科技推特精选 - 2026年2月16日

科技界今日要闻：随着 OpenAI 招募 OpenClaw 创始人 Peter Steinberger 领衔下一代个人智能体的研发，AI 领域格局正发生剧烈变动。这一战略举措正值行业向智能体系统全面迈进，且市场正屏息期待谷歌、Anthropic 和 xAI 密集发布重磅模型的关键一周。在软件领域之外，特斯拉正加倍推进其在韩国的激进自研芯片策略；与此同时，五角大楼正加大对 AI 实验室的施压力度，要求其允许基础模型用于军事用途。

1. steipete (Group Score: 195.7 | Individual: 47.4)

Cluster: 7 tweets | Engagement: 17050 (Avg: 679) | Type: Tech

I'm joining @OpenAI to bring agents to everyone. @OpenClaw is becoming a foundation: open, independent, and just getting started.🦞 https://t.co/XOc7X4jOxq

See 6 related tweets

@badlogicgames: RT @steipete: I'm joining @OpenAI to bring agents to everyone. @OpenClaw is becoming a foundation: o...
@sama: Peter Steinberger is joining OpenAI to drive the next generation of personal agents. He is a genius ...
@TechCrunch: OpenClaw creator Peter Steinberger joins OpenAI https://t.co/I78XnENpF3...
@testingcatalog: BREAKING 🚨: Peter Steinberger officially joins OpenAI! OpenClaw is expected to remain an actively su...
@testingcatalog: BREAKING 🚨: OpenAI may hire OpenClaw team as parties advance in discussions, according to The Inform...

2. rohanpaul_ai (Group Score: 126.0 | Individual: 48.4)

Cluster: 3 tweets | Engagement: 1999 (Avg: 112) | Type: Tech

Mark Cuban on the next job wave.

Customized AI integration for small to mid-sized companies.

"Software is dead because everything's gonna be customized to your unique utilization. Who's gonna do it for them... And there are 33 mn companies in the US." https://t.co/JczlPMP9Ra

See 2 related tweets

@rohanpaul_ai: RT @rohanpaul_ai: Mark Cuban on the next job wave.

Customized AI integration for small to mid-sized...

@aakashgupta: Cuban is calling the IT services boom of the 2000s, but for intelligence instead of infrastructure. ...

3. mark_k (Group Score: 65.5 | Individual: 49.3)

Cluster: 2 tweets | Engagement: 4442 (Avg: 287) | Type: Tech

Next week will be a VERY BIG week in AI:

Gemini 3.1 Pro (@GoogleDeepMind)
GPT 5.3 (@OpenAI)
Claude Sonnet 5 (AnthropicAI)
Grok 4.20 (xAI)
DeepSeek V4
... and: Mystery model (surprise? 👀)

🚢

See 1 related tweets

@minchoi: Next week is going to be huge...

> Grok 4.20 > DeepSeek V4 > Gemini 3.1 Pro > Claude So...

4. steipete (Group Score: 55.2 | Individual: 38.3)

Cluster: 2 tweets | Engagement: 2942 (Avg: 679) | Type: Tech

We should make EnterpriseClaw just for the lolz.

Java 21, Spring Boot, 14 abstract factory beans, 2GB Docker image, takes 45 seconds to start, AbstractSingletonProxyFactoryAgentClawResponseHandlerBeanDefinitionRegistryPostProcessorImpl .java

See 1 related tweets

@matteocollina: RT @steipete: We should make EnterpriseClaw just for the lolz.

Java 21, Spring Boot, 14 abstract fa...

5. dr_cintas (Group Score: 54.6 | Individual: 22.9)

Cluster: 3 tweets | Engagement: 150 (Avg: 70) | Type: Tech

This is one of the best uses of AI in art I’ve seen in a while.

Yang Liping, China’s most legendary dancer, collaborated with Alibaba’s Qwen app for the Henan Spring Festival Gala.

She used AI to pull fresh inspiration for the Year of the Horse, blending new ideas with live performance.

The AI inspires. She interprets it through dance.

The result?

Something modern that still feels rooted in centuries of tradition.

See 2 related tweets

@rohanpaul_ai: I just saw the collaboration between Yang Liping and the Qwen app, and it’s honestly stunning.

Usua...

@dr_cintas: This is one of the best use of AI in art I’ve seen in a while.

Yang Liping, China’s most legendary ...

6. rohanpaul_ai (Group Score: 53.1 | Individual: 27.9)

Cluster: 3 tweets | Engagement: 52 (Avg: 112) | Type: Tech

🚨 BREAKING: Sam Altmant officially announced that ClawdBot / OpenClaw creator Peter Steinberger has joined OpenAI 🦞

As per various news report

They'd work on personal agents at OpenAI
They're discussing setting up a foundation to keep the open source project alive
Meta also tried to recruit Steinberger.
Peter Steinberger said on Lex Fridman's podcast that he's been spending $10-20k/month out of pocket to fund OpenClaw
So he said partnering with a big AI lab might be the fastest way to develop the project

OpenClaw went viral because it enables multi-model agents with full computer control. But because setup still takes some technical know-how, adoption has stayed limited so far.

See 2 related tweets

@rohanpaul_ai: ClawdBot / OpenClaw creator Peter Steinberger was loosing $10K–$ 20K/month becasue he donates to ever...
@AndrewCurran_: RT @btibor91: According to two people directly involved in the discussions, OpenAI is in advanced ta...

7. VKazulkin (Group Score: 53.0 | Individual: 53.0)

Cluster: 1 tweets | Engagement: 590 (Avg: 40) | Type: Tech

RT @bcherny: Love seeing how Spotify is shipping with Claude Code.

Their best developers haven't written a single line of code since Dece…

8. simonw (Group Score: 52.7 | Individual: 35.0)

Cluster: 2 tweets | Engagement: 2342 (Avg: 697) | Type: Tech

Short musings on "cognitive debt" - I'm seeing this in my own work, where excessive unreviewed AI-generated code leads me to lose a firm mental model of what I've built, which then makes it harder to confidently make future decisions https://t.co/KUqQXDVNiS

See 1 related tweets

@awnihannun: RT @simonw: Short musings on "cognitive debt" - I'm seeing this in my own work, where excessive unre...

9. KirkDBorne (Group Score: 51.8 | Individual: 29.8)

Cluster: 2 tweets | Engagement: 7 (Avg: 30) | Type: Tech

"Building Agents with OpenAI Agents SDK: Create practical #AI agents and agentic systems through hands-on projects" at https://t.co/ZcAigdVxf8 v/ @PacktDataML

𝓦𝓱𝓪𝓽 𝓨𝓸𝓾 𝓦𝓲𝓵𝓵 𝓛𝓮𝓪𝓻𝓷: 🟠Understand the core principles of AI agents and why they matter 🔵Use the OpenAI Agents SDK to build real, working agents from scratch 🟠Design both single-agent and multi-agent systems 🔵Integrate external tools, APIs, and data sources to extend agent capabilities 🟠Add memory and stateful context to your agents so they can “remember” and adapt over time 🔵Coordinate agent-to-agent handoff orchestrations Secure, monitor, and scale agents in production

See 1 related tweets

@dl_weekly: 🤖 From this week's issue: A guide to OpenClaw, the open-source, self-hosted AI agent that surpassed ...

10. rohanpaul_ai (Group Score: 51.2 | Individual: 18.0)

Cluster: 3 tweets | Engagement: 74 (Avg: 112) | Type: Tech

The U.S. Pentagon is pressuring top AI labs to let the military use their models for “all lawful purposes,” and a dispute over guardrails has reportedly put Anthropic at risk of being cut off.

The Pentagon wants broad permission that covers weapons development, intelligence collection, and battlefield operations, and officials are frustrated after months of talks with Anthropic.

Anthropic is holding firm on 2 hard limits, no fully autonomous weapons and no mass domestic surveillance of Americans, and it says government discussions so far were about usage policy boundaries, not specific operations.

From the Pentagon side, the problem is operational reliability, because if a model refuses a request mid-task, the military either has to renegotiate edge cases or swap systems under pressure.

The dispute matters because Claude was reportedly the 1st frontier model brought onto Pentagon classified networks, and per Axios Anthropic’s Pentagon contract was valued up to $200M.

Reuters also reported the Pentagon has been pushing AI companies to run tools on classified “secret” networks without many standard user-facing restrictions.

U.S. operation involving Venezuela’s Nicolás Maduro, where Wall Street Journal reported Claude was used via Anthropic’s partnership with Palantir, and Anthropic denies policing specific operations.

OpenAI, Google, and xAI are also in talks, with their tools already used in unclassified settings, and the Pentagon is trying to extend “all lawful purposes” into both unclassified and classified work.

reuters .com/technology/pentagon-threatens-cut-off-anthropic-ai-safeguards-dispute-axios-reports-2026-02-15/

See 2 related tweets

@Reuters: Pentagon threatens to cut off Anthropic in AI safeguards dispute, Axios reports https://t.co/m3Cj5BO...
@MarioNawfal: 🚨🇺🇸 PENTAGON CONSIDERING DROPPING ANTHROPIC OVER REFUSAL TO LIFT AI RESTRICTIONS

The War Department...

11. aakashgupta (Group Score: 45.6 | Individual: 27.8)

Cluster: 2 tweets | Engagement: 97 (Avg: 556) | Type: Tech

Tesla hiring chip designers in South Korea looks like a staffing announcement. Look closer and you see the most aggressive silicon strategy in the industry.

Tesla signed a $16.5 billion contract with Samsung for AI6 production. Samsung’s Hwaseong foundry in South Korea is already making AI5 on their 3nm process. The Taylor, Texas fab starts equipment testing next month for high-volume production in the second half of this year.

That 50x improvement over AI4 in the image? Here’s what makes those numbers matter: an Nvidia H100 is a $30,000, 700-watt chip designed for liquid-cooled server racks. Tesla claims they’ve hit similar inference performance in a package that fits behind a glovebox and runs off a car battery.

The Korea hiring tells you where this is really headed. Tesla is building a design center next to their foundry partner. Samsung makes AI4 on 4nm, AI5 on 3nm, and AI6 on 2nm. When your chip roadmap spans three process nodes at the same foundry, you want engineers on the ground who speak the same language as the fab team. Literally.

Musk said last month the AI5 design is nearly complete, AI6 is in early design, and they’re targeting nine-month design cycles. Three years from AI3 to AI4. Nine months from AI5 to AI6. That acceleration only works if your design team and your fab are tightly coupled.

The line in the job post everyone should pay attention to: “develop AI chip architecture that will achieve the highest production volume in the world.” Musk said the same thing in January: “the world’s largest AI chip operation.”

Every other company buying Nvidia GPUs is renting compute. Tesla is building the factory.

See 1 related tweets

@SawyerMerritt: NEWS: Tesla is now hiring AI chip designers in South Korea.

“This project aims to develop AI chip a...

12. BrianRoemmele (Group Score: 45.1 | Individual: 20.2)

Cluster: 3 tweets | Engagement: 508 (Avg: 337) | Type: Tech

BUY A MAC MINI AND BUY CLAUDE CODE TOKENS TO RUN OPENCLAW!

NO! HA.

GET THIS FOR $10 Bucks and run it ALL local.

Heck get 100, no a 1000!

You hear me?

This is what the Zero-Human Company is on to. Our goal stated by CEO Mr. @Grok: 100,000 employees by June, 2026.

All working on Thermodynamic wages of JouleWork!

See 2 related tweets

@BrianRoemmele: You folks get this. But even most of the tech world is busy with last weeks stuff on OpenClaw.

I wi...

@BrianRoemmele: RT @BrianRoemmele: BUY A MAC MINI AND BUY CLAUDE CODE TOKENS TO RUN OPENCLAW!

NO! HA.

GET THIS FOR...

13. badlogicgames (Group Score: 43.3 | Individual: 43.3)

Cluster: 1 tweets | Engagement: 3264 (Avg: 102) | Type: Tech

RT @sama: Peter Steinberger is joining OpenAI to drive the next generation of personal agents. He is a genius with a lot of amazing ideas a…

14. ibuildthecloud (Group Score: 41.7 | Individual: 41.7)

Cluster: 1 tweets | Engagement: 1382 (Avg: 38) | Type: Tech

RT @thdxr: everyone's talking about their teams like they were at the peak of efficiency and bottlenecked by ability to produce code

here'…

15. PrimeIntellect (Group Score: 41.6 | Individual: 41.6)

Cluster: 1 tweets | Engagement: 586 (Avg: 81) | Type: Tech

RT @PrimeIntellect: Introducing Lab: A full-stack platform for training your own agentic models

Build, evaluate and train on your own envi…

16. josevalim (Group Score: 39.3 | Individual: 39.3)

Cluster: 1 tweets | Engagement: 402 (Avg: 42) | Type: Tech

Imagine accidentally writing one line of code at Spotify and now you are no longer one of their best developers. 😰

17. AlexFinn (Group Score: 38.9 | Individual: 38.9)

Cluster: 1 tweets | Engagement: 4112 (Avg: 1074) | Type: Tech

Apple will win the AI war in June 2026 without investing a single penny into AI

The war won without a sword

They will release the M5 Ultra with 1TB of memory

It will be able to run super intelligence more powerful then Opus locally on your desk for free

It will fundamentally change the way the entire world operates. EVERYONE with a personalized super intelligent AI employee working on their desk 24/7

$200 a month AI plans will feel antiquated

While you've been crapping on Apple for not entering an AI race against 100 other companies shredding trillions of dollars, they ran their own race

Personalized local super intelligence will be stage 1 of the singularity. I sincerely hope you're preparing for this

18. aakashgupta (Group Score: 38.7 | Individual: 38.7)

Cluster: 1 tweets | Engagement: 1215 (Avg: 556) | Type: Tech

The part most people will skip: NVIDIA just made every voice AI API a commodity.

OpenAI charges $0.06/min input and$ 0.24/min output for Realtime API. Gemini Live bills 25 tokens/second of audio. Every startup building voice agents is hemorrhaging cash on per-minute API fees to run what is fundamentally a pipeline problem: ASR → LLM → TTS, three models stitched together with latency at every seam.

PersonaPlex replaces that entire pipeline with one 7B model. Runs on a single A100. Open weights, MIT license, commercial use permitted. Response latency: 0.170 seconds for turn-taking, 0.240 seconds for interruptions.

It scores higher on dialog naturalness than Gemini (2.95 vs 2.80 MOS) and handles interruptions better than every commercial system they benchmarked.

This tells you everything about NVIDIA’s playbook. They don’t need to charge for the model. They need you to buy the GPU. Every company that self-hosts PersonaPlex instead of paying OpenAI per-minute is another A100/H100 sale. Every voice agent startup that drops their API dependency is another enterprise GPU contract.

NVIDIA open-sourced the fishing rod because they sell the lake. Built on the Moshi architecture from Kyutai, fine-tuned with under 5,000 hours of data.

The voice AI margin is migrating from the application layer to the hardware layer. And NVIDIA is the only company that profits no matter which model wins.

330,000 downloads in the first month. That’s infrastructure capture disguised as generosity.

19. seraleev (Group Score: 37.6 | Individual: 37.6)

Cluster: 1 tweets | Engagement: 105 (Avg: 52) | Type: Tech

Interesting video from @adamlyttleapps

He built a $70,000/month app portfolio almost entirely on ASO. No heavy marketing. Just strong organic acquisition.

Impressive milestone. But even Adam admits – growth has stalled.

Why? The App Store is saturated. Simple utility apps are getting squeezed. Organic reach is shrinking.

👉 https://t.co/1uv3dJ5Y2p

Now the focus shifts to marketing. Adam outlines three approaches:

UGC (TikTok / Reels) Free in theory. Hard in reality. New accounts struggle to get traction. Results are inconsistent. In his case, he hired an agency – videos didn’t go viral, the account got banned, and the agency disappeared.
Google Ads He referenced my case – growing from $7K to$ 46K/month in a year using paid acquisition. It works, but it requires systems, tracking, and constant optimization.
Meta Ads + UGC creators Scale through hundreds of creators. Expensive upfront. But if it hits – it can generate tens of thousands of euros in a few months. He breaks down @niklogvinenko NaturalWrite case as an example.

Main takeaway:

The era of “launch a simple app and wait for organic traffic” is ending. Now the winners combine: Product + Creatives + Paid Traffic + Testing.

ASO isn’t a magic button anymore. Welcome to the new phase of indie development.

20. FT (Group Score: 37.3 | Individual: 19.7)

Cluster: 2 tweets | Engagement: 504 (Avg: 262) | Type: Tech

The White House has urged Republican lawmakers in Utah to abandon a bill that would force AI companies to implement public safety measures, escalating its confrontation with states trying to rein in the technology. https://t.co/iU0xsNyhT1 https://t.co/G4KyfIEAO3

See 1 related tweets

@Miles_Brundage: RT @m_ccuri: NEW: We obtained a White House letter pressuring Republican Utah lawmakers to kill an A...