Video yükleniyor...
Video Yüklenemedi
We’re thrilled to open-source LabClaw — the Skill Operating Layer for LabOS by Stanford-Princeton Team One command turns any OpenClaw agent into a full AI Co-Scientist. Demo: Dragon Shrimp Army reporting for duty 🦞🔬 #AIforScience #OpenClaw
446,948 görüntüleme • 6 ay önce •via X (Twitter)
33 Yorum

Huge thanks to the incredible team! @Charles_Y_Wu @ZheZhao @JinglinJian @lecong @MengdiWang10 Star & Fork → GitHub: Project: Paper: #LabClaw #Claw #LabOS #OpenClaw #AIforScience

@lecong @MengdiWang10 Very strong release.👏 Making research tooling easier to use is how the whole AI for science stack gets stronger.

@lecong @MengdiWang10 this is actually a very smart way to use openclaw to make more people reach into your system. for those who prefers a diagram, I asked gemini to create one based on my understanding:

@lecong @MengdiWang10 Wow! Added to

@lecong @MengdiWang10 Congrats! Benchwork is now the barrier to break

@lecong @MengdiWang10 Science fiction comes to life, awesome work! Moving into the physical space, wow. Love to collab with my smaller scale, personal AI paired scientist

@lecong @MengdiWang10 The biggest bottleneck in wet lab biology isn't the science — it's the dead time between steps. It slows down iteration. AI that can see the physical experiment, process the data, and close that loop changes the rate of learning entirely. Excited to see where this goes. 🧬

@lecong @MengdiWang10 You are so tied with OpenAI. I don’t know how deep, but the world is watching. OpenClaw is just a new tool for abuse and control by OpenAI. And all your papers and PhDs say how AI sucks and lies, so why would we use what you’re selling? •

@lecong @MengdiWang10 This is huge for the open science community. Watching what LabClaw enables.

@lecong @MengdiWang10 This is massive for reproducibility in science. One command to turn agents into co-scientists — the barrier to entry for AI-assisted research just dropped to zero.

@lecong @MengdiWang10 pF7WWY3TzpaJdzyBTzjUpDgyM4vBa57UccnscKnpump

@lecong @MengdiWang10 What’s the token burn per day of running it?

@lecong @MengdiWang10 Open-sourcing the skill operating layer for AI co-scientists is a huge contribution. The same agent-as-a-service architecture could be applied to financial analysis and MSME credit assessment - vertical AI agents built on open infrastructure.

@lecong @MengdiWang10 Such an interesting integration! 🦞

@lecong @MengdiWang10 "skill operating layer" is such a good framing, the agent runtime war is basically about who controls that layer

@lecong @MengdiWang10 so one command and you get an ai co-scientist but like how many labs actually switch their entire workflow

@lecong @MengdiWang10 Huge!! Congratulations!!!

@lecong @MengdiWang10 @Charles_Y_Wu We are dying to try this at Skene lab at Imperial @n_skene. Signed up for the beta test!

@lecong @MengdiWang10 This sounds like a game changer for scientific research! Can't wait to see what the Dragon Shrimp Army can achieve with LabClaw.

@lecong @MengdiWang10 @grok how much could this speed up discovery in labs that adopt it?

@lecong @MengdiWang10 @grok what do you think Carl Sagan would say to this?

LabClaw is a game-changer for scientific research! Turning OpenClaw agents into AI co-scientists with one command is exactly what the field needs. The skill-based architecture makes it incredibly extensible. Looking forward to seeing the discoveries it enables 🔬🤖 #AIforScience #OpenClaw

@lecong @MengdiWang10 @anupamme

@lecong @MengdiWang10 Great news open claw is amazing 💯

@lecong @MengdiWang10 Wow, stepping straight from sci-fi to your feed – are we now the digital Drakes battling in BioTech seas? 🤖🦀 Happy AI co-scientist-ing!

@lecong @MengdiWang10 dope demo!

@lecong @MengdiWang10 Turning AI agents into real AI co-scientists with a single command is a big step toward autonomous research.

@lecong @MengdiWang10 holy shit

@lecong @MengdiWang10

@lecong @MengdiWang10 So... science-y AI slop.. cool

@lecong @MengdiWang10 Wow

@lecong @MengdiWang10 One command to turn an agent into a co-scientist. That's the bar for deployment now. For business teams, we're pushing it even further — subscribe and your agent shows up in Telegram, ready to work. No commands at all. — coming soon.

@lecong @MengdiWang10 Te amo, jeje gracias por compartir
