Dario Amodei had a private dinner at the White House on Sunday night, the first one-on-one meeting between him and President Trump. Earlier this month, Trump had publicly written that the Anthropic CEO was "pretending to be a 'perfect little angel'" after Amodei called for the AI industry to slow down. Two days earlier, a federal appeals court had upheld the Pentagon's decision to label Anthropic a supply chain risk, which bars the US military from using its models.

Meanwhile, TechCrunch's reporters spent the weekend asking a simpler question about Meta's new AI agent: would you actually hand it your bank account?

• Trump hosted Anthropic's CEO for their first one-on-one meeting, after months of open conflict.

• Critics circulated a memo attacking Amodei to the White House before the dinner, according to the Wall Street Journal.

• Trump and House Speaker Mike Johnson are set to meet top AI CEOs at the White House on Tuesday.

• A federal appeals court upheld the military ban on Anthropic's models two days before the dinner.

• A TechCrunch reporter's Meta Muse agent found him unclaimed money, then ran into Meta's ad-business problem.

• Saturday Night Live put an Amodei impression on air, complete with a Gollum-style monologue.

…and a whole lot more that you can read about below.

Trump confirmed to Fox News that he would host Amodei for a private dinner at the White House on Sunday evening. "So now I'm having dinner with Dario from, as you know, very highly respected, from Anthropic," Trump said. According to Axios, Trump personally invited Amodei after he missed Thursday's state dinner for Xi Jinping, where Sam Altman, Jeff Bezos, and Mark Zuckerberg were among the tech executives in attendance.

Anthropic's standing with the administration has been strained for months. The company was sanctioned after refusing in February to let its models be used for fully autonomous weapons or domestic mass surveillance, and on Friday a federal appeals court upheld the Pentagon's supply chain risk designation.

Before the dinner, according to the Wall Street Journal, critics circulated a memo to the White House arguing Amodei has "a long record of attacking Trump," "deep Democratic ties," and "doesn't trust President Trump." Trump's own position hasn't moved. "We're about a year, maybe a year and a half up on China," he said, adding that when asked about the string of security incidents involving advanced AI models, "I'm not worried about it." A White House official told The Hill: "America will lead the world in Super Intelligence, while protecting American consumers." Trump and Speaker Mike Johnson are set to meet top AI executives Tuesday, and with both Anthropic and OpenAI heading toward IPOs valued near or above $1 trillion, one Axios source said, "The WH is trying to get everyone seated at the table."

On TechCrunch's Equity podcast, reporter Sean O'Kane described trying Muse, Meta's new personal AI agent, for himself. One of its first suggestions was to scan for unclaimed funds. It found some, and a check is on its way to him. But he called that "a party-trick type thing," a one-time win. The real pitch Meta is making, he said, is handing Muse your credit card, Gmail, and finances to cancel unused subscriptions and catch double charges, and that is where the trust problem starts. "Meta's business is to sell you ads," O'Kane said. "And yes, they'll make the argument that the more they know about you, the more accurate and interesting the ads will be. Wake me up when we get to that fever dream." He noted Muse initially treated him like a stranger rather than plugging into his Facebook or Instagram history, which made him more willing to try it, though the app gradually pulls those accounts in. Anthony Ha pointed to what makes Meta's move unusual: while OpenAI and Anthropic chase enterprise revenue ahead of their IPOs, Meta is betting on consumers instead.

Software Engineer | Candid Health
San Francisco, CA / New York, NY / Denver, CO (Full-time, Full stack)
Build the autonomous revenue cycle management platform used by healthcare providers.

Solutions Engineer | Mach9
San Francisco, CA (Full-time, Full stack)
Work with customers on AI CAD software built for civil engineering.

Staff Full Stack Engineer | Sully
Mountain View, CA (Full-time, Full stack)
Build the autonomous AI agents that run entire hospital operations.

Product Engineer, New Products | Y Combinator
San Francisco Bay Area (Full-time, Full stack)
Build new products for the accelerator behind Airbnb, Stripe, and Coinbase.

Full Stack Engineer | Ohm
San Francisco, CA (Full-time, Full stack)
Help engineering teams accelerate hardware development and testing.

Forward Deployed Engineer | AviaryAI
Chicago, IL (Full-time, Full stack)
Build outbound AI voice agents for the financial services industry.

Full-Stack Engineer | Encord
London, UK (Full-time, Full stack)
Build the data layer for physical AI.

Teams like Parallel, Turbopuffer, and Wordsmith are already setting the pace on Attio. Get an always-on revenue engine, with agents and workflows that build pipeline, chase every buying signal, and move deals forward with your team. Whether you're working in your browser, inbox, or favorite agent, connect to your customer data in real-time through Attio's web app, MCP, API, and SDK.

Try Attio now

Today's Growth Recommendation

Make money from the weatherA Los Angeles ice cream shop uses weather markets to hedge against cold days—turning unexpected weather into an additional revenue stream.

28 Wishes sells ice cream in Los Angeles. Below 70°F, sales fall about 20%. So the owners put about $20 a day into Kalshi weather markets, taking the cold side. The days that keep customers away now pay something back. See how other owners are doing it

See how hedging works

Let AI Interview You Before It Writes

Most weak AI drafts come from a thin request. You type one line, the AI guesses the rest, and you spend the next ten minutes fixing the guesses. Flip the order. Tell the AI the task, then ask it to question you first, one question at a time, and only draft after you've answered.

Start with the task and who the result is for. Let the AI ask up to five questions, and answer each in a sentence or two. If a question is one you can't answer yet, say so, since that gap is worth knowing about before you write anything. Once the questions are done, ask for the draft and a short list of any assumptions it made. This takes about two minutes and usually saves several rounds of edits. It works for emails, proposals, lesson plans, job applications, and anything else where the AI would otherwise be guessing what you meant.

The Interview-First Prompt

Invite friends & get instant rewards

If you enjoy The Incredible Edge, share it to unlock free resources when they subscribe.