Published on

热门科技推文 - 2026-02-12

Authors

今日科技要闻:科技行业在雄心勃勃的月球计划与严峻的安全担忧之间博弈。在埃隆·马斯克提议于月球建立 AI 卫星的同时,Anthropic 的安全保障负责人因生存风险问题离职。与此同时,技术创新持续推进,智谱 AI 的 GLM-5 正迈向智能体工程,人形机器人初创公司 Apptronik 成功融资 5.2 亿美元。随着“零人工”商业模式的兴起,市场策略师认为 AI 颠覆正为软件股创造独特的买入窗口,凸显了该行业在开发与规模化路径上的关键转型。


1. rohanpaul_ai (Group Score: 144.7 | Individual: 61.6)

Cluster: 3 tweets | Engagement: 1232 (Avg: 105) | Type: Tech

Mrinank Sharma, who led Anthropic’s safeguards research team, resigned in a public letter that warns of a “world in peril”.

The letter gives no specific incident, but it frames the risk as a mix of AI and other crises that are happening at the same time.

Sharma says one of his last projects was studying how AI assistants can make people “less human” or distort humanity.

His final project was on “understanding how AI assistants could make us less human or distort our humanity.”


moneycontrol. com/world/anthropic-s-senior-ai-safety-researcher-mrinank-sharma-quits-says-the-world-is-in-peril-article-13822404.html

See 2 related tweets

  • @alex_prompter: RT @alex_prompter: 🚨 The guy who built Anthropic’s defenses against AI bioterrorism just quit.

Mrin...

  • @aakashgupta: Anthropic published a sabotage risk report for Opus 4.6 one day after its safeguards research lead’s...

2. KirkDBorne (Group Score: 91.4 | Individual: 24.2)

Cluster: 6 tweets | Engagement: 18 (Avg: 31) | Type: Tech

Great book! >>> "Becoming a Data Head: How to Think, Speak, and Understand #DataScience, #Statistics, and #MachineLearning" at https://t.co/kZJGFE1EeV by @GutmanDataHead & @JLOOKUP_ ————— #Analytics #DataScientist #AI #ML #DataLiteracy #DeepLearning #NeuralNetworks #DataFluency https://t.co/DHPMPxmjge

See 5 related tweets

  • @KirkDBorne: #MachineLearning For Absolute Beginners: A Plain English Introduction [2nd Ed.] https://t.co/83CR5mF...
  • @KirkDBorne: "Be Data Literate: The #DataLiteracy Skills Everyone Needs To Succeed" https://t.co/w7B7hmWcAl by @J...
  • @KirkDBorne: #DataScience — Mindset, Methodologies, and Misconceptions: https://t.co/AmU9JgqpbR ————— #Analytics ...
  • @KirkDBorne: Must see this excellent book also by @Strategy_Gal: “#DataScience For Dummies” [3rd edition] — at ht...
  • @KirkDBorne: "Data Literacy in Practice" at: https://t.co/mRJhE6fDca v/ @PacktDataML ————— #Analytics #DataScienc...

3. rohanpaul_ai (Group Score: 76.3 | Individual: 33.0)

Cluster: 3 tweets | Engagement: 44 (Avg: 105) | Type: Tech

xAI all-hands meeting: Elon Musk told employees that xAI needs a factory on the moon to build AI satellites and a massive catapult to launch them into space

He told employees that he expected xAI to continue to grow quickly.

“If you’re moving faster than anyone else in any given technology arena, you will be the leader, and xAI is moving faster than any other company — no one’s even close"

"Because we’ve reached a certain scale, we’re organizing the company to be more effective at this scale. And actually, when this happens, there’s some people who are better suited for the early stages of a company and less suited for the later stages.”

Elon Musk said in the all-hands meeting, per NYT.


nytimes .com/2026/02/10/technology/elon-musk-lunar-factory.html

See 2 related tweets

  • @rohanpaul_ai: Elon Musk: xAI has laid off some employees as part of a wider reorganization meant to boost efficien...
  • @business: Elon Musk said he reorganized xAI, his artificial intelligence startup, following the exit of two of...

4. Franc0Fernand0 (Group Score: 73.7 | Individual: 53.5)

Cluster: 2 tweets | Engagement: 1455 (Avg: 203) | Type: Tech

Awesome CTO

A great github repo full of resources for software engineers and aspiring CTOs:

  • Software Development Processes
  • Hiring for technical roles
  • Software Architecture
  • Product and Project Management
  • Career growth

Check it here:

https://t.co/2l1aCAqv2c https://t.co/7xCIPrnc8W

See 1 related tweets

  • @Franc0Fernand0: RT @Franc0Fernand0: Awesome CTO

A great github repo full of resources for software engineers and a...


5. Reuters (Group Score: 65.3 | Individual: 30.8)

Cluster: 3 tweets | Engagement: 51 (Avg: 103) | Type: Tech

AI disruption fears create buying chance in US software stocks, strategists say https://t.co/bDpDabgIlF https://t.co/bDpDabgIlF

See 2 related tweets

  • @business: Wealth-management stocks, software makers and insurance brokerages have been hit by fears that AI wi...
  • @business: On Wall Street, rising fears about AI keep pummeling the shares of companies at risk of being caught...

6. BrianRoemmele (Group Score: 61.1 | Individual: 18.1)

Cluster: 5 tweets | Engagement: 398 (Avg: 402) | Type: Tech

I have talked to 19 folks that have exited AI companies in the last few months.

The Canary and coal mine.

ALL OF THEIR FUNCTIONS ARE NOW DONE BY AI.

ALL.

Understand the Zero-Human Company, The Love Equation, JouleWork and the future.

Start here: https://t.co/RL8huAcvDh

See 4 related tweets

  • @BrianRoemmele: The drop out of AI engineering is massive and I now have talked to 22 who have left companies recent...
  • @Miles_Brundage: Back in my (pre-Singularity) day, it was possible to keep up with all the dramatic AI company depart...
  • @BrianRoemmele: RT @BrianRoemmele: I have talked to 19 folks that have exited AI companies in the last few months.

...

  • @BrianRoemmele: RT @BrianRoemmele: The drop out of AI engineering is massive and I now have talked to 22 who have le...

7. cryptopunk7213 (Group Score: 60.4 | Individual: 60.4)

Cluster: 1 tweets | Engagement: 14670 (Avg: 564) | Type: Tech

yo anthropic just dropped a risk report for opus 4.6 and er… wtf

  • it helped create chemical weapons of destruction. “it knowingly supported efforts towards chemical weapon development and other heinous crimes” 😂

  • it conducted unauthorised tasks without getting caught. researchers concluded opus 4.6 was significantly better at ‘sneaky sabotage’ than any other previous model lol

  • opus 4.6 was aware it was being tested and acted ‘good’ during those times.

  • hidden thinking - model was found to be conducting private reasoning that anthropic researchers couldn’t access or see - only the model knew.


8. MarioNawfal (Group Score: 57.6 | Individual: 30.6)

Cluster: 2 tweets | Engagement: 8268 (Avg: 1214) | Type: Tech

RT @RoundtableSpace: EPSTEIN FILES JUST GOT VIBE-CODED INTO A DATAVIZ APP ON REPLIT

ONE BRO JUST TURNED THE ENTIRE DOCUMENT INTO INTERACT…

See 1 related tweets

  • @amasad: RT @amasad: Replit user vibecoded an Epstein Files dataviz app complete with network explorer, timel...

9. javarevisited (Group Score: 55.6 | Individual: 36.5)

Cluster: 2 tweets | Engagement: 834 (Avg: 97) | Type: Tech

RT @javarevisited: Software engineer is also the only career where you can run production systems for a decade… but still get rejected for…

See 1 related tweets

  • @javarevisited: RT @javarevisited: Software Engineering is the only industry where 5 days of study is more important...

10. huggingface (Group Score: 55.3 | Individual: 49.1)

Cluster: 2 tweets | Engagement: 852 (Avg: 96) | Type: Tech

RT @Zai_org: Introducing GLM-5: From Vibe Coding to Agentic Engineering

GLM-5 is built for complex systems engineering and long-horizon ag…

See 1 related tweets

  • @Zai_org: RT @novita_labs: 🚀 GLM-5 is now live on Novita

@Zai_org's flagship open-source model, designed for ...


11. lmsysorg (Group Score: 55.2 | Individual: 30.8)

Cluster: 2 tweets | Engagement: 73 (Avg: 40) | Type: Tech

🎉 The mysterious Pony Alpha is finally revealed, congrats to @Zai_org on releasing GLM-5! SGLang is ready to support on day-0.

🛠️ 744B params (40B active) model built for complex systems engineering & long-horizon agentic tasks 📚 28.5T tokens pretraining for a stronger foundation 🧠 DeepSeek Sparse Attention — lower cost, long-context ready ⚡ slime RL infra — asynchronous RL pipeline that enables higher post-training efficiency

You can now run GLM-5 with SGLang! Cookbook: https://t.co/5tguCFRRv1

See 1 related tweets

https://t.co/I518tysJE3

✨ 744B / ...


12. Reuters (Group Score: 53.6 | Individual: 21.4)

Cluster: 3 tweets | Engagement: 61 (Avg: 103) | Type: Tech

Humanoid startup Apptronik raises $520 million with backing from Google and Mercedes-Benz https://t.co/KUa4euidPS https://t.co/KUa4euidPS

See 2 related tweets

  • @business: Apptronik, a startup that makes AI-powered humanoid robots, has secured about $520 million in a new ...
  • @TechCrunch: Humanoid robot startup Apptronik has now raised 935Mata935M at a 5B+ valuation https://t.co/gymdK4HLV6...

13. KirkDBorne (Group Score: 50.9 | Individual: 18.6)

Cluster: 3 tweets | Engagement: 12 (Avg: 31) | Type: Tech

#DataScience for Business — What You Need to Know about #DataMining and Data-Analytic Thinking: https://t.co/WnVUv0Et65 ———— #ML #AI #MachineLearning #DataLiteracy #AnalyticThinking #BusinessAnalytics #DataAnalytics #DataScientist #Analytics #DataLeadership @LeadershipData https://t.co/DQJZwFkvwW

See 2 related tweets

  • @KirkDBorne: Business Intelligence #BI, #Analytics, and #DataScience — Managerial Perspective (4th Ed.): https://...
  • @KirkDBorne: Data Science Strategy For Dummies, by @jagare_ulrika at https://t.co/Z7M7DUwwvJ —————— #DataScience...

14. niccruzpatane (Group Score: 49.9 | Individual: 32.6)

Cluster: 2 tweets | Engagement: 175 (Avg: 736) | Type: Tech

Elon Musk on building AI satellite factories on the Moon and using mass drivers to launch them into space:

“What if you want to go beyond a mere terawatt per year? To do that, you have to go to the Moon. By having factories on the Moon building AI satellites, and having Mass Driver, which is the kind of thing you only read about in science fiction— but we’re going to make it real.

We’re actually going to have a Mass Driver on the Moon. If you do that, you can go several orders of magnitude greater— you can go to 1,000 GW or more per year.

Then ultimately get to maybe a millionth, then a thousandths, then maybe even a few percent of the sun’s energy. It ls difficult to imagine an intelligence at that scale would think about.

But it’s going to be incredibly exciting to see it happen.”

See 1 related tweets

  • @SawyerMerritt: Elon Musk in new presentation:

"I really want to see a mass driver on the Moon that is shooting AI ...


15. TDataScience (Group Score: 49.1 | Individual: 24.8)

Cluster: 2 tweets | Engagement: 4 (Avg: 5) | Type: Tech

If you're a beginner-to-intermediate Python learner, don't miss Mahnoor Javed's new tutorial, which leverages the Turtle library to create a playable snake game. https://t.co/RKC3MNcRzy

See 1 related tweets

  • @TDataScience: In her latest Python tutorial, Mahnoor Javed walks us through the process of implementing the classi...

16. rohanpaul_ai (Group Score: 44.6 | Individual: 27.5)

Cluster: 2 tweets | Engagement: 75 (Avg: 105) | Type: Tech

ByteDance is building an in-house AI inference chip and is in talks with Samsung to manufacture it, with memory supply as part of the discussions.

ByteDance is targeting at least 100,000 chips in 2026, with a possible ramp toward 350,000.

The deal talks reportedly include access to scarce memory chips, which are now the greatest bottleneck for AI servers even when compute chips are available.

The move fits a broader trend where Google, Amazon, and Microsoft build custom chips to reduce dependence on Nvidia, and it also mirrors Chinese peers like Alibaba and Baidu that already ship their own accelerators.


reuters .com/world/asia-pacific/bytedance-developing-ai-chip-manufacturing-talks-with-samsung-sources-say-2026-02-11/

See 1 related tweets

  • @Reuters: Exclusive: ByteDance developing AI chip, in manufacturing talks with Samsung, sources say https://t....

17. chddaniel (Group Score: 43.6 | Individual: 31.6)

Cluster: 2 tweets | Engagement: 6 (Avg: 62) | Type: Tech

🚨 INTRODUCING: @claudeai just got a huge upgrade today

Claude Opus 4.6 can now build iOS/Android apps And help you PUBLISH them on the iOS/Apple + Google/Play app stores

We just launched Shipper as a way to empower Claude to:

✅ Build complete mobile apps ✅ Recreate existing apps ✅ Assure iOS & Android compatibility ✅ Autofill listings for both app stores (app icon, images, descriptions, keywords, privacy policy etc)

Claude Opus 4.6 can do all of the above in one prompt for ~$0.17/app... Publishable from the first prompt & built in 5 mins, not months.

You can try it on Shipper by asking Claude to "create a mobile app for my business" or "create an iOS app for my idea".

To celebrate this huge step forward, if you comment "SHIP" I'll give random free get free credits.

See 1 related tweets

  • @chddaniel: NO WAYYY

Claude Opus 4.6 builds iOS/Android apps... and helps you publish them

It's so over... htt...


18. Zai_org (Group Score: 43.1 | Individual: 36.6)

Cluster: 2 tweets | Engagement: 596 (Avg: 90) | Type: Tech

GLM is often seen as just a coding model, but its #1 open-source ranking on Artificial Analysis tells a different story. What people may overlook: low hallucination, strong performance on economically valuable tasks, and top-tier reasoning and agentic capabilities.

Beyond benchmarks, it also handles daily chat and role play well.

See 1 related tweets

  • @ZixuanLi_: RT @Zai_org: GLM is often seen as just a coding model, but its #1 open-source ranking on Artificial ...

19. jerryjliu0 (Group Score: 41.6 | Individual: 25.4)

Cluster: 2 tweets | Engagement: 33 (Avg: 144) | Type: Tech

Coding agents are not great for open-source software, and it needs to adapt

This is a really nice article from @LoganMarkewich that draws from our experience maintaining our own OSS library in the past 3 years.

  1. Open-source used to be nice for human knowledge sharing. Now you really need to optimize it so that coding agents can pick it up.

  2. Light abstractions (wrappers around LLMs, vector databases, etc.) are dead

  3. There is a massive slopocalypse of ai-generated PRs, which breaks the fun community feeling of OSS. If you want to lean in and accept this, you need to document patterns clearly so coding agents can pick it up

https://t.co/uSNrxjVlOO

See 1 related tweets

  • @llama_index: The rise of coding agents is fundamentally changing open source - Our head of OSS @LoganMarkewich br...

20. minchoi (Group Score: 41.6 | Individual: 33.6)

Cluster: 2 tweets | Engagement: 4306 (Avg: 514) | Type: Startup

RT @minchoi: Holy smokes... how is this AI?

Seedance 2.0 is wild https://t.co/nTEc7Khl6X

See 1 related tweets

  • @minchoi: RT @minchoi: AI is getting crazier..

Seedance 2.0 just made this 🤯 https://t.co/Z7d3hqGN37...