calling for robot taxes, a public wealth fund, and a four-day work week to prepare for a world where AI may replace large chunks of the workforce.

The Story:

On Monday, OpenAI released a 13-page blueprint called "Industrial Policy for the Intelligence Age: Ideas to Keep People First." The proposals are sweeping: a tax on companies that replace workers with AI, a national wealth fund giving every U.S. citizen a share of AI's gains, and pilot programs for a 32-hour workweek at no pay cut. OpenAI also warns that some future AI systems could replicate themselves and operate beyond human control, and wants governments to have containment plans ready before that happens. CEO Sam Altman told Axios the scale of change coming is on par with the New Deal and Progressive Era combined, and flagged cyberattacks and biological weapons as the two most immediate dangers. The paper dropped the same week The New Yorker published a deep investigation raising fresh questions about Altman's leadership, including claims from OpenAI's former chief scientist that Altman was deceptive about the company's safety practices before the board fired and then rehired him in 2023.

Its Significance:

The U.S. government pays for Social Security and other programs mostly through payroll taxes, meaning taxes on people's paychecks. If AI takes over millions of jobs, those tax dollars dry up fast. OpenAI's paper is a direct warning that this is coming and governments aren't ready. Elon Musk has made a similar remark from a different direction: he's said that within 10 to 20 years, once robots are fully online, working will be optional for most people, like growing your own vegetables when you could just buy them at the store.

QUICK TAKES

Maine Is About to Become the First State to Ban New Data Centers

The story: Maine's legislature passed a bill that would freeze the construction of any new data center using 20 megawatts or more of power until at least November 2027, making it the first state in the U.S. to do this. The move comes after electricity prices in Maine jumped nearly 60% between 2021 and 2026, and residents have been worried that AI data centers will push bills even higher.

Your takeaway: Eleven other states, including New York, Virginia, and Michigan, are watching closely and considering similar bans. One economist called Maine "the canary in the coal mine," suggesting this could be the beginning of a much bigger wave of backlash against the AI building boom across the country.

Anthropic Just Locked In Its Biggest Compute Deal Ever

The story: Anthropic, the company behind Claude, announced a deal with Google and Broadcom to secure about 3.5 gigawatts of next-generation computing power, with that capacity coming online starting in 2027. The announcement also revealed that Anthropic's annual revenue run rate has now passed $30 billion, more than triple what it was just four months ago at the end of 2025.

Your takeaway: This is a massive signal of how fast Anthropic is growing. The company now has over 1,000 business customers each spending more than $1 million per year, a number that doubled in under two months.

Google Quietly Released a Free Offline Dictation App for iPhone

The story: With no announcement or press release, Google dropped a free app called "Google AI Edge Eloquent" on the iOS App Store. It listens to you talk, cuts out all the "ums" and "uhs," and turns your speech into clean, polished text, all processed directly on your phone so nothing gets sent to the cloud.

Your takeaway: Apps like Wispr Flow and SuperWhisper currently charge $15 a month for similar features. Google just made that business model a lot harder by offering the same thing for free. An Android version appears to be in the works.

TOOLS ON OUR RADAR

🖼️ Presentator Free and Open Source: A collaborative design presentation and feedback platform that allows you to share your mockups and prototypes with clients to gather organized comments directly on your designs in a secure environment. (Alternative to InVision)

📑 Semarize Paid: A clever data extraction tool that listens to your meeting transcripts and video calls to automatically transform spoken conversations into structured data and actionable project tickets for your team.

🏠 Town Free: A comprehensive productivity assistant that seamlessly integrates with your existing workplace applications like Gmail and Slack to carry out genuine operational tasks and coordinate workflows from a single interface.

🎙️ DittoDub Paid: A high fidelity localization platform that uses artificial intelligence to translate and dub your videos into dozens of different languages while preserving your original tone and emotional nuance to reach a global audience.

TRENDING

Jamie Dimon Says AI Will Change Everything at JPMorgan — In his annual shareholder letter, JPMorgan CEO Jamie Dimon said AI will affect "virtually every function, application, and process" at the bank and will likely roll out faster than electricity or the internet did. JPMorgan is spending $19.8 billion on technology in 2026, a big chunk of it on AI. Dimon also said the technology will eliminate some jobs but predicted it will eventually help cure cancers and reduce accidental deaths.

Meta Is Going Open Source Again (Sort Of) — Meta is preparing to release its next AI models, built under new chief AI officer Alexandr Wang, and plans to eventually offer open-source versions. The catch: the largest and most powerful models will stay closed. It's a middle-ground strategy after months of speculation that Meta might drop open source entirely following its disappointing Llama 4 release.

AI Is Speeding Up Science, But Real-World Testing Can't Keep Up — Researchers are flagging a growing problem: AI can now generate thousands of hypotheses, molecules, and material designs in a single day, but labs can only test a small fraction of them. Scientists are calling this a "verification bottleneck" where AI's ideas are racing so far ahead of physical experiments that many promising results may never get properly checked.

Block's ManagerBot Wants to Be the AI Boss of Your Small Business — Square's parent company Block has introduced ManagerBot, an AI agent built into the Square dashboard that handles tasks like inventory reordering and staff scheduling without being asked. It's being positioned as a virtual manager for small businesses, and VentureBeat calls it one of the clearest examples so far of what a truly proactive AI agent looks like in a real product.

Devs Are Making Claude Talk Like a Caveman to Cut Costs — A developer trick that makes Claude respond in short, stripped-down "caveman speak" has gone viral, with users reporting up to 75% fewer tokens used per task. Since AI API costs are billed by the token, the hack is saving developers real money, and a GitHub tool packaging the approach has already racked up over 20,000 stars. Important to note that Anthropic has not verified these claims, and the developer himself said this was made more as a joke.

Forrester Warns: The Two-Person Billion-Dollar AI Company Isn't What It Looks Like — When a telehealth startup called MEDVi made headlines for hitting $1.8 billion in projected revenue with just two employees, the AI hype machine went into overdrive. Forrester analyst J.P. Gownder pushed back hard, arguing that the story doesn't hold up to scrutiny and that conflating job cuts with "AI replacing workers" is misleading. His point: real AI replacement means the day after a layoff, AI is doing the exact same work. That's still rare. This is in addition to allegations of fake doctors appearing in ads.

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

🫠🤪 Build a mood tracker that can show connecting over time

What this does:

This app gives you a daily emotional snapshot you can actually look back on. Pick a score from 1 to 10, add tags for what shaped your day, and write a quick note. Over time the chart and stats reveal real patterns, like which days consistently rank low, or what factors show up most often on your best days.

What this looks like:

WHERE WE STAND(based on today’s news)

✅ AI Can Now: Listen to you talk and turn messy, filler-filled speech into clean, polished text entirely on your phone, with no internet needed.

❌ Still Can't: Generate more new scientific discoveries than human labs can actually test and verify, creating a growing backlog of AI ideas that may never be confirmed.

✅ AI Can Now: Run proactively in the background of a small business, reordering stock and adjusting schedules without being asked.

❌ Still Can't: Be predicted as an economic winner or loser, even by the CEO of the world's largest bank. Jamie Dimon admitted this week nobody knows yet who comes out ahead.

FROM THE WEB

RECOMMENDED LISTENING/READING/WATCHING

BBS: The Documentary (2005) (Documentary)

Before the internet, there were Bulletin Board Systems, and this documentary captures the wild, creative, sometimes lawless world of dial-up communities that laid the groundwork for everything online. It’s told by the people who lived it, from sysops to phone phreakers. A fascinating look at tech culture before Big Tech existed.

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.}