The Tea in Tech has seen tremendous growth over the past few months. So I figured it’s time for a thesis statement. To share what I’m all about. Tune in below.

^ that’s the structure modern marketing was built on. And I’m here to tell you it’s breaking.

Those lanes aren’t distinct anymore. They’re collapsing into one continuous loop of influence, distribution, and trust. And that’s where I come in.

As Tea in Tech has grown, something has become very clear:

This is no longer a passion project, silly little videos I shoot in my spare time. It’s become bigger than that; bigger than me.

Tea in Tech is a media platform. And for good reason:

Only 28% of Americans trust mainstream media today.
(Just 8% of Republicans do.)
Institutional trust is at an all-time low.

I’ve felt it too. The tech narrative has been broken for a while: it’s boring. Full of jargon. Driven by hype cycles. And most irritating to me, it’s covered from the outside in.

So I’m doing the exact opposite.

Tea in Tech is tech, from the inside out. From someone who’s actually built and shipped the products, sat in all the rooms.

You come to me because you trust my point of view. Not just for headlines, but for what they actually mean. The signal amid all the AI noise.

And I take that responsibility very seriously. I’m committed to bringing you the sharpest, most relevant signal (well, Tea) every single day.

Not to be dramatic (ok, maybe just a little dramatic), but I consider this my life’s mission.

So thank you for being here. This newsletter community is my most leaned-in and engaged, and I’m always here for your thoughts, requests, and feedback.

Grateful, committed, and riled up. – M

1. Mark Zuckerberg is building an AI version of himself to help run Meta

My Tea: Zuck should probably work on humanizing himself first tbh.

2. Anthropic explores building its own AI chips

My Tea: There’s chatter that Anthropic is hitting compute limits, especially around Mythos. Suddenly the chip convo makes sense.

3. Leaked OpenAI memo reveals aggressive internal push into Enterprise AI

My Tea: OpenAI is coming for Anthropic’s core business. sips tea

4. Cerebras, an AI infrastructure company, files for the largest IPO of the year

My Tea: The AI gold rush is indeed great for the picks/shovel sellers.

5. Google is in talks with the Defense Department to deploy AI in classified environments

My Tea: In 2018, Google walked away from a DoD project after employee backlash. Oh, how times have changed.

I had an inspiring, thought-provoking conversation with Jaime Teevan, Chief Scientist at Microsoft and one of Time’s 100 most influential people in AI. We covered all things AI and the future of work.

Check out the sneak peek below; full interview drops later this week on socials!

The next big thing in AI? World models.

World models go way beyond LLMs. Instead of just generating text, these systems ingest images, video, and real-world signals to build a working understanding of how the world actually functions; the physics of it all.

That shift is meaningful: it’s what unlocks embodied AI. Think robots and autonomous systems that don’t just respond to prompts, but can perceive, reason, and act in real environments.

Yann LeCun has brought a lot of attention to this space with his new venture focused on world models, reportedly raising a $1B seed at a $5B valuation. I recently spoke with Niantic Spatial, another company in this space that’s training models on real-world geospatial data. The models inform machines so that they can better understand and navigate physical spaces. Their technology blew me away.

The implications of world models are significant. This is where AI starts to move beyond generating content and into interacting with reality. Not just better tools, but entirely new ways of working alongside technology, unlocking both productivity and a more human experience with AI.

I cover world models in Episode 2 of my new exclusive series with LinkedIn and IBM, Innovation in Action. Tune in below.

Peak Silicon Valley this week: Allbirds, the unofficial uniform of VCs and founders in the 2010s, is now rebranding as an AI infrastructure company.

Yes, you read that right. “NewBird AI” plans to buy GPUs and become a compute provider.

Honestly, I don’t hate the pivot. Demand for AI compute is real.

But I do think the sheer audacity of the wool sneakers to GPU-as-a-service rebrand needs to be studied somewhere, somehow…

(And FWIW, I hated the shoes)

Until the next sip,
Meghana