顶尖科技推文 - 2026年06月25日

2026年6月25日科技每日简报

Today's top tech conversations are led by @Dimillian, whose post about 'RT @OpenAI: We’ve designed and...' garnered the highest engagement. Key themes trending across the top stories include design, https, environments, built, building. The community is actively discussing recent developments in AI, engineering practices, and startup strategies.

1. Dimillian (Group Score: 1430.3 | Individual: 59.5)

Cluster: 48 tweets | Engagement: 3238 (Avg: 191) | Type: Tech

RT @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 47 related tweets

@kimmonismus: OpenAI just unveiled Jalapeño, its first custom AI chip designed from scratch for LLM inference-

It...

@kimmonismus: Absolutely insane:

"Jalapeño was co-developed from initial design to manufacturing tape-out in jus...

@testingcatalog: OpenAI 🤝 Broadcom

OpenAI announced its first AI chip, designed and produced in a partnership with B...

@StockMKTNewz: OPENAI AND BROADCOM $AVGO JUST UNVEILED OPENAI'S FIRST CUSTOM AI CHIP

It is called Jalapeño. It is ...

@mark_k: Today @OpenAI announced Jalapeño, its first custom Intelligence Processor developed with Broadcom fo...

2. brunoborges (Group Score: 474.4 | Individual: 57.6)

Cluster: 13 tweets | Engagement: 1650 (Avg: 80) | Type: Tech

RT @JPoehnelt: Two months ago I was fired by Google for creating the Google Workspace CLI. It went viral, hit #1 on Hacker News, gained thousands of GitHub stars and many thousands of actual users in just a couple days.

It was an incredible, confusing journey, from directors and leaders asking what they could learn from the tool to getting grilled by legal about why the Google logo and brand colors are on the Google Workspace GitHub code repositories.

I think the cause was that Workspace and certain leaders (and projects) were afraid of being disrupted. But the fear wasn't specific to my CLI, it was a broader fear in what agents meant for Workspace. Either way, the irony of my termination was the announcement at Google Cloud Next two days before I was fired that an official Workspace CLI was coming.

I want this out there because it is easier for me to explain my story and it is an experience I want to fully own. It's also part of my healing.

Nearly 7 years at Google was an incredible opportunity for me and I was fortunate to have wonderful teammates and a manager that fully supported me through these last few months. Thank you.

See 12 related tweets

@steipete: Google fired the guy that made the google workspace cli, because he made the google workspace cli.

...

@kunchenguid: the person who created google workspace CLI got fired by google

having been at big tech for decades...

@levelsio: Insane

@sundarpichai you should rehire and then promote him, these are the people Google needs!\n\n...

@PaulSolt: RT @steipete: Google fired the guy that made the google workspace cli, because he made the google wo...
@clairevo: Something I think a lot of bigco takes on this are missing is the nuance of devrel and devtools. The...

3. Scobleizer (Group Score: 393.1 | Individual: 48.1)

Cluster: 16 tweets | Engagement: 4174 (Avg: 187) | Type: Tech

RT @berman66: Today, we’re announcing Runlayer has raised $30M from Felicis and Khosla Ventures to help companies go all in on AI.

Runlayer is the golden path for AI: enablement, security, and control in one platform.

So, how does it give your team the right tools for AI? 🧵

https://t.co/nMjWuv3Oc3

See 15 related tweets

@Scobleizer: Proud to have @runlayer supporting my work watching and serving the AI industry at https://t.co/8L5x...
@Parul_Gautam7: Feels like Runlayer is building the missing layer that finally lets AI agents do real work. https://...
@wallstengine: Runlayer raised a $30M Series A led by Felicis.

They’re building a secure layer for employees to co...

@heyshrutimishra: Your employees are building AI agents you don't know about.

They're connecting them to production ...

@zaynmcps: sales reps right now watching an AI agent close the deal they forgot to follow up on for 3 weeks htt...

4. ivanfioravanti (Group Score: 280.6 | Individual: 50.5)

Cluster: 10 tweets | Engagement: 826 (Avg: 82) | Type: Tech

RT @Alibaba_Qwen: 📣📣 Meet Qwen-AgentWorld — a native language world model that simulates 7 agent environments (MCP, Search, Terminal, SWE, Web, OS, Android) within a single model. Environment modeling is the training objective from day one, not a post-hoc adaptation.

🤔 LLMs are trained to be better agents — better at acting in environments. But nobody has trained them to model the environments themselves.

🗺️ Our roadmap: investigate how language world modeling can push the boundaries of general agent capabilities, along two routes:

1️⃣ Build a foundation model for environment simulation — outperforming Claude Opus 4.8 and GPT-5.4 on AgentWorldBench

2️⃣ Investigate how world modeling enhances agent training: 🔬 Controllable Sim RL (agentic RL with LWM as environments) surpasses training in real environments 🧠 Learning to predict environments (LWM warm-up) makes agents stronger — remarkably, even without any agent-specific training, this predictive knowledge transfers to agentic tasks with zero fine-tuning

📑 Paper: https://t.co/Jx2l5RKq71 📖 Blog: https://t.co/7tVcKyhsx2 💻 GitHub: https://t.co/B5Lvb1UZCn 🤗 HuggingFace: https://t.co/Kw3QBL1TM5 🧩 ModelScope: https://t.co/YBnGYgMWWI

See 9 related tweets

@BrianRoemmele: This is BIGGER than the DeepSeek Moment and makes the story of “regulate AI” point stupid.

The toot...

@BrianRoemmele: While Dario of IBM tries to get the government to shut down any and all open source AI because it sc...
@Ali_TongyiLab: A new approach to general agent capabilities: Qwen-AgentWorld. 🤖

While LLMs are typically trained t...

@BrianRoemmele: 20 Blackwell class AIs networked running Qwen-Agent will run circles around Mythos.

There are more ...

@VaibhavSisinty: Alibaba just open sourced an AI agent model. And it is beating every closed model on the market 🤯

I...

5. shiri_shh (Group Score: 215.2 | Individual: 30.8)

Cluster: 8 tweets | Engagement: 39 (Avg: 638) | Type: Tech

You can now run a biotech startup from a coffee shop.

Tetsuwan is building AWS but for BIOLOGY labs.

Instead of buying millions in lab equipment, you write your experiment in code and a robotic cloud lab executes it perfectly.

Flawless data, delivered straight to your laptop without ever stepping inside a physical building.\n\nQT @c_m_ponce: Automating the lab bench is the best thing we can do for AI in biology.

Most experiments are still run by hand. Every biologist's handiwork is unique and every lab is a little different. So, biology faces widespread reproducibility issues.

But, AI will demand more reproducible data than we can produce, and generate more ideas than we can test.

Experiments must be communicated and executed in a standardized way to generate reproducible data.

So, Tetsuwan is building a lab where users specify experiments in an exact syntax. These experiments are executed by an automated platform to generate transparent, reproducible output. The user never needs physical access to a lab. This is a biology lab you can use like a computer.

Our platform, built and tested with pilot labs over the past two years, lets users configure automated workflows rapidly & precisely.

Later this year, we will bring our first services online, focusing on functional screens for protein design.

Alex and I met at Caltech, where she was one year my senior. We're building this company because we were little kids who wanted to be biologists that grew up into adults who resented the lab bench. We want biology to be about asking questions, not the painful and frustrating manual process of asking them.

For those looking to receive updates on our pilot services, looking for a meaningful job, or just to learn more about our work, see the links in the comments!

See 7 related tweets

@zaynmcps: biology has a reproducibility problem nobody talks about

every experiment run by hand. every lab sl...

@Mayhem4Markets: Lab automation hardware is a commodity. You can buy robot arms and plate readers off the shelf. Conf...
@wallstengine: Tetsuwan is launching its automated biology cloud lab platform, built to let researchers submit expe...
@rohanpaul_ai: This is so cool. Tetsuwan is building a browser-based cloud biology lab.

Because, AI can now genera...

@VadimStrizheus: 🚨 Do you understand what just happened to science?

these guys just launched the AWS for biology, an...

6. PatrickMoorhead (Group Score: 214.7 | Individual: 36.1)

Cluster: 9 tweets | Engagement: 53 (Avg: 47) | Type: Tech

I’m interested to see what $QCOM does with Modular. Qualcomm is very good at edge enabling software, but that’s not the same as datacenter software capability. Strategically, this could help to better answer the datacenter question. I’m wondering if this is more of an aquihire or Modular services will continue. I’m attending the financial analyst today to learn more.\n\nQT @Modular: We’re excited to announce that Modular has entered an agreement to be acquired by @Qualcomm. The future of unified compute has never been stronger. Read the full announcement: https://t.co/FiQUL5CvNj

See 8 related tweets

@danielnewmanUV: $QCOM is buying Modular. An AI native software layer to deepen its tools and strengthen its datacent...
@PatrickMoorhead: $QCOM CEO @cristianoamon kicks off its 2026 investor day:

• Opened by framing this as a new chapte...

@wallstengine: QUALCOMM LAYS OUT DATA CENTER PUSH WITH $META AND MICROSOFT

$QCOM said META plans to use its Dragon...

@KristinaParts: BREAKING: Qualcomm to acquire AI software startup Modular. Deal terms not disclosed. Modular's softw...
@StockSavvyShay: $QCOM has acquired Modular for ~$ 3.9B in stock to add a hardware-agnostic AI software and compiler l...

7. rorypreddy (Group Score: 211.4 | Individual: 57.3)

Cluster: 6 tweets | Engagement: 2506 (Avg: 172) | Type: Tech

RT @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 5 related tweets

@irl_danB: > the legend tweets something 6-24 months after it is conventional wisdom and well-trod in certain c...
@dani_avila7: Karpathy named 3 paradigms for LLM UX… but I think there’s a fourth

when agents stop receiving task...

@jdkanani: In Morphic slack, we have two agents: Binod and Hamza.

We were using new paradigm before it became ...

@rickasaurus: I see very little benefit to having them on chat vs in a nice UI. In many ways it’s worse.\n\nQT @ka...
@anvisha: Agreed, but it’s not that new. We built a bespoke version of this for our internal use a couple mont...

8. Parul_Gautam7 (Group Score: 207.9 | Individual: 31.0)

Cluster: 7 tweets | Engagement: 29 (Avg: 77) | Type: Tech

The most valuable AI products aren't adding more features.

They're removing handoffs.

Genspark Design bringing design, prototyping, animation, and code into one place feels like a step in that direction.\n\nQT @genspark_ai: 🎨 Introducing Genspark Design

The next-gen AI for design and creation, powered by Claude Opus 4.7.

From rough idea to professional design, no design background needed.

One tool for everything you design. UI prototypes, videos, HTML animations, posters — all in one place and production-ready.
Stay on-brand with your own design system. Upload your Figma files or save your designs, reuse them across every project, and share them with your team.
Ship it, don't just mock it. Turn any app or website design into working code with one click, built on Genspark Code.

Start building at our launch price 👉 https://t.co/Fn9YJHgRx8

See 6 related tweets

@rohanpaul_ai: Genspark just launched Genspark Design.

Turns rough prompts into editable UI, posters, videos, HTML...

@VadimStrizheus: Designers watching Genspark replace their job with AI.

vibe-design is finally here.👇 https://t.co/...

@heyrobinai: every AI design tool i've tried makes a beautiful mockup and then leaves you stranded. great, now re...
@AiBreakfast: Genspark is going to swallow Adobe.

Mark my words.\n\nQT @genspark_ai: 🎨 Introducing Genspark Desig...

@dani_avila7: Use the right tool for the job... for design specifically, Genspark is strong

What stands out to me...

9. StockMKTNewz (Group Score: 179.7 | Individual: 48.1)

Cluster: 6 tweets | Engagement: 1430 (Avg: 321) | Type: Tech

SK Hynix stock is about to begin trading on the US exchanges on July 10th\n\nQT @IPONewsroom_: SK HYNIX IS GOING PUBLIC IN THE 🇺🇸 IN JUST 3 WEEKS

SK Hynix is planning a $29.4 Billion US listing with trading set to begin on July 10th

This would be one of the top five share sales in history, comparable in size to Saudi Aramco's record 2019 IPO.

SK Hynix controls 57% of the global high-bandwidth memory market by revenue. HBM is the chip that sits inside every NVIDIA GPU powering the AI buildout. Without it, the data centers do not work.

The numbers behind the business:

Q1 2026 operating profit: $24.3 billion, a record high Q1 2026 sales:$ 34 billion, nearly tripling year over year Stock performance: +300% in Seoul this year Market cap: over $1 trillion

The proceeds will fund additional capacity expansion and purchases of EUV lithography machines needed to make next-generation chips.

Lead underwriters: Bank of America, Citigroup, Goldman Sachs, JPMorgan.

See 5 related tweets

@IPONewsroom_: SK HYNIX IS GOING PUBLIC IN THE 🇺🇸 IN JUST 3 WEEKS

SK Hynix is planning a $29.4 Billion US listing ...

@WOLF_Financial: South Korean memory chip giant SK Hynix is about to list its stock on US exchanges making it much ea...
@shanaka86: SK Hynix, the Korean firm that makes the memory all AI runs on, just filed to raise $29.4B on Nasdaq...
@moneycontrolcom: #Business | SK Hynix seeks $29 billion with new US listing to fund AI boom
SK Hynix plans $29.4 b...
@Reuters: South Korea's SK Hynix said it plans to raise up to $29.43 billion via the listing of American Depos...

10. heyshrutimishra (Group Score: 173.6 | Individual: 34.2)

Cluster: 6 tweets | Engagement: 12 (Avg: 258) | Type: Tech

Every Big Tech company is racing to build AI assistants. A startup called Nori just showed the first one that actually runs a household.

SuperNori auto-restocks groceries, books travel, plans weekly meals around your family's nutrition goals, and controls smart home devices. All without being asked.

The real moat in AI isn't the model. It's who owns the daily routine.\n\nQT @IsaacDrgn: Most AI helps you write, design, code, and ship faster at work. Nothing was built for the person quietly holding the family together.

Introducing SuperNori: the first Proactive Family AI Agent built for the family caretaker in every family.

Here's how it works: https://t.co/mxhtDiULIA

See 5 related tweets

@heyrobinai: i had a calendar, a list, a meal planner, and a dead notes doc. four apps, none of them talking to e...
@Gorden_Sun: Nori launched SuperNori, a proactive AI agent designed for families.

Traditional AI only responds w...

@AiBreakfast: New consumer AI category:

Not an assistant. Not a copilot.

SuperNori = Proactive Family AI The Fam...

@teslaownersSV: Tesla redefined the car.

SuperNori is doing the same for the home.

AI built for the family 🖤\n\nQT...

@dr_cintas: RT @IsaacDrgn: Most AI helps you write, design, code, and ship faster at work. Nothing was built for...

11. chandrarsrikant (Group Score: 169.4 | Individual: 38.7)

Cluster: 6 tweets | Engagement: 240 (Avg: 253) | Type: Tech

Shah was approached directly by Meta’s chief product officer, Chris Cox, while searching for a leader "with an intuitive grasp of the immense, global product potential of WhatsApp, who could navigate the shifts AI will bring, and who has the seriousness to lead the world’s largest communication service.”

“Kunal became the clear choice as we got to know him,” Cox said in an internal note announcing the transition earlier this week.\n\nQT @chandrarsrikant: 🚨$900 million for Kunal Shah: Meta brings Silicon Valley's ‘hackquisition' playbook to India with WhatsApp-Cred deal

Meta's $900 million deal to bring Cred founder Kunal Shah into its fold as WhatsApp's new global leader has all the hallmarks of a growing trend in the technology industry: the "hackquisition", a strategy in which companies use investments and other strategic deals to recruit entrepreneurial talent.

The term, a portmanteau of hack and acquisition, refers to a strategy whereby large tech companies poach a startup’s key talent through a range of deals, including licensing deals, rather than outright buyouts.

What's unique is that Shah's hire is not directly tied to AI, although the technology is likely to play a significant role in WhatsApp's evolution as Meta pursues its ambition of delivering "personal superintelligence" to everyone.

Shah also differs from many Indian-origin leaders in global technology companies who rose through the ranks of Silicon Valley companies. Instead, he built and scaled two large-scale businesses in India.

He is also among the country's most prolific angel investors, having backed more than 250 startups, including several unicorns, and actively mentors hundreds of founders.

Meta CEO Mark Zuckerberg said Shah "brings the kind of builder mentality and global perspective that will serve him well in running the world's biggest messaging app."

“Kunal became the clear choice as we got to know him,” Cox said in an internal note announcing the transition.

To be sure, Meta is no stranger to such deals. In June 2025, the company invested $14.3 billion in data-labeling startup Scale AI for a 49 percent stake and recruited its CEO, Alexandr Wang, to spearhead the social networking giant's AI efforts, including its superintelligence labs.

Meta also subsequently acquired a 49 percent stake in NFDG, the $1.1 billion AI venture fund run by Nat Friedman and Daniel Gross, through a tender offer last year and hired both of them to join its leadership team.

In India, Meta tapped Ezetap co-founder Abhijit Bose to lead WhatsApp India in November 2018, marking the first time it had appointed a country head in any market. However, the move was not accompanied by any investment.

Good one by @tsuvik

https://t.co/sfeZugDk3z

See 5 related tweets

@chandrarsrikant: 🚨$900 million for Kunal Shah: Meta brings Silicon Valley's ‘hackquisition' playbook to India with Wh...
@chandrarsrikant: 🚨 🆕 How WhatsApp's billion-dollar India business influenced Meta's bet on Kunal Shah

India now cont...

@moneycontrolcom: $900 million for Kunal Shah: Meta brings Silicon Valley’s ‘hackquisition’ playbook to India with Wha...
@moneycontrolcom: #TechWithMC | How WhatsApp's billion-dollar India business influenced Meta's bet on Kunal Shah
Wh...
@Goenka_Tushar1: RT @chandrarsrikant: Shah was approached directly by Meta’s chief product officer, Chris Cox, while ...

12. firstadopter (Group Score: 162.1 | Individual: 52.9)

Cluster: 4 tweets | Engagement: 449 (Avg: 76) | Type: Tech

HOLY CRAP $MU

Micron: “Fourteen of the 16 SCAs that we have signed have a cumulative revenue at minimum price per our contracts of approximately $100 billion over the remaining agreement term”

“We are excited to announce that we have now signed 16 strategic customer agreements, or SCAs, which we expect will fundamentally transform our business model.”

“We are pleased to announce that we have completed 16 SCAs with customers across the data center, consumer and auto market segments. These SCAs accelerate the transformation of our business model, enhance partnership in technology and innovation, and provide customers with contracted supply assurance.”

“Typically, these agreements have a five-year term, from calendar 2026 through the end of calendar 2030. Automotive agreements generally have a three-year term. The 16 signed agreements represent roughly 20% of our DRAM volume and a third of our NAND volume over this period. These SCAs include four very large customers and three medium-sized customers. The remaining agreements relate to smaller customers from the automotive industry and represent our commitment to this important sector.”

“When completed, we expect approximately half or more of our company revenue to be under these SCAs with customers across end markets. Our customers value our U.S. supply plans, and this is reflected in our SCAs.”

“These SCAs are structured as take-or-pay agreements, with binding commitments to purchase specific volumes over this multi-year term.”

See 3 related tweets

@firstadopter: Micron on LT agreements: floor price is $100B. they expect "MUCH HIGHER"

$100 billion is floor pric...

@firstadopter: Micron: "We now expect calendar 2026 industry server units to grow high-teens percent, above our pri...
@firstadopter: Back to highs we go\n\nQT @firstadopter: HOLY CRAP $MU

Micron: “Fourteen of the 16 SCAs that we hav...

13. TheoBuildsAI (Group Score: 155.1 | Individual: 45.8)

Cluster: 4 tweets | Engagement: 1262 (Avg: 290) | Type: Tech

RT @DJ_CURFEW: The 100x org went viral. Half the internet hated it. The other half was curious. One month later: output is up. productivity is spiking. we're approaching a 5:1 agent-to-human ratio. And contrary to popular belief, we're doing the OPPOSITE of tokenmaxxing. We're tokenSAVING.... Why 5,000 agents for 1,000 people? Because agents AREN'T workflows... Agents define permissions and skills. the vast majority of ours don't bother a single human, they run as triggers and loops in the background. My rule is simple: if you're doing work that ISN'T purely using your judgment → build an agent or a system to remove that bottleneck. So that's what we did. And yes, it broke things. Honestly it created real pain for first couple weeks. To be expected. we have TONS of examples now people who had NEVER built a system in their life and are now running large-scale agentic systems that power much of our company. I always say - if you automate your own job with AI, you'll ALWAYS have a job. Systems is arguably the most important thing of all. Today, regardless of what your role used to be, everyone does the IC work end-to-end: → Before, you riffed on messaging. Now you ship the landing page yourself. → Before, you were a PM waiting on design and research. Now you're a builder iterating in real-time. → Before, you waited on someone to finish something. Now you don't. The 100x future isn't coming. It's HERE. Bottlenecks are mostly gone and those that aren't are in sight. And obviously gotta lead by example. I even designed and built the Brain² landing page. I built most of the product. The in-app announcements. The emails. All of it.

Breaking our existing processes let us rethink everything WITHOUT prior constraints. Enter, Process Miner: my agent that mines all the processes in our company. This is just one of the infinite use cases for setting up an LLM to process EVERY SINGLE event that happens in your company. These types of agents simply are not sustainable without compressed context. Instead of starting from scratch each time and deep researching across the ~100,000 daily activity items in our company, it starts from context that has already been processed, summarized, and organized. Everyone's jumping on the context and memory train now. I love it. We've been building this for two years. and we STARTED OVER 4 times! We came out the other side with what we call live intelligence: → LLM pipeline: every event runs through a cheap LLM that summarizes, organizes, triggers, and rolls up

→ Self-organizing memory. Don't overcomplicate it. Start simple.

→ Self-improving orchestration. THIS IS THE KEY. BUT... don't over-focus on orchestrating models... it's WAY more about orchestrating CONTEXT. It wasn't more tokens. It wasn't a better model. It was CONTEXT. Every LLM is average by default. Think about that. If you're using the same model as everyone else, your intelligence is by definition, AVERAGE. That's what really changed everything for us internally and it's why we even have a product to release in the first place. Our company AI is personalized (and yes these learnings are applicable regardless what system you use).

Every time you use Brain², we dynamically rewrite the system prompt based on real-time context and intent. Your activity, your decisions, your projects, your memory, your preferences. And your FEEDBACK.. that's WAY more important than decisions that everyone goes overboard on being so valuable. All injected AHEAD of time. When AI has 100% context, you don't need prompts in the same way. You need INTENT. I go anywhere in ClickUp and type "@brain slides." Done. "@brain campaign." Shipped. "@brain app." Live. And it REMEMBERS. Every interaction compounds. One Brain for your whole company, and every person makes it smarter for everyone else. I flew our AI team to my house for a week to finish this. Monitors and whiteboards everywhere. SO proud of this 100x team. Brain² is truly the best work AI in the world... and I firmly believe it's also the best personal AI. Naturally, easy for me to say and honestly I assumed it was just my bias originally. but then EVERYONE who tried it started saying the same thing. So we tested it. A three-week study, randomly recruited participants, blind pairwise comparisons. They picked Brain² as the best nearly 100% of the time over ChatGPT, Gemini, and Claude. Here's the kicker: you can run ANY of those models inside Brain². Same frontier models. Radically different results.

Brain² doesn't win because the model is better. It wins because of CONTEXT. It learns from every single interaction and new piece of context it sees. OH and we built our harness around being OBJECTIVE. Soo it can be sassy - you will definitely see some 'fun' answers in your day to day.

AI that is built to agree with you is NOT productive. That makes no one better. Run it against whatever you're using right now. It compounds in quality the more that you use it.

Prove me wrong → https://t.co/k9J7r9VuxS

Brain² is your company's AI.

I built this for us originally to realize 100x, and now we're shipping it externally. Even if you're building your own I just hope sharing in public helps in some small way. 💜 Let's make the world more productive

See 3 related tweets

@zaynmcps: cada LLM es promedio por defecto

estás usando el mismo modelo que todos los demás, tu inteligencia ...

@FellMentKE: The “one brain for your whole company” thing actually makes so much sense when you think about it.

...

@TheoBuildsAI: Wish we had that culture Thats a BIG statement to say it is better than all other AI’s. Gotta try th...

14. wallstengine (Group Score: 153.4 | Individual: 39.3)

Cluster: 5 tweets | Engagement: 294 (Avg: 132) | Type: Tech

Humanoid robotics company Agility Robotics is going public through a merger with Churchill Capital Corp XI $CCXI.

The deal values Agility at $2.5B pre-money and is expected to generate$ 620M+ in gross proceeds, including a ~$200M PIPE led by Foxconn.

The combined company is expected to trade under ticker $AGLT.

Agility says it has $300M+ of multi-year orders for its next-gen Digit v5 robot from a 30+ customer pipeline.

The company has logged 65,000+ operating hours across nine customer facilities, with deployments including Schaeffler, GXO, Toyota Motor Manufacturing Canada and Mercado Libre.

Backers include Nvidia, Amazon, SoftBank Vision Fund 2, Foxconn and DCVC.

See 4 related tweets

@IPONewsroom_: AGILITY ROBOTICS IS GOING PUBLIC VIA SPAC AT A $2.5 BILLION VALUATION

The company behind the Digit ...

@WOLF_Financial: Agility Robotics is going public\n\nQT @IPONewsroom_: AGILITY ROBOTICS IS GOING PUBLIC VIA SPAC AT A...
@StockSavvyShay: Agility Robotics is going public via $CCXI at a$ 2.5B valuation and will trade under $AGLT.

The Dig...

@IPONewsroom_: HUMANOID ROBOT MAKER AGILITY ROBOTICS IS GOING PUBLIC IN A SPAC DEAL VALUING IT AT AROUND $2.5 BILLI...

15. julien_c (Group Score: 149.0 | Individual: 34.1)

Cluster: 5 tweets | Engagement: 36 (Avg: 121) | Type: Tech

more and more large-scale robotics datasets on the @huggingface hub!!\n\nQT @BitRobotNetwork: 1/ Introducing HIW-500 (Humanoids-in-the-Wild 500): the largest open-source humanoid teleop dataset collected in real homes

Built w/ @UnitreeRobotics @huggingface across 12 homes in Southeast Asia, it covers: > 500+ hrs > 23K+ episodes > 10+ TB > 10+ household tasks https://t.co/KUfFJm4wno

See 4 related tweets

@Thom_Wolf: Bitrobot casually dropping the largest humanoid teleop dataset ever collected in real homes

HIW-500...

@chris_j_paxton: Bitrobot is doing the work, making large amounts of data available to the community\n\nQT @BitRobotN...
@AdinaYakup: Super cool collab work from @BitRobotNetwork @UnitreeRobotics and @huggingface 🤗

https://t.co/B6OvO...

@chris_j_paxton: RT @BitRobotNetwork: 1/ Introducing HIW-500 (Humanoids-in-the-Wild 500): the largest open-source hum...

16. ClickHouseDB (Group Score: 145.1 | Individual: 34.3)

Cluster: 5 tweets | Engagement: 6 (Avg: 10) | Type: Tech

Congrats to the Unify team on their launch. 🚀

Unify runs on ClickHouse Cloud, because agentic products live or die on how fast the data answers.

They turned outbound into an agentic workflow, and the early growth is off the charts.

See why the agentic era demands a new data foundation.

https://t.co/k6hfkK5knl\n\nQT @austinh___: We built Claude for outbound sellers.

AEs & SDRs can harness GTM engineering through chat across 40+ data sources, no technical skills required.

We’ve had 57,548 queries in our first few weeks of beta, growing 45% w/w.

The teams winning outbound in 2026 are already on it. Self-serve now live here: https://t.co/pwdSGTG0Hq

See 4 related tweets

@alex_prompter: Your best salesperson is still hunting for email addresses.

That is insane.

40 tabs. Stale lists. ...

@hwchase17: awesome launch\n\nQT @austinh___: We built Claude for outbound sellers.

AEs & SDRs can harness GTM ...

@hwchase17: RT @austinh___: We built Claude for outbound sellers.

AEs & SDRs can harness GTM engineering throug...

@StockMKTNewz: Unify just said "We built Claude for outbound sellers."\n\nQT @austinh___: We built Claude for outbo...

17. PatrickMoorhead (Group Score: 141.1 | Individual: 31.5)

Cluster: 7 tweets | Engagement: 34 (Avg: 47) | Type: Tech

Some more deets on $META adopting datacenter$ QCOM CPUs in 2H/28 from the press release: -strategic multi-generation collaboration for Qualcomm Technologies to be a supplier for data center CPUs for Meta. (Note "collaboaration " which is MOU versus contract) -the Qualcomm Dragonfly C1000, is planned to power Meta’s next-generation server fleet (sounds firmer) -Qualcomm’s first-generation Qualcomm Dragonfly C1000 CPU will be in production starting in the second half of 2028.\n\nQT @PatrickMoorhead: Not one, but three different Qualcomm datacenter CPUs. One with HBC. $META CEO @finkd says they have a MOU to use them.$ QCOM https://t.co/6qfMvSD6GB

See 6 related tweets

@StockMKTNewz: QUALCOMM $QCOM JUST UNVEILED ITS FIRST DATA CENTER CPU AND META$ META IS THE LAUNCH CUSTOMER

The ...

@StockSavvyShay: $META plans to use$ QCOM C1000 CPU in its data centers giving the company a major hyperscaler design...
@LaurenGoode: RT @PatrickMoorhead: Some more deets on $META adopting datacenter$ QCOM CPUs in 2H/28 from the press...
@zerohedge: *QUALCOMM SAYS DRAGONFLY C1000 CPUS COMING IN 2028

*QUALCOMM SHARES QUICKLY EXTEND DECLINES TO 6%, ...

@financialjuice: Qualcomm: Dragonfly C1000 CPUs coming in 2028, Meta plans to use the C1000 CPU in its data centers. ...

18. wallstengine (Group Score: 140.1 | Individual: 65.6)

Cluster: 3 tweets | Engagement: 1253 (Avg: 132) | Type: Tech

$MU Q3’26 EARNINGS HIGHLIGHTS

🔹 Revenue: $41.46B (Est.$ 35.5B) 🟢 🔹 Adj. EPS: $25.11 (Est.$ 20.4) 🟢 🔹 Adj. Gross Margin: 84.9% (Est. 81.8%) 🟢

Q4 Guide: 🔹 Revenue: $50.0B ±$ 1.0B (Est. $43.4B) 🟢 🔹 Adj. EPS:$ 31.00 ± $1.00 (Est.$ 24.3) 🟢 🔹 Gross Margin: Approximately 86% (Est. 83%) 🟢 🔹 OpEx: Approximately $1.65B

Segment Performance: 🔹 Automotive & Embedded Revenue: $4.63B (Est.$ 3.51B) 🟢 🔹 Cloud Memory: Revenue $13.769B; Gross Margin 83%; Operating Margin 78% 🔹 Core Data Center: Revenue$ 11.524B; Gross Margin 87%; Operating Margin 83% 🔹 Mobile & Client: Revenue $11.521B; Gross Margin 87%; Operating Margin 86% 🔹 Automotive & Embedded: Revenue$ 4.634B; Gross Margin 79%; Operating Margin 75%

Other Metrics: 🔹 Operating Cash Flow: $25.39B 🔹 Adjusted FCF:$ 18.3B 🔹 HBM4: In high-volume shipments for lead customer platform; qualification samples shipped to multiple end-customers 🔹 HBM4E: Volume production expected in calendar 2027 🔹 G9-based PCIe Gen6 high-performance SSD: In high-volume production 🔹 High-capacity 245TB QLC SSD: Shipments commenced

Financials: 🔹 Adj. Net Income: $28.86B 🔹 Adj. Operating Income:$ 33.68B 🔹 Adj. OpEx: $1.52B 🔹 Investments in CapEx, net:$ 7.1B 🔹 Cash, marketable investments, and restricted cash: $30.2B

Capital Return: 🔹 Dividend: $0.15/share payable July 21, 2026 to shareholders of record as of July 6, 2026

Commentary: 🔸 “Micron’s record fiscal Q3 financial results and even stronger outlook for Q4 reflect the strategic value of memory in the AI era.”

See 2 related tweets

@wallstengine: MICRON $MU Q3’26 EARNINGS HIGHLIGHTS

🔹 Revenue: $41.46B (Est.$ 35.5B) 🟢 🔹 Adj. EPS: $25.11 (Est.$ 2...

@tengyanAI: tldr: BofA thinks AI adds another $1t to chip sales in five years
that's about what the industry ...

19. TeksEdge (Group Score: 138.5 | Individual: 29.0)

Cluster: 6 tweets | Engagement: 7 (Avg: 14) | Type: Tech

Google made Computer Use native in Gemini 3.5 Flash. It was previously only available as a standalone Gemini 2.5 model, it’s now built directly into the main Flash model.

Can't wait to see the benchmarks like OSWorld, Mind2Web, WebArena, WebVoyager.

Key features: 🖥️ Gemini power agents can now see, reason, and take actions on browser, desktop & mobile ⚡ Better speed and lower cost for long-horizon tasks 🛡️ Added adversarial training + optional safeguards (e.g. human confirmation for risky actions)\n\nQT @GoogleAIStudio: https://t.co/oFKtt2LToj

See 5 related tweets

@testingcatalog: Google released Computer Use capability for Gemini 3.5 Flash, scoring 78.4 on OSWorld-Verified.

&g...

@omarsar0: This is a cool release from Google.

I've started to rely a lot on computer use for agentic loops/lo...

@DynamicWebPaige: 🐨⚡️ @trycua had early access from @GoogleDeepMind to Gemini 3.5 Flash's native Computer Use! On Cua-...
@Techmeme: Google says computer use is now a built-in tool supported in Gemini 3.5 Flash, available via the Gem...
@Browserbase: RT @Stagehanddev: Gemini 3.5 Flash + CUA is impressive: on our benchmarks, it performs similarly to ...

20. zerohedge (Group Score: 135.7 | Individual: 19.0)

Cluster: 8 tweets | Engagement: 444 (Avg: 789) | Type: Tech

Anthropic accused Alibaba of waging a large-scale effort to “illicitly” access its Claude artificial intelligence model using thousands of fraudulent accounts: BBG

See 7 related tweets

@Mayhem4Markets: *COPYRIGHT HOLDERS AROUND THE PLANET ACCUSE ANTHROPIC OF STEALING HUMANITY'S INTELLECTUAL PROPERTY W...
@business: Anthropic accused Chinese technology giant Alibaba of waging a large-scale effort to “illicitly” acc...
@negligible_cap: *ANTHROPIC ACCUSES ALIBABA OF ‘ILLICITLY’ ACCESSING AI MODELS

In a letter to US officials, Anthropi...

@FirstSquawk: ANTHROPIC ACCUSED ALIBABA OF UNAUTHORIZED ACCESS TO ITS AI MODELS, OUTLINING ITS ALLEGATIONS IN LETT...
@Techmeme: Sources: in a letter to US officials, Anthropic accused Alibaba of adversarial distillation, accessi...