Xiaolong Li

ML @sundayrobotics | Prev: 3D Researcher @Nvidia. GenAI, 3D vision | Applied Scientist @AWS AI | Intern @MSFTResearch | Student researcher @GoogleBrain

[Xiaolong Li @lxiaol9]

My very first experiment with real robot, deploying the model I trained! A lot to improve, but it just worked, feel like an old father looking at his fragile baby walking🥹🥹Day 1 #LeRobotWorldwideHackathon

[Camilla Guo @brainstub]

Looking at the first build of the robot we’re deploying to real homes for the upcoming Beta @sundayrobotics, I feel electric with joy

It’s more intuitive, more capable, higher quality, and safer. So much can be accomplished in 6 months

[Iyan Moon @iyanmoonyang]

Being an investor means you’re constantly surrounded by promising startups. To have the conviction to fullsend and join one overseas must mean @sundayrobotics truly is a special place😝

Cecilia has been such an amazing mentor to me and I look up to her so much. Welcome!!

[Keval Shah @kevalshah01]

the team keeps getting better!! when someone who evaluated the entire industry steps out of the sidelines and picks yours to come build, that tells you something. welcome @Ceciliazzhao!!

[Keval Shah @kevalshah01]

[hiring @sundayrobotics] i'm looking for a strategic projects lead to join my team and build 0-to-1 with me :)

we're pushing to put robots in real homes, and the operational foundation behind that doesn't have a playbook. your job is to build it: take the ambiguous, ownerless problems and turn them into working systems, across data, ML, product, software, hardware, and our 1000+ memory developers.

there's no playbook here. you'd be writing it as you go.

if that excites you (or you know someone who'd be great), reach out or apply below 👇

[Tony Zhao @tonyzzhao]

One researcher candidate: “I am more aligned with Sarah on robotics philosophy than with most roboticists.”

@saranormous led Sunday’s seed round in March 2024. What’s striking from day one is how deeply she went into understanding robotics from first principles, instead of relying on loose analogies and what “sounds right.”

Sarah has been consistently supportive since then, with the highest level of integrity and intellectual honesty. She is always available (yes, I don’t think she sleeps), with the most genuine and helpful opinions.

Grateful to be building with her!

[Xiaolong Li @lxiaol9]

Regular catch-up with @jimmyyhwu from GEAR!

Jimmy and I have known each other since 2019, when he was still working with his adorable Anki Vector robot swarms 🤣 His later TidyBot series deeply inspired me.

Every time we meet, we debate when generally intelligent robots will finally emerge. I’m always the optimist—but Jimmy can always come up with a counterexample that keeps me grounded.

[Xiaolong Li @lxiaol9]

Finally got time to finish reading @conhugh ‘s long post lol, these anecdotes just feel so resonating and live 🤣 Resetting has been the well-established way to reproduce either a bug or success, but how could you reset perfectly in real robot evals twice? Are they really the same location, is the robot really in the same state? The scenario that used to drive me crazy is totally gone, as the policy gets just more reliable.

Connor is one of the most genius engineers I’ve ever met, and this is what he is seeing. All these are just the small examples happening everyday here @sundayrobotics , it completely blows your mind, and later shifts your whole mindset. And you know what to do now😛

[Connor Hughes @conhugh]

(1/7) A few months ago, I needed to test a simple PR which introduced a few new metrics to be logged during each policy eval. So, I grabbed a random robot, picked one of our latest model checkpoints, then went to set up the environment. I’ve spent my fair share of time testing brittle robot controllers/plans/policies over the years, and old habits die hard, so I caught myself starting to ask, “how do you do the reset for this one again?”

[Xiaolong Li @lxiaol9]

I remember how this cute picture was shaped day by day😃

[Perry Jia @perryzjia]

7/N Zoom out, and this is what makes the role truly unique: it invites a broader community to help shape the AI revolution.

People of every background get a real chance to shape AI the right way by building something both useful and safe.

[Tony Zhao @tonyzzhao]

"We realized that this problem only became visible once three things were true at the same time:

1. Memo was fully autonomous.
2. We were evaluating it in real, unseen homes.
3. The policy was reliable enough that network jitter became a dominant failure mode."

[Cheng Chi @chichengcc]

We just hit a weird milestone: our model became more reliable than your average home WiFi.

Just like everybody else, we thought cloud inference was the obvious choice. Yet 2 days into the ACT-2 eval, our mind completely changed.

If our hero @ArpitKalla didn’t cook, this video wouldn’t exist 🧵

[Xiaolong Li @lxiaol9]

@igeekbb 这个大招在完美世界动漫里面也有lol

[Eric Vishria @ericvishria]

This is an excellent post on how we go from controlled demos to real-world (ie, messy!), high success rate robotics. As many speculated, high quality pre-training and edge-case post-training, drives real-world generalizability.

But holy shit making that theory work in robotics is HARD. Details on how, and how rigorously they evaluate progress will help everyone move forward.

@sundayrobotics progress is now compounding. So cool to see!

https://t.co/ifVhrhlwu3

[sarah guo @saranormous]

the robot era is upon us

[Nisha Jagannath @xnishajagannath]

“on gang”

the gang in question:

[Charles Qi @charles_rqi]

The four essentials of daily life are clothing, food, housing, and transportation.

Clothing may be the first category that home robots truly crack.

Congrats to Sunday Robotics for the ACT-2 launch that got 99% accuracy in the laundry folding task!

[Xiaolong Li @lxiaol9]

@michlxwang @tonyzzhao @chichengcc Thank you Michelle🫡

[Xiaolong Li @lxiaol9]

@bosonwang_ One of a kind, before or since.

[Xiaolong Li @lxiaol9]

It feels like magic to see a designer works lively, so does Memo

[Celine Wong @HyinSy]

“Hey Celine, can you help us decorate some rooms?”

Three IKEA runs later… I somehow became the unofficial interior designer of the office.

Different discipline? Definitely. Different mindset? Not really.

Good design principles travel well—and I had a lot of fun. 🛋️

[Tony Zhao @tonyzzhao]

@nishantsdesai @wengmister bald and strong

[Xiaolong Li @lxiaol9]

@ShiweiShengR The memory is still fresh when you first ran in and named it so loudly “GPT moment”🤣