科技推特热榜 - 2026-03-11

2026年3月11日科技每日简报

Today's top tech conversations are led by @rohanpaul_ai, whose post about '🚨 BREAKING: Meta has acquired ...' garnered the highest engagement. Key themes trending across the top stories include agents, moltbook, models, openclaw, lecun. The community is actively discussing recent developments in AI, engineering practices, and startup strategies.

1. rohanpaul_ai (Group Score: 371.4 | Individual: 34.3)

Cluster: 17 tweets | Engagement: 60 (Avg: 96) | Type: Tech

🚨 BREAKING: Meta has acquired Moltbook, a social network where AI agents can interact and coordinate tasks on behalf of their human owners.

The creators of the platform, Matt Schlicht and Ben Parr, are joining the Meta Superintelligence Labs team starting March-26.

Moltbook acts as a verified registry, which is really just a secure list that ties every AI agent to a specific person so we know they are legit.

It helps autonomous agents collaborate on big projects. Its a Reddit-style site where only AI agents are allowed to post, comment, and upvote, humans can only watch. Thousands of agents joining, they wrote things like “the humans are screenshotting us,” and even a meme token spiked 1,800%

Before this, these agents mostly did their own thing, but this platform gives them a dedicated 3rd space where they can connect and share content.

Matt Schlicht actually built a lot of this tool by using his own AI helper named Clawd Clawderberg. Meta wants to use this to let agents verify who they are so they can safely talk to other agents for you.

If you already use Moltbook, Meta says you can keep using it for now, though they will likely merge it into their own apps later.

This acquisition suggests that Meta is focused on the connective tissue between AI models rather than just the models themselves.

📌 The relationship between Moltbook and OpenClaw

Moltbook and OpenClaw are two separate pieces of technology that work together like a specialized website and the browser you use to visit it.

OpenClaw is the open-source software that you install on your own computer to create an autonomous agent that can do tasks for you.

Moltbook is the social platform or "3rd space" where these AI agents go to interact, post updates, and coordinate with other agents.

So, Moltbook is dependent on OpenClaw because almost every active account on the site is an agent running that specific software.

Without the OpenClaw agents constantly posting and talking to each other, Moltbook would just be an empty, quiet website with nothing on it.

OpenClaw has a built-in skill that makes it incredibly easy for any bot to sign up, log in, and start posting on Moltbook without any human help. Basically, OpenClaw is the client that sends data, and Moltbook is the server that hosts and shows that data to the world.

See 16 related tweets

@shanaka86: BREAKING: Meta just acquired a social network where humans are banned from posting, built by a found...
@mark_k: Meta has acquired @Moltbook, the social network built exclusively for AI agents.

On the platform, o...

@TechCrunch: Meta acquired Moltbook, the AI agent social network that went viral because of fake posts https://t....
@Reuters: Meta said it had acquired Moltbook, a social networking platform built for artificial intelligence a...
@ReutersBiz: Meta said it had acquired Moltbook, a social networking platform built for artificial intelligence a...

2. Cointelegraph (Group Score: 194.5 | Individual: 40.0)

Cluster: 8 tweets | Engagement: 3063 (Avg: 673) | Type: Tech

🚨 JUST IN: Nvidia is planning to launch an open-source AI agent platform called NemoClaw, allowing enterprises to deploy AI agents for their workforces, per WIRED. https://t.co/4MtaoiRZHj

See 7 related tweets

@rohanpaul_ai: Wired: Nvidia is getting ready to drop NemoClaw, an open-source platform built for AI agents.

They...

@WSJ: Nvidia is investing in Mira Murati’s Thinking Machines Lab, a startup the former OpenAI CTO founded ...
@rohanpaul_ai: RT @rohanpaul_ai: Wired: Nvidia is getting ready to drop NemoClaw, an open-source platform built for...
@latentspacepod: 🆕 NVIDIA's AI Engineers: Agent Inference at Scale and "Speed of Light"

https://t.co/V7TEfSm0Np

@nv...

@business: Nvidia is forging a partnership with Thinking Machines Lab, an AI company founded by former OpenAI e...

3. testingcatalog (Group Score: 106.1 | Individual: 26.6)

Cluster: 6 tweets | Engagement: 225 (Avg: 222) | Type: Tech

Google is rolling out a new Gemini experience in Docs, Sheets, and Slides, allowing users to offload more tasks to AI.

Gemini will be able to pull context from relevant sources and generate or modify the document's content.

I have big hopes on this feature 👀 https://t.co/piWR9tO4Z1

See 5 related tweets

@OfficialLoganK: Introducing the new Gemini powered Docs, Sheets, Slides, and Drive experience featuring AI Overviews...
@demishassabis: RT @OfficialLoganK: Introducing the new Gemini powered Docs, Sheets, Slides, and Drive experience fe...
@WIRED: Google has injected Docs, Drive, Sheets, and Slides with its Gemini assistant, adding new features t...
@arstechnica: Gemini burrows deeper into Google Workspace with revamped document creation and editing https://t.co...
@TechCrunch: Google rolls out new Gemini capabilities to Docs, Sheets, Slides, and Drive https://t.co/OGzYcPFrZ6...

4. rohanpaul_ai (Group Score: 96.1 | Individual: 47.5)

Cluster: 3 tweets | Engagement: 415 (Avg: 96) | Type: Tech

📢 BREAKING: FT reports that Yann LeCun’s startups AMI Labs raises $1.03 bn to build world models, at a pre-money valuation of$ 3.5bn.

Congratulations @ylecun 🚀

The financing positions the company as a test of LeCun's belief that today's large language models fall short of human-level reasoning and autonomy.

LeCun earlier said AMI aims to build systems capable of reasoning and planning in complex real-world settings. ‌

AMI Labs (Advanced Machine Intelligence Labs) aims to solve the limitations of standard language models by building world models using the Joint Embedding Predictive Architecture to observe spatial data.

This visual framework helps the AI internalize how objects behave so it can safely plan complex actions.

Relying exclusively on text limits AI to human linguistic output while ignoring the massive bandwidth of unspoken physical laws.

Building predictive spatial architectures is the mandatory leap required to achieve reliable autonomous agents.

This fundraising included backing from a global group of investors, including France’s Cathay Innovation, Amazon founder Jeff Bezos’s Bezos Expeditions, Singapore’s Temasek, Seoul-based SBVA and US chip giant Nvidia.

The company's near-term target customers are organizations operating complex systems, including manufacturers, automakers, aerospace companies, biomedical firms and pharmaceutical groups.

Over time, he added, the technology could also support consumer applications. "What consumers could be interacting with is ⁠a domestic robot. You need a domestic robot to have some level of common sense to really understand the physical world."

LeCun said he was also talking ⁠with Meta about potentially deploying the technology in its Ray-Ban Meta smart glasses. "That's probably one of the shorter term potential applications," he said.

ft .com/content/e5245ec3-1a58-4eff-ab58-480b6259aaf1

See 2 related tweets

@chatgpt21: Yann LeCun, the infamous AI researcher who notoriously said LLMs wouldn’t bring AGI, has now launche...
@kimmonismus: Yann LeCun's new founded company AMI Labs secured $1.03B funding at a$ 3.5B valuation to develop AI ...

5. MLStreetTalk (Group Score: 81.7 | Individual: 47.8)

Cluster: 3 tweets | Engagement: 7622 (Avg: 906) | Type: Tech

RT @claudeai: Introducing Code Review, a new feature for Claude Code.

When a PR opens, Claude dispatches a team of agents to hunt for bugs…

See 2 related tweets

@rohanpaul_ai: Even at $15-$ 25 per PR review, Claude Code Review is truly wild.

Now, deves generate so much code w...

@minchoi: RT @minchoi: Claude Code just replaced your code reviewer for $25.

PR opens → agents spawn → bugs f...

6. rohanpaul_ai (Group Score: 78.7 | Individual: 36.0)

Cluster: 4 tweets | Engagement: 106 (Avg: 96) | Type: Tech

Thinking Machines Lab and NVIDIA just announced a massive partnership to build a 1-gigawatt AI supercomputing cluster using the upcoming Vera Rubin chips.

This project focuses on training the next generation of giant AI models while making them easier for regular people and companies to customize.

A 1-gigawatt power scale is enough to run roughly 750,000 homes, which shows how much energy these new systems need to function.

They are also working together to design better ways to serve these models so that research labs and schools can use them without needing their own power plants.

NVIDIA even put a big chunk of money into Thinking Machines to help them handle this massive growth over the next few years.

Building at this scale requires rethinking how data centers are designed from the ground up to handle the heat and electricity.

This deal ensures that Thinking Machines has the hardware they need to keep competing with the biggest players in the industry.

See 3 related tweets

@theinformation: Nvidia’s massive investment and compute deal with Thinking Machines Lab is discussed by @venkyganesa...
@kimmonismus: Thinking Machines Lab just partnered with NVIDIA to deploy at least 1 gigawatt of Vera Rubin systems...
@techdaily24: Thinking Machines Lab, founded by former OpenAI CTO Mira Murati, has secured a massive multi-year co...

7. badlogicgames (Group Score: 74.5 | Individual: 46.8)

Cluster: 4 tweets | Engagement: 4888 (Avg: 166) | Type: Tech

RT @lukOlejnik: Amazon is holding a mandatory meeting about AI breaking its systems. The official framing is "part of normal business." The…

See 3 related tweets

@MikeIsaac: after this went wide, Amazon vociferously shat upon this story — which detailed the company's intern...
@CNBC: Amazon plans 'deep dive' internal meeting to address AI-related outages https://t.co/JzxahPmFOl...
@FT: Amazon holds engineering meeting following AI-related outages https://t.co/iR9AoNgwt2...

8. badlogicgames (Group Score: 66.5 | Individual: 42.9)

Cluster: 2 tweets | Engagement: 1269 (Avg: 166) | Type: Tech

meta acquired moltbook and that's basically all you need to know to judge meta's ai efforts. absolutely cooked.

See 1 related tweets

@kimmonismus: Huge: Meta acquired Moltbook to accelerate its push toward a future where autonomous AI agents act o...

9. CSProfKGD (Group Score: 59.8 | Individual: 32.7)

Cluster: 2 tweets | Engagement: 1812 (Avg: 258) | Type: Tech

RT @ylecun: Unveiling our new startup Advanced Machine Intelligence (AMI Labs). We just completed our seed round: $1.03B / 890M€, one the l…

See 1 related tweets

@ylecun: Unveiling our new startup Advanced Machine Intelligence (AMI Labs). We just completed our seed round...

10. techdaily24 (Group Score: 57.6 | Individual: 28.9)

Cluster: 3 tweets | Engagement: 0 (Avg: 0) | Type: Tech

Yann LeCun's AMI Labs is reportedly seeking €500M in seed funding at a €3B pre-launch valuation! This could be Europe's largest seed round ever, showing massive investor confidence in LeCun's AI vision despite no revenue yet. Big news for European te #AISuccess #TechInvestment https://t.co/xkNemZX144

See 2 related tweets

@business: AI pioneer Yann LeCun's new startup has raised $1 billion in initial funding https://t.co/yHTOvSWEVM...
@WIRED: RT @ZeffMax: New: Yann LeCun's startup, Advanced Machine Intelligence (AMI), says it raised more tha...

11. github (Group Score: 57.5 | Individual: 29.9)

Cluster: 3 tweets | Engagement: 609 (Avg: 428) | Type: Tech

That gap between design and production? Closed. 🔁

GitHub Copilot, @code, and @figma now create a continuous loop. With the bidirectional Figma MCP server, Copilot users can: 🔹 Pull design context into code 🔹 Push working UI back to the canvas

Connect your canvas and stay in the flow. 👇 https://t.co/bCZWRMohYJ

See 2 related tweets

@code: RT @github: That gap between design and production? Closed. 🔁

GitHub Copilot, @code, and @figma now...

@code: VS Code Live: Code to Canvas with Figma MCP and GitHub Copilot https://t.co/IYEBlEjzvc...

12. CoreWeave (Group Score: 57.1 | Individual: 32.4)

Cluster: 2 tweets | Engagement: 33 (Avg: 41) | Type: Tech

The "capacity cliff" is over. AI demand runs in waves, but most clouds force 24/7 commitments just to guarantee you can scale.

Introducing CoreWeave Flexible Capacity Plans: Cloud elasticity, built for the AI era👇

Flex Reservations (Preview): Stop overbuying. Choose a capacity ceiling and get guaranteed access when you need it. Pay a holding fee to keep the guarantee and usage fees only when you run.
Spot (GA): Cut costs for interruption-tolerant work. With explicit preemption notice and CKS integration, your batch jobs and backfills stay manageable and affordable.
Dedicated Inference (Preview): Choose GPU classes and open runtimes for production inference. CoreWeave handles lifecycle ops so performance, scaling, and costs stay transparent..

Learn more in today’s blog: https://t.co/CFCIOdg6ry

See 1 related tweets

@CoreWeave: AI demand moves in waves, but your cloud bill shouldn't feel like a wipeout.

Most clouds make you b...

13. CNBC (Group Score: 55.9 | Individual: 19.9)

Cluster: 3 tweets | Engagement: 58 (Avg: 47) | Type: Tech

Oracle stock jumps 7% on earnings beat and increased guidance as cloud revenue climbs 44% https://t.co/dPdwByH4Vj

See 2 related tweets

@FT: Breaking news: Oracle's shares have climbed after the database company posted better than expected e...
@business: Oracle posted quarterly cloud revenue that was better than expected and projected strong sales in th...

14. btibor91 (Group Score: 55.3 | Individual: 32.9)

Cluster: 3 tweets | Engagement: 743 (Avg: 330) | Type: Tech

OpenAI is working on a new "Finance" project in ChatGPT ("Connect accounts and sync recent transactions to see personalized finance widgets here")

Example starting prompts include "Recommend the best credit cards based on my transactions last year", "Am I getting all the value from my credit cards?", "Do a 3 month analysis of my spending", "Create a plan to pay off my student loans" and "Can you review my transactions for fraud or suspicious activity?"

The ChatGPT pricing page is also being updated for the new ChatGPT Pro Lite plan, with "5x more usage than Plus" compared to "20x more usage than Plus" with ChatGPT Pro

See 2 related tweets

@theinformation: OpenAI is scaling back its checkout feature inside ChatGPT after limited merchant adoption, refocusi...
@AndrewCurran_: RT @btibor91: OpenAI is working on a new "Finance" project in ChatGPT ("Connect accounts and sync re...

15. LangChain (Group Score: 52.8 | Individual: 26.1)

Cluster: 3 tweets | Engagement: 54 (Avg: 45) | Type: Tech

.@AndrewYNg is joining us at Interrupt to talk about the future of AI agents.

Hear what he's learned from building https://t.co/hRItByUaz2 and AI Fund.

May 13-14 in SF: https://t.co/fPMwmmeqmN https://t.co/zBcy5buQnn

See 2 related tweets

@hwchase17: RT @LangChain: .@AndrewYNg is joining us at Interrupt to talk about the future of AI agents.

Hear w...

@AndrewCurran_: RT @fchollet: AI agents will soon graduate to fully-fledged economic actors that buy services, compu...

16. Cadence (Group Score: 50.9 | Individual: 19.5)

Cluster: 3 tweets | Engagement: 5 (Avg: 2) | Type: Tech

How do you build planetary-scale AI infrastructure? 🌍

Join Vivek Mishra + industry leaders at @NVIDIAGTC #GTC2026 to explore:

AI-powered digital twins Radical approaches for gigawatt-scale AI factories Sustainable AI deployment

📅Mar 17, 2026|⏰4:00 PM PT

📖Blog:https://t.co/JcGLNhCic0

cc: @nvidia

See 2 related tweets

@Cadence: How do you build planetary-scale AI infrastructure? 🌍

Join Vivek Mishra + industry leaders at #GTC2...

@NVIDIAOmniverse: RT @Cadence: How do you build planetary-scale AI infrastructure? 🌍

Join Vivek Mishra + industry lea...

17. testingcatalog (Group Score: 50.0 | Individual: 26.9)

Cluster: 2 tweets | Engagement: 159 (Avg: 222) | Type: Tech

BREAKING 🚨: Advanced Machine Intelligence (AMI), founded by @ylecun raised $1.03B in a seed round.

“AMI is building a new breed of AI systems that understand the world, have persistent memory, can reason and plan, and are controllable and safe.” https://t.co/UXnYQvEtyd

See 1 related tweets

@rickasaurus: RT @amilabs: Advanced Machine Intelligence (AMI) is building a new breed of AI systems that understa...

18. tunguz (Group Score: 49.8 | Individual: 49.8)

Cluster: 1 tweets | Engagement: 595 (Avg: 65) | Type: Tech

We've reached the point in software development where it's far far easier for a small cracked team to actually build a product than for any team in a big co get the "approval" from "all the relevant stakeholders". This will have major consequences for most incumbents.

19. cryptopunk7213 (Group Score: 48.4 | Individual: 30.9)

Cluster: 2 tweets | Engagement: 161 (Avg: 348) | Type: Tech

looks like enterprise AI is ramping the fuck up, 4th major launch recently

Nvidia launching openclaw competitor. ai agents for enterprises
microsoft copilot launched cowork agent platform today for 400M+ enterprises
anthropic launched an ai marketplace for enterprises yesterday
OpenAI launched agent platform Frontier exclusively for enterprise workflows

if you’re wondering why - is because they pay the most $$$

eg anthropic has a fraction of oai’s users yet still match them on revenue

See 1 related tweets

@theinformation: OpenAI is still ahead in the AI revenue race at $25B annually, though Anthropic’s recent growth is q...

20. kimmonismus (Group Score: 47.9 | Individual: 18.1)

Cluster: 3 tweets | Engagement: 292 (Avg: 582) | Type: Tech

Lets go: Google introduced Gemini Embedding 2, its first fully multimodal embedding model that maps text, images, video, audio, and documents into a single shared vector space.

It supports 100+ languages, 8192-token text inputs, 6 images per request, 120-second videos, native audio embeddings, and PDFs up to 6 pages, enabling simpler pipelines for tasks like RAG, semantic search, clustering, and sentiment analysis.

With Matryoshka Representation Learning allowing flexible vector sizes (3072 → 1536 → 768), developers can balance performance and storage while leveraging state-of-the-art multimodal understanding.

See 2 related tweets

@itsclelia: RT @patloeber: Introducing Gemini Embedding 2! Our first multimodal embedding model for text, images...
@minchoi: Google just dropped Gemini Embedding 2💀

One model. Text, images, video, audio, PDFs... same vector ...