How should we plan for AI in the workplace? And what do we need to rethink about our relationship with this technology to make it work?

That’s the conversation I had with Microsoft last week. I was invited to contribute to Microsoft’s Work Trend Index, their annual report on the future of work.

Honestly, it was a full-circle moment: being interviewed by a company I once interviewed with for a job. And I was in great company, too: Karim Lakhani of Harvard Business School (my former b-school professor!) and Aneesh Raman, Chief Economic Opportunity Officer, at LinkedIn also contributed to the report.

One of the big topics we discussed: as AI agents take on more execution, how do we ensure human judgment, creativity, and agency stay at the center of work?

Tune in below for a clip from my interview, more on my socials later today.

1. Anthropic cuts massive compute deal with SpaceX

My Tea: Safety-first lab meets Elon-powered compute empire. Strategically necessary? Maybe. Brand-aligned? Questionable. See The Steep, below.

2. Google, Microsoft, and xAI agree to U.S. government reviews of their AI models

My Tea: Frontier AI just got a government checkpoint. The question is whether oversight can move as fast as the models do.

3. NVIDIA strikes massive AI infrastructure deals with Corning and IREN

My Tea: Jensen is buying up the bottlenecks: fiber, data centers, power. Makes sense, especially as competition for AI infrastructure heats up.

4. AI traffic is becoming retail’s new growth channel

My Tea: Adobe’s newest research shows AI-driven retail traffic surging and converting better than other channels. Is AI the future of shopping?

5. Vine is back, and it’s fighting AI slop

My Tea: Jack Dorsey is backing Divine, a Vine reboot built around human-made videos. OG Vine brought me unspeakable joy. Rooting for this.

Last week, OpenAI invited me to a dinner with the Codex leadership team at their SF office.

Since then, I’ve gone a little Codex crazy™. I’ve built several automations for my own business workflows (from emails, calendaring, to invoicing), plus a few fun ones, including a daily astrology/affirmation agent (because why not?!) through the Codex desktop app. I’m starting to think of the app as my very own agentic ChatGPT.

Right now, I’m sipping on this: could Codex be the future interface of ChatGPT itself? There’s something to be said about frontier models evolving beyond the chat UI. Brewing up a Tea on this for you… coming soon.

The Tea in Tech went live at Linkedin! Here’s a short clip from my talk at LinkedIn’s Global Offsite, where I spoke to 100+ LinkedIn employees about how to use AI effectively at work (over Tea, of course).

The Tea in Tech is hiring.

I’m looking for two incredible people to join the team:

Both roles will have opportunities to grow within the org.

If this sounds like you or someone you know — spill that tea.

Anthropic’s $4B compute deal with Elon Musk is giving me the ick tbh.

Something feels off. And it’s not about Dario or Elon specifically.

It’s about a company that built its entire brand on ethical, responsible AI…then hit compute limits and made a curiously convenient deal with someone whose public values often seem wildly at odds with its own.

There’s definitely some “enemy of my enemy is my friend” energy here as both companies battle OpenAI.

But I wonder if this deal could cost Dario more than the dollar amount. What is the price on a moral high ground?

My full take:

Digital artist Beeple’s work has made its way to Silicon Valley, and the exhibit is a delightful, absurdist, and pointed commentary on tech and media culture.

See below. Hope you enjoy seeing tech CEOs on all fours as much as I did.

The exhibit is hosted at Node Foundation’s Palo Alto location and is free to the public. Highly recommend stopping by if you are in the Bay Area.

Stay sharp. Stay savvy. Stay steeped (in the Tech Tea). -M