Emerson S

The feature is called “Name Tag.” Point your glasses at a stranger and @Meta AI tells you exactly who they are. I agree the use cases are real and the tech is ready. But what’s your opinion, does knowing someone’s name before they introduce themselves change the whole energy of meeting people? Or even privacy? techcrunch.com/2026/02/13/met…

@skyler_chan_ Very ready

i don’t know if yall are ready for our demo day reveal

@chetan_ A humanoid 👀

guess what's inside 👀

@tbpn @mcuban I don’t believe this lol, we will all merge one day. Both will exist!

.@mcuban says humanoid robots won't last more than 5-10 years. Instead, we'll "design the house to fit the robot, and design the robot to fit the house." "You could create a house where the pantry, the refrigerator, and the washing machines were hidden behind the garage, if you even have a garage. That way you could redesign the house so that all the living space was for people."

@ychngji6 Impressive!

Introducing ForeHOI - One small step toward Physical AI! Accepted to #CVPR2026 🎉 It turns real human experiences — the natural way you touch, grab & use objects every day — into 3D instantly. This will equip robots to perform precise sequential actions in the physical world 🚀

@packyM @PimDeWitte This is incredibly fascinating

There is a tremendous amount of progress happening in World Models. Multiple labs have raised more than $1B. WMs were the star of GTC. They are a real path to embodied AI. So @PimDeWitte & I wrote a comprehensive 19k word overview of World Models. notboring.co/p/world-models

We are building the future with our hands and our data and our curiosity and it is absolutely wild that this is real life.

The reason I’m so bullish on humanoid robots specifically isn’t just the hardware….. It’s the data advantage. 8 billion humans already doing everything a robot needs to learn. In the same environments. With the same objects. Every single day.

@hendrikomg Love this demo

A tiny ego world model. Metric depth. 5 FPS local on a MacBook. <10 min of training data. Not the most general. But extremely relevant for real robotics.

@maticrobots This is so cool

With Matic's new smart home integration, you can control cleaning from anywhere with the help of Siri and Google Assistant. Ready to roll? Try asking it to clean your kitchen.

@audrlo 👏🏽👏🏽🥳

we are launching tomorrow. stay tuned.

I’ve been deep in the physical AI and robotics space and this piece from @AnneliesGamble is the most interesting take I’ve read in a while. Everyone is racing to build a robot that can do everything. But a UC Berkeley professor just pointed out that generality hasn’t worked in production yet. And the path forward is actually the opposite of what most people think. Start with one specialized task. Get it to work reliably every single time. Deploy it. Let it collect real functional data from actual production environments. Then use that data to expand into adjacent skills. Generality isn’t where you start. It’s where you end up after you ship.

@andrewchen I genuinely believe access to AI and compute is the great equalizer of our generation

Tired: Universal basic income Wired: Universal basic compute

$2.1M into a wearable that trains your brain to focus better and stress less. In a world drowning in notifications, distractions, and information overload, this might be the most useful product I’ve seen all month.

@thatssodhawal @mavehealth @Google @ufc @ycombinator @BlumeVentures Congrats! This is big!!!

We've raised $2.1M to fix your focus. Our wearable headset @mavehealth improves attention & stress regulation in just 20 minutes a day for users at @Google, @ufc, @ycombinator. Backed by @BlumeVentures, alongside existing and new investors. Order now at mavehealth.com 🇺🇸🇮🇳

The past week in robotics funding alone Mind Robotics raised $500M backed by a16z and Accel. Sunday raised $165M and hit unicorn status at $1.15B. Over $1.2B into physical AI robots in a single week. Sunday is building Memo, a household humanoid with 1,000 people already on the waitlist. Mind Robotics is spun out of Rivian to build robots that can actually handle factory work humans still do. 2026 is not warming up. It already started. techcrunch.com/2026/03/12/hum…

@itsbenjyyy @TenkaraAI @trueventures Congrats Ben!

Ten years ago I was building factories. Today I'm building the tools I wish I had inside them. @TenkaraAI raised $7M led by @trueventures.

@niccruzpatane Tesla all the way

I know which one I’m taking on date night.

@JulianSaks Love it!!!

v2 community design 🔥🔥🚀🚀

I think people underestimate how messy the data problem in robotics actually is. You can’t just scrape the internet the way text models do. If you want a robot to learn something simple like clearing a table or loading a dishwasher, you need someone doing the task in the real world, from a first person perspective, interacting with real objects. And you need that example thousands of times. That’s why companies like @GenrobotAI and @sundayrobotics are interesting. GenRobot is building egocentric datasets by capturing everyday human behavior across thousands of homes and tasks, while Sunday is exploring how this kind of human demonstration data can translate into real world robotic capability. What’s also interesting from recent research is that more diverse data alone might not be the answer. Models seem to learn faster when the same task is paired across different embodiments. A human doing it, a robot arm doing it, another robot doing it. That alignment helps the model understand how actions translate across bodies. Which honestly makes sense. Robotics isn’t just perception, it’s behavior and physics. And the real advantage in robotics might end up being whoever builds the best dataset of real world interaction.

@perryzjia Applied 👏🏽🥳

Want to break into robotics but don't have prior experience? Sunday is opening the door! We're hiring Memory Developers to help train the next generation of general-purpose robots. As a Memory Developer, you won’t just collect data—you will shape the “memory” of our robot and play a vital role in training the AI that powers it. Using provided hardware, you will record high-quality demonstrations of household tasks that directly influence how our robot learns to perceive and act in real environments. This is a flexible, remote, part-time role for individuals who are excited to contribute to the future of robotics. Compensation follows our pay-per-task model: - $30 per hour of approved data during a 1-week screening period - Up to $60 per hour of approved data for more complex tasks after screening If you're excited about robotics and want to be part of building something real, this is your chance to get started. 🚀 Apply here: jobs.ashbyhq.com/sunday/f92adbe…