精选科技推文 - 2026年6月26日

2026年6月26日科技每日简报

Today's top tech conversations are led by @AWSNewsroom, whose post about 'RT @ajassy: Really enjoyed my ...' garnered the highest engagement. Key themes trending across the top stories include https, openai, agents, google, including. The community is actively discussing recent developments in AI, engineering practices, and startup strategies.

1. AWSNewsroom (Group Score: 372.3 | Individual: 36.3)

Cluster: 19 tweets | Engagement: 558 (Avg: 89) | Type: Tech

RT @ajassy: Really enjoyed my meeting with Prime Minister @narendramodi about what’s ahead for Amazon in India.

We’ve been serving customers, sellers, developers, startups, and enterprises in India for more than a decade and just getting started.

Shared that we’re investing $48 billion over the coming five years, including$ 21+ billion in AI and cloud infrastructure.

By 2030, we plan to support 3.8 million jobs, enable $80 billion in ecomm exports, and bring benefits of AI to 15 million small businesses and 4 million government school students.

Excited about what’s ahead. Still early days for what we can build. https://t.co/sxkRka2Qng

See 18 related tweets

@moneycontrolcom: #WATCH | 🇮🇳 Amazon CEO Andy Jassy met PM Narendra Modi at Seva Teerth and praised his leadership, ca...
@narendramodi: A great meeting with Mr. Andy Jassy. I welcome Amazon's record $48 billion investment in India. This...
@ANI: #WATCH | President and CEO of Amazon, Andy Jassy says, "The Prime Minister's vision for the last 12 ...
@mattsgarman: Hundreds of thousands of organizations across India are already building on @awscloud. Today we're a...
@ANI: President and CEO of Amazon, Andy Jassy tweets, "Enjoyed my meeting with Prime Minister Narendra Mod...

2. victormustar (Group Score: 307.6 | Individual: 52.0)

Cluster: 9 tweets | Engagement: 586 (Avg: 77) | Type: Tech

RT @ornith_: Aloha! 🌺 Meet Ornith-1.0, a family of open-source LLMs specialized for agentic coding.

Ornith-1.0 spans the full parameter sizes including 9B Dense, 31B Dense, 35B MoE, and 397B MoE. It achieves state-of-the-art performance among open-source models of comparable size on coding benchmarks including: ✅Terminal-Bench 2.1(77.5) ✅SWE-Bench(82.4 on verified, 62.2 on pro, 78.9 on Multilingual) ✅NL2Repo(48.2) ✅SWE Atlas(41.2 on QnA, 42.6 RF, 39.1 TW) ✅ClawEval(77.1)

Post-trained on top of gemma4 and qwen3.5, Ornith-1.0 employs a novel self-improving training strategy in which reinforcement learning is used to generate not only solution rollouts, but also the task-specific scaffolds that drive those rollouts. By jointly optimizing the scaffold and the resulting solution, the model generate higher-quality solutions in agentic coding.😎

All models are released under the MIT license, enabling full commercial and research use.

📖Tech Blog: https://t.co/qT9N2HYWFn 🤗Huggingface: https://t.co/PRrwqjeBtM

See 8 related tweets

@0xSero: New best in class models for:
8-16gb memory
16-96gb memory
256-512gb memory

Gotta improve my...

@Parul_Gautam7: Open-source coding models are moving ridiculously fast.

MIT licensed, agent-focused, and benchmark ...

@teortaxesTex: almost all engagement with this seems synthetic I'd be interested in serious people like @_xjdr givi...
@ivanfioravanti: Time to test it! Download starting in 3,2,1! 🚀\n\nQT @ornith_: Aloha! 🌺 Meet Ornith-1.0, a family of...
@Marktechpost: Most agentic coding is a capable model wrapped in a hand-built harness. The scaffolding is human-eng...

3. shiri_shh (Group Score: 285.9 | Individual: 32.0)

Cluster: 10 tweets | Engagement: 56 (Avg: 647) | Type: Tech

Traditional observability is completely broken for the agent economy.

You can't monitor a swarm of autonomous AI agents using tools built for legacy SaaS. The failure modes are completely unpredictable.

If AI is writing the code and running the workloads, it only makes sense that AI fixes it when it breaks.

Self-healing software is officially here.\n\nQT @shcallaway: We raised $8m to build self-healing software.

In 2026, software moves fast.

But monitoring and observability are still manual and slow.

@sazabi is a next-generation observability platform for fast-moving engineering teams.

Not another AI SRE.

Not another LLM observability tool.

A new, general solution to observability that works for any workload (including agents) and leverages AI for maximum speed and automation.

Time to suit up.

See 9 related tweets

@alexisaftalion: the guy is cracked, the product is great, the branding is nice. excited to see what comes next! Soli...
@Parul_Gautam7: Software engineers after Claude Code + Sazabi: https://t.co/Hiq55DdDdW\n\nQT @shcallaway: We raised ...
@Scobleizer: Nobody saw this gap until the AI-coding-agent wave hit hard:

Cursor, Claude Code, and Codex fundame...

@cgtwts: Software engineers when they realize they’ll never have to wake up at 3am to fix a bug again: https:...
@wallstengine: Sazabi raised an $8M seed led by J2 Ventures, Village Global and YC to build AI-native observability...

4. wallstengine (Group Score: 261.8 | Individual: 33.7)

Cluster: 10 tweets | Engagement: 58 (Avg: 132) | Type: Tech

TRUMP ADMINISTRATION ASKED OPENAI TO STAGGER GPT-5.6 RELEASE OVER SECURITY CONCERNS, PER THE INFORMATION

OpenAI CEO Sam Altman reportedly told staff that GPT-5.6 will be released first in a limited preview to a small group of partners after a federal government request.

In a memo, Altman said the government would be “approving access customer by customer during this preview period.”

The request reportedly came after OpenAI worked with government agencies over the past month to preview the model, including the Office of the National Cyber Director and the Office of Science and Technology Policy.

The Information also reports Commerce Secretary Howard Lutnick cautioned OpenAI against launching without approvals from other agencies.

Altman told staff this is “not our preferred long term model” and said OpenAI will work with the government and the rest of the industry on a more sustainable process for future releases.

The situation comes after the White House’s recent clash with Anthropic, where export controls reportedly forced Anthropic to withdraw its Fable and Mythos models from the market over national security concerns.

See 9 related tweets

@MTSlive: SITUATION BREWING: GPT-5.6 has reportedly launched for OpenAI enterprise partners ahead of a wider r...
@shensi: Model governance is coming\n\nQT @steph_palazzolo: New w/ @leomschwartz @amir:

The Trump admin has...

@TFTC21: The Trump administration just asked OpenAI to stagger the release of GPT-5.6 over national security ...
@timshi_ai: the future has arrived, but not evenly distributed\n\nQT @kimmonismus: OpenAI is reportedly releasin...
@MTSlive: SITUATION DETECTED: The Trump administration asked OpenAI to stagger the release of GPT-5.6 over nat...

5. alex_prompter (Group Score: 254.5 | Individual: 33.6)

Cluster: 12 tweets | Engagement: 60 (Avg: 38) | Type: Tech

Your agent should not need a night shift.

Yet half of “AI automation” right now is someone checking logs, restarting runs, and hoping their laptop stays awake.

That is not leverage. That is a new job with worse hours.

@hyperagentapp gives agents their own machine, browser, code execution, and a way to keep running when you close your laptop.

$100 to try it.$ 500 to move your OpenClaw setup.\n\nQT @andrewbusse: If OpenClaw feels more like babysitting, you're not alone.

Skip the daily crashes, leaked secrets, and constant monitoring.

Move to Hyperagent.

For a limited time, get $100 in inference credits at signup and$ 500 when you migrate your first agent.

https://t.co/S9xfEUmIr3 https://t.co/6CBQ3mosjt

See 11 related tweets

@TheoBuildsAI: RT @andrewbusse: If OpenClaw feels more like babysitting, you're not alone.

Skip the daily crashes,...

@vasuman: If you’re hungover from OpenClaw, try migrating to Hyperagent by Airtable.

Run your agents in the ...

@omarsar0: Running agents these days shouldn't be too hard.

But local agents are tricky to operate.

@hyperag...

@alex_prompter: RT @alex_prompter: Your agent should not need a night shift.

Yet half of “AI automation” right now ...

@shiri_shh: People who spent $599 on a Mac Mini to run OpenClaw… now watching HyperAgent give the same thing in ...

6. jxnlco (Group Score: 228.4 | Individual: 48.3)

Cluster: 11 tweets | Engagement: 737 (Avg: 85) | Type: Tech

tldr: https://t.co/7rgT5VhP2L\n\nQT @OpenAINewsroom: Codex usage at OpenAI gives us a preview of what agentic work may look like in the future.

In a new paper, the OpenAI Economic Research team looks at the broader shift from chat to delegation: people using agents not just to get answers, but to hand off longer, more complex work.

https://t.co/CQzNwpEevY

See 10 related tweets

@TFTC21: OpenAI just published internal data on how its own employees use AI agents, and the numbers paint a ...
@cryptopunk7213: looks like openai’s silently taking the lead for long horizon agentic ai tasks

25% of codex request...

@OpenAINewsroom: Codex usage at OpenAI gives us a preview of what agentic work may look like in the future.

In a new...

@gdb: Agents are being adopted very quickly and accelerating work. How this looks across OpenAI itself: ht...
@emollick: This is a fascinating and important set of data which shows us where things are going, using OpenAI ...

7. kimmonismus (Group Score: 208.5 | Individual: 34.3)

Cluster: 7 tweets | Engagement: 116 (Avg: 543) | Type: Tech

Everyone's teaching AI agents to shop. Almost nobody is solving the boring part:

letting them move the money, pay across borders, keep the books, stay compliant.

That's the gap Airwallex is going after, with Airi (faster agent checkout today, a real wallet in the works) and T:0 (an AI-native platform for the finance back office, still in beta).

The interface gets all the attention. The money layer is what decides whether agents can do the work, and Airwallex already owns the rails for it.

It’s great that the important aspects are now also being addressed with the help of AI agents.\n\nQT @awxjack: We’ve raised $320M at an$ 11B valuation, led by Addition.

AI is changing how companies are built. Teams are smaller, global from day one, and using agents more. We’ve spent 10 years building the financial rails for that world.

We’re now building the intelligent layer on top. https://t.co/71AcyjF3KD

See 6 related tweets

@vasuman: Every AI agent demo dies the moment real money gets involved. AI can plan the trip, add the right it...
@itsolelehmann: i keep thinking about how good agentic commerce is gonna feel.

it's just going to oneshot all the s...

@etnshow: BREAKING: Airwallex raises a $320M Series H at$ 11B https://t.co/C1xWspjfGb\n\nQT @awxjack: We’ve ra...
@airwallex: 10 years building the infrastructure that helps businesses operate globally. Now, we're building the...
@exec_sum: Some fintechs are building banking platforms for startups. Some fintechs are building AI agent platf...

8. OfficialINDIAai (Group Score: 206.3 | Individual: 30.9)

Cluster: 8 tweets | Engagement: 0 (Avg: 6) | Type: Tech

As AI becomes an integral part of public service delivery, building trust and accountability is more important than ever.

The panel discussion on "Responsible AI, Governance, Policy & Pathways to Scale", as part of the Workshop on Leveraging AI for Rural Development (Viksit Gram for Viksit Bharat) examined how policy frameworks, governance mechanisms, and institutional readiness can support the safe and effective deployment of AI. The conversation focused on creating an ecosystem where innovation can thrive while ensuring transparency, fairness, and public trust.

#IndiaAI #ResponsibleAI #AIGovernance #DigitalIndia #ViksitBharat

@AshwiniVaishnaw @JitinPrasada @sudeep_tweets @SecretaryMEITY @_DigitalIndia @mygovindia @kansalrohit69 @Alagu19066600

See 7 related tweets

@OfficialINDIAai: Workshop on Leveraging AI for Rural Development (Viksit Gram for Viksit Bharat), organised by the De...
@OfficialINDIAai: Swati Sharma, Joint Secretary, Rural Livelihoods, Ministry of Rural Development, shared her perspect...
@OfficialINDIAai: Swati Sharma, Joint Secretary, Rural Livelihoods, @MoRD_GoI, shared her perspectives on "AI, DPI & ...
@OfficialINDIAai: Building AI for India, and for every Indian.

At the Workshop on Leveraging AI for Rural Development...

@OfficialINDIAai: While speaking at the Workshop on Leveraging AI for Rural Development (Viksit Gram for Viksit Bharat...

9. XFreeze (Group Score: 175.5 | Individual: 49.2)

Cluster: 5 tweets | Engagement: 3526 (Avg: 657) | Type: Tech

Jensen looking at Sam like:

“I gave him my heart, my soul, my CUDA, and $30B worth of GPUs… And he used it to build my competitor behind my back. He basically scammed me” https://t.co/qUBKquyDKS\n\nQT @OpenAI: We’ve designed and built our first AI chip: Jalapeño.

Designed from the ground up by OpenAI and brought to production with @Broadcom, Jalapeño is purpose-built for the LLM workloads powering ChatGPT, Codex, the API, and future agentic products.

Chips are foundational to the AI economy. Building our own expands our full-stack platform from products to models to infrastructure, and will help us scale intelligence, serve more people, and expand access to AI.

See 4 related tweets

@minchoi: It's happening...

Amazon has Trainium. Google has TPUs. Anthropic is exploring custom chips. xAI / ...

@WesRoth: OpenAI unveiled Jalapeño, its first custom AI chip built specifically for running large language mod...
@AskYoshik: I predicted this way before. Every AI company started by training models.

Now they're designing chi...

@rezoundous: can AI finally get cheaper?\n\nQT @OpenAI: We’ve designed and built our first AI chip: Jalapeño.

De...

10. StockMKTNewz (Group Score: 158.1 | Individual: 45.0)

Cluster: 7 tweets | Engagement: 1179 (Avg: 330) | Type: Tech

GOOGLE $GOOGL JUST UPGRADED GOOGLE FINANCE

Three major features launched today:

AI-POWERED PORTFOLIO TRACKING

Drop in screenshots, upload CSVs or PDFs, or just describe your holdings and Google Finance builds your portfolio automatically. Once set up you can ask questions like "what sectors are underrepresented in my portfolio?" or "how does my fixed income allocation impact my long-term growth potential?"

PERSONALIZED MARKET BRIEFINGS ON A SCHEDULE

Describe a task like "send me a daily pre-market briefing analyzing significant overnight moves across major cryptocurrencies" and Google Finance runs it automatically on your schedule. You can tie it to your watchlist or portfolio for personalized insights. Notifications push through the Google app.

NEW GOOGLE FINANCE APP FOR ANDROID

Launched today. Real-time data, live financial news feed, AI research tool, and AI-powered "key moments" that explain why a stock moved. iOS app coming later this year.

See 6 related tweets

@AndrewCurran_: Google just announced that Google Finance is coming out of beta, and they are launching a new financ...
@StockSavvyShay: $GOOGL upgraded Google Finance with AI-powered portfolio tracking that can build portfolios from scr...
@jenzhuscott: I live on Google Business and must say the team has been ramping up a lot of very useful features up...
@Google: RT @thefox: Whenever I post about Google Finance, I get two top questions: (1) when are portfolios c...
@Techmeme: Google launches a Google Finance app for Android, with market data, financial news, and an AI-powere...

11. TeksEdge (Group Score: 157.3 | Individual: 36.0)

Cluster: 5 tweets | Engagement: 56 (Avg: 15) | Type: Tech

🧠Localmaxxers 👀 LLM built for CPU performing at 200+ tps! I have to test this on a Raspberry Pi.

Liquid AI dropped LFM2.5-230M, a 230M model built from the ground up for CPU and edge devices, not GPUs.

⚡ 213 tokens/s on Galaxy S25 Ultra (CPU) ⚡ 42 tokens/s on Raspberry Pi 5 🧠 Strong at instruction following, data extraction & tool use 📦 Tiny memory footprint + runs great on llama.cpp

🎯 These models rock on-device agents for local automation, robots, and privacy-first workflows where you don’t have (or want) a GPU.\n\nQT @liquidai: Introducing LFM2.5-230M: our smallest model yet, built to run fast anywhere (CPUs, NPUs, and GPUs) to enable agentic tasks on phones, robots, home and network automation devices.

230M parameters, built on the LFM2 architecture Pre-trained on 19T tokens, with a 32K context extension Post-trained with distillation from LFM2.5-350M 213 tok/s decode speed on Galaxy S25 Ultra (CPU) 42 tok/s on a Raspberry Pi 5 (CPU) Competes with and often beats models more than twice its size on instruction following, data extraction, and tool use. use it for large-scale data extraction pipelines or lightweight on-device agentic workloads.

🧵

See 4 related tweets

@lmsysorg: 🎉 Meet LFM2.5-230M from @liquidai, their smallest model yet at 230M params, but it punches way above...
@vllm_project: 🚀 Day-0 support for LFM2.5-230M on vLLM!

Ready to serve @liquidai's smallest LFM2 model for fast G...

@BrianRoemmele: Wow!

Liquid AI just released LFM2.5-230M, their smallest and most efficient model yet. This 230 mil...

@liquidai: RT @TeksEdge: 🧠Localmaxxers 👀 LLM built for CPU performing at 200+ tps! I have to test this on a Ra...

12. inafried (Group Score: 152.2 | Individual: 55.0)

Cluster: 5 tweets | Engagement: 43384 (Avg: 1109) | Type: Tech

RT @Pontifex: We cannot consider #AI to be morally neutral. In reality, every technical tool embodies choices and priorities through what it measures, ignores, and optimizes, and how it classifies people and situations. Ethical discernment cannot be limited to asking whether we are using a system for good or bad purposes. It must also examine how that system is designed and what vision of the human person and society is embedded in the data and models that guide it. #MagnificaHumanitas

See 4 related tweets

@sriramk: The Pope captures something that every tech company that has tried to determine “what is right” - wh...
@Grady_Booch: Every line of code represents an ethical or moral decision.\n\nQT @Pontifex: We cannot consider #AI ...
@ChristopherHale: Pope Leo XIV is back on X with more AI bangers.\n\nQT @Pontifex: We cannot consider #AI to be morall...
@Kyrannio: Interesting\n\nQT @Pontifex: For AI to respect human dignity and truly serve the common good, respon...

13. moneycontrolcom (Group Score: 150.3 | Individual: 19.4)

Cluster: 10 tweets | Engagement: 1 (Avg: 8) | Type: Tech

#AIWithMC | Anthropic accuses its Chinese AI rival Alibaba of using fake accounts to 'steal' Claude AI capabilities

Anthropic has accused Alibaba of using thousands of fake accounts to extract Claude AI capabilities, alleging a large-scale model distillation campaign that violated its terms and intellectual property rights.

@shauryashubham writes: https://t.co/oaHPE5DAeY

See 9 related tweets

@ReutersBiz: WATCH: Anthropic has accused Alibaba, the Chinese technology and e-commerce giant, of illicitly extr...
@FirstSquawk: ANTHROPIC ACCUSES ALIBABA OF USING 25,000 FRAUDULENT ACCOUNTS TO PROBE CLAUDE AI MODELS

Anthropic⁠ ...

@Reuters: Anthropic has accused Alibaba, the Chinese technology and e-commerce giant, of illicitly extracting ...
@Reuters: Anthropic says Alibaba illicitly extracted Claude AI model capabilities https://t.co/aSJpadcBXI http...
@Reuters: US AI company Anthropic accused Alibaba, the Chinese technology and e-commerce giant, of illicitly e...

14. wallstengine (Group Score: 150.3 | Individual: 19.2)

Cluster: 9 tweets | Engagement: 168 (Avg: 132) | Type: Tech

Google is reorganizing its AI coding strike team to better compete with Anthropic. The team is being expanded into a formal “midtraining” group focused on improving model capabilities in coding and other Biz tasks, after recent exits of Noam to OpenAI & John Jumper to Anthropic. https://t.co/tR8rOgmldp

See 8 related tweets

@financialjuice: Google reorganizing new AI coding tool strike team - The Information $GOOGL https://t.co/BQ5wf4M6fE...
@WesRoth: Anthropic is reportedly hiring two more leading Google AI researchers: Jonas Adler and Alexander Pri...
@FirstSquawk: GOOGLE REVAMPS AI CODING STRIKE TEAM TO CATCH ANTHROPIC IN LUCRATIVE MARKET- THE INFORMATION...
@Techmeme: Sources: Google expands the scope of its months-old AI coding strike team to "midtraining" to better...
@StockSavvyShay: $GOOGL is reorganizing its AI coding team as it tries to catch Anthropic in one of AI’s most lucrati...

15. Rasmic (Group Score: 147.0 | Individual: 35.5)

Cluster: 5 tweets | Engagement: 105 (Avg: 98) | Type: Tech

200M seed round, chat we're so back\n\nQT @a16z: We're thrilled to lead Mirendil's $200M seed round.

Frontier AI work has been locked inside a few big labs.

@mirendil is building a system that can help anyone do AI work: they train frontier models that are expert at AI R&D and build the product around it. The result is a system that loops over research and engineering problems on its own, making progress without human intervention.

It's like a coding agent built for AI research that controls its own GPUs.

Mirendil is one of the few teams with the experience and the priors to make the end-to-end system work. @bneyshabur, @HarshMeh1a, @shayan_, and @tararezaeikh come out of Anthropic, Google DeepMind, and xAI.

Welcome to the a16z family, Mirendil.

By @BornsteinMatt and @MaikaThoughts

See 4 related tweets

@SanthProject: What exactly was original about this?\n\nQT @a16z: We're thrilled to lead Mirendil's $200M seed roun...
@HessianFree: RT @bneyshabur: Today, I’m excited to formally announce @mirendil with my amazing co-founders Harsh ...
@chiefofautism: RT @a16z: We're thrilled to lead Mirendil's $200M seed round.

Frontier AI work has been locked insi...

@bneyshabur: RT @MirendilAI: We are proud to formally announce Mirendil. Our mission is to accelerate science and...

16. Konstantine (Group Score: 142.1 | Individual: 34.3)

Cluster: 5 tweets | Engagement: 32 (Avg: 48) | Type: Tech

Neil Movva @neilmovva and Samir Menon @blintzbase are two of the most brilliant young AI infrastructure architects in the valley.

With big company experience at Apple, NVIDIA, and Intel, high-growth experience at Together AI, and repeat-founder experience at Blyss, they were a triple threat. Not only were they deep engineers, but also charismatic, visionary, and extremely driven.

So when they laid out a vision for a new infrastructure platform built specifically for agents working in the background, Lauren Reeder @laurenmhreeder and I at @sequoia enthusiastically led their Seed.

Agents work autonomously for hours. They read entire codebases. They screen hundreds of candidates. They run thousands of simultaneous tasks. Sail Research is custom built for AI Agents.

Sail has assembled a dream team and are already processing trillions of tokens per week on behalf of their customers.

Today, Sail is announcing their $80M Seed & Series A, with us, Kleiner Perkins, Redpoint, Theory Ventures, Vine Ventures, and CRV.

Neil and Samir: the best is yet to come.

Special thanks to the extraordinary Jason Ginsberg @JasonBud who introduced me to Neil and the company.

See 4 related tweets

@simonguozirui: RT @neilmovva: Samir Menon @blintzbase and I are thrilled to announce Sail @sailresearchco ! We buil...
@Redpoint: Congratulations to @neilmovva, Samir Menon, and the @sailresearchco team on their $80M Seed & Series...
@Redpoint: RT @reneeshah123: Sail Research (@sailresearchco) is building an inference platform for long-running...
@theinformation: Today on TITV:

-Google takes on Anthropic with revamped AI team | @jasonrdean

-AI demand fuels hug...

17. Dagnum_PI (Group Score: 134.5 | Individual: 29.9)

Cluster: 7 tweets | Engagement: 18 (Avg: 272) | Type: Tech

The World Economic Forum estimates AI will create 170M jobs and displace 92M by 2030. Net gain: 78 million.

$AIAI CEO Todd Furniss on why the companies that win won't be the ones cutting headcount. They'll be the ones using AI to grow revenue and build new products.

This is exactly what the Ai² acquisition model does.

Buy companies, deploy AI across the portfolio, make people more productive. Not fewer people. More output.\n\nQT @_AiSquared: CEO of AIAI Holdings, Todd Furniss discussed on https://t.co/lPxyTYVT4m why AI will drive employee growth rather than increase job losses. Read full article here:

https://t.co/5vg3gOO2TA #AI #AIAIHoldings #Artificialintelligence https://t.co/LQtl9FGdoj

See 6 related tweets

@Jameswise: Really excellent detail in this on the AI economy from @azeem and team.
AI revenues are growing a...
@xeophon: Azeem and the @ExponentialView team did an insane deep dive about the ai economy and made it freely ...
@business: Revenue from AI has reached a tipping point, showing that the hundreds of billions of dollars tech c...
@MoorInsStrat: 📈 Companies are spending more on AI than ever—but many still haven’t modernized how they operate.

V...

@GoI_MeitY: Speaking after the bilateral meeting in Washington, DC, Secretary, MeitY, Shri S. Krishnan highlight...

18. ClementDelangue (Group Score: 128.7 | Individual: 66.0)

Cluster: 3 tweets | Engagement: 2371 (Avg: 156) | Type: Tech

We just crossed $100M annual run-rate. I know many AI companies are capturing much more $$$ these days, but still proud of the milestone!

Maximizing short-term revenue has never been our priority. In fact, we're proud to manage to store and serve hundreds of petabytes of models and datasets while keeping HF free and open-source for 97% of our users. As a platform, we’re happy to hopefully create orders of magnitude more value for the community than what we capture. To me, that’s the very definition of a platform.

And it has helped us build one of the most loved platform in tech, with network effects, a defensible position and a sustainable business which is quite unique in AI.

Many many thanks to all the community members for building with us, we wouldn't be anywhere without you! Can’t wait for what’s next, especially as more companies start to see the value of open and local AI! Next milestone $1B?

See 2 related tweets

@Bouazizalex: Let’s to @ClementDelangue @huggingface !!!

@dani_avila7: Massive milestone for opensource AI 👏

Congrats to the @huggingface team

The whole ecosystem wins w...

19. levie (Group Score: 128.2 | Individual: 35.4)

Cluster: 4 tweets | Engagement: 482 (Avg: 572) | Type: Tech

There are some subtleties in this launch that are very important in practice. This isn’t just you interacting with Claude in a 1:1 format via Slack. In this case, Claude acts as a coworker that any user can tap into in a shared way.

We’ve already seen some agentic coding systems start to adopt this pattern (as well as OpenClaw and Hermes), and doing it for general purpose knowledge work continues to push the idea forward.

As a result, what this means is that this agentic coworker needs its own set of resources, access to tools, and data to work with. This is not the same as you giving it access to your personal resources and tools, because the agent then could accidentally then share those out with anyone.

The agent needs to instead be like any other user in the system, and you need to be thoughtful about what it should have access to, and make sure its information that is safe to share with that group.

When you can pull that off, it’s quite powerful. For instance, by connecting to Claude Tag to Box, you could have Claude access corporate sales materials for questions in sales conversations or generating RFPs, brand guidelines and marketing assets for campaign creation, product roadmap materials and product documentation for coding agents to use, contracts that anyone in the legal team can access, and more.

But this is just the Box example. You can equally have it access your product or customer analytics data, CRM information, codebase, and other resources and agents that would make sense to work on in a collective manner.

It’s awesome to see continued innovation in what the future of work may begin to look like with agents.\n\nQT @karpathy: This is a new paradigm for interacting with Claude that is significantly more "inline" with all the other human activity org-wide. Once you do all of the under the hood engineering work to make this "just work" (e.g. across tools, integrations, compute environments, memory, security, etc.), Claude basically joins the team in a seamless way - you can talk to it as you would talk to a person and it can help with a very large variety of workloads.

Imo this is the 3rd major redesign of LLM UIUX. The first paradigm was that the LLM is a website you go to, the second was that it is an app you download to your computer. This third one is that it is a self-contained, persistent, asynchronous entity with org-wide tools and context, working alongside teams of humans. It really takes a while to wrap your head around it, but it works and it is awesome.

See 3 related tweets

@tugot17: "You either die a hero, or live long enough to see yourself doing B2B sales"\n\nQT @karpathy: This ...
@guohao_li: It feels like Claude Tag is much more than just a Slack bot feature. Under the hood, it points to a ...
@noahzweben: RT @levie: There are some subtleties in this launch that are very important in practice. This isn’t ...

20. zerohedge (Group Score: 126.7 | Individual: 25.7)

Cluster: 7 tweets | Engagement: 2179 (Avg: 774) | Type: Tech

Apple's price increases: MacBook Air rose $200 to$ 1,299 Base MacBook Pro increased $300 to$ 1,999 Entry-level MacBook Neo increased $100 to$ 699 iPad Air increased $150 to$ 749 iPad Pro increased $200 to$ 1,199 iPhone prices unchanged for now, though the company hinted at more increases in a statement.\n\nQT @zerohedge: APPLE SAYS AI DATA CENTER BOOM CAUSED HAS UNPRECEDENTED DEMAND FOR MEMORY AND STORAGE, FORCING PRICE INCREASES

See 6 related tweets

@FirstSquawk: APPLE SAYS AI DATA CENTER BOOM HAS CAUSED UNPRECEDENTED DEMAND FOR MEMORY AND STORAGE, FORCING PRICE...
@markgurman: Also, the Apple TV and HomePod mini haven’t been updated in half a decade. We’re not talking about c...
@zerohedge: APPLE SAYS AI DATA CENTER BOOM CAUSED HAS UNPRECEDENTED DEMAND FOR MEMORY AND STORAGE, FORCING PRICE...
@imranye: im hoping a side effect of this, is that we will see a new class of more efficient software

i think...

@Techmeme: Apple says "the rapid expansion of AI data centers has created an extraordinary surge in demand for ...