热门技术推文 - 2026-03-25

2026年3月25日科技每日简报

Today's top tech conversations are led by @AndrewCurran_, whose post about 'RT @claudeai: New in Claude Co...' garnered the highest engagement. Key themes trending across the top stories include figma, claude, design, https, agents. The community is actively discussing recent developments in AI, engineering practices, and startup strategies.

1. AndrewCurran_ (Group Score: 507.4 | Individual: 36.0)

Cluster: 17 tweets | Engagement: 2998 (Avg: 794) | Type: Tech

RT @claudeai: New in Claude Code: auto mode.

Instead of approving every file write and bash command, or skipping permissions entirely, auto mode lets Claude make permission decisions on your behalf.

Safeguards check each action before it runs. https://t.co/kHbTN2jrWw

See 16 related tweets

@kimmonismus: This is insane. They literally drop every freaking day.

Today: Claude code auto mode. Permission d...

@omarsar0: Claude Code now supports auto mode.\n\nQT @claudeai: New in Claude Code: auto mode.

Instead of appr...

@bcherny: no 👏 more 👏 permission prompts 👏\n\nQT @claudeai: New in Claude Code: auto mode.

Instead of approvi...

@rohanpaul_ai: Claude Code auto mode is a new middle setting between approve-everything and '--dangerously-skip-per...
@testingcatalog: Anthropic released Auto Mode for Claude Code CLI, which allows Claude to make its own decisions on w...

2. seiiiiiiiiiiru (Group Score: 349.5 | Individual: 55.2)

Cluster: 12 tweets | Engagement: 2980 (Avg: 244) | Type: Tech

RT @soraofficialapp: We’re saying goodbye to the Sora app. To everyone who created with Sora, shared it, and built community around it: thank you. What you made with Sora mattered, and we know this news is disappointing.

We’ll share more soon, including timelines for the app and API and details on preserving your work. – The Sora Team

See 11 related tweets

@testingcatalog: BREAKING 🚨: OPENAI DISCONTINUES SORA, MORE DETAILS WILL BE SHARED LATER.

Video generation is movin...

@chatgpt21: Superapp incoming.\n\nQT @soraofficialapp: We’re saying goodbye to Sora. To everyone who created wit...
@tunguz: This is crazy.\n\nQT @soraofficialapp: We’re saying goodbye to Sora. To everyone who created with So...
@danshipper: whoa! really interesting\n\nQT @soraofficialapp: We’re saying goodbye to Sora. To everyone who creat...
@seconds_0: they need the compute back\n\nQT @soraofficialapp: We’re saying goodbye to Sora. To everyone who cre...

3. starbuxman (Group Score: 332.2 | Individual: 48.1)

Cluster: 14 tweets | Engagement: 26992 (Avg: 1201) | Type: Tech

RT @claudeai: You can now enable Claude to use your computer to complete tasks.

It opens your apps, navigates your browser, fills in spreadsheets—anything you'd do sitting at your desk.

Research preview in Claude Cowork and Claude Code, macOS only. https://t.co/sVymgmtEMI

See 13 related tweets

@EHuanglu: brutal… Claude just cloned OpenClaw

now you can text AI agents on your laptop to edit hundreds of p...

@dejavucoder: claude is becoming every wrapper startup's biggest competitor more and more everyday\n\nQT @claudeai...
@kimmonismus: One has to admit, Anthropic sees a trend (openClaw) and develops their own version.

The speed of th...

@WesRoth: Claude Cowork and Claude Code can now directly operate a user's Mac, executing workflows exactly lik...
@rohanpaul_ai: You can now enable Claude to use your computer to complete tasks.

When computer use is needed, Clau...

4. eastdakota (Group Score: 171.8 | Individual: 62.4)

Cluster: 4 tweets | Engagement: 768 (Avg: 67) | Type: Tech

This is a big deal.\n\nQT @Cloudflare: We’re introducing Dynamic Workers, which allow you to execute AI-generated code in secure, lightweight isolates. This approach is 100 times faster than traditional containers. https://t.co/c36Vkb7I0R

See 3 related tweets

@Cloudflare: We’re introducing Dynamic Workers, which allow you to execute AI-generated code in secure, lightweig...
@larsencc: Trillions I tell you\n\nQT @Cloudflare: We’re introducing Dynamic Workers, which allow you to execut...
@eastdakota: RT @KentonVarda: This is all made possible by the isolation model that Cloudflare Workers has been b...

5. MSBIntel (Group Score: 151.1 | Individual: 31.4)

Cluster: 8 tweets | Engagement: 14 (Avg: 14) | Type: Tech

TERAFAB might be the most ambitious industrial project announced this decade.

What Elon Musk announced Saturday:

A $25 billion chip factory. Joint venture between Tesla, SpaceX, and xAI. Built in Austin.
Targeting 1 terawatt of annual compute. The entire US currently produces about half a terawatt.
At full capacity: 1 million wafer starts/month. That's ~70% of TSMC's entire global output. From one facility.
Two chips: AI5 for Tesla vehicles and Optimus robots. D3 for AI data centers in space.
80% of output goes to orbit. Not Earth. Space.
SpaceX filed with the FCC for 1 million data center satellites in LEO.
Musk says AI training in space will be cheaper than on Earth within 2-3 years. Solar power is 5x stronger in orbit.

No timeline. No semiconductor manufacturing experience. Tesla's Battery Day promised a revolution in 2020 that still hasn't fully delivered.

But if even 10% of this ships, it reshapes the entire chip industry.

"We either build the Terafab or we don't have the chips. And we need the chips."

See 7 related tweets

@MarioNawfal: TERAFAB unveiled. $20–25B. Not a side project.

A trillion watts of compute per year. Yes, trillion....

@MarioNawfal: RT @MarioNawfal: 🇺🇸 Elon says he’s done waiting on the chip industry

Tesla and SpaceX are planning ...

@FirstSquawk: MUSK PLANS AI CHIP MEGA-FACTORY

Elon Musk says Tesla and SpaceX will build a major chip facility in...

@SawyerMerritt: Dan Ives in new $TSLA note:

"We believe this (Terafab project) is the first step to ultimately what...

@MarioNawfal: 🇺🇸 Tesla and SpaceX are planning to build a massive chip facility in Austin called Terafab.

Two fa...

6. aakashgupta (Group Score: 140.8 | Individual: 35.6)

Cluster: 5 tweets | Engagement: 91 (Avg: 746) | Type: Tech

Figma turned the design canvas into a write API for AI agents this morning.

Free! (During beta.) Will be a paid, usage-based API after. They said they're "learning how to account for agentic behavior in our paid seats," which tells you this is as much a business model experiment as a product launch.

The timing makes sense. Every major coding agent already has MCP support. Claude Code, Codex, Cursor, Copilot, Warp, Augment. Figma just made itself the shared context layer across all of them. Whichever agent wins, Figma wins, because the design system lives in their files.

That's the real play. AI-generated code looks generic because agents have zero design context. No tokens, no components, no spacing rules, no brand. Figma is selling the fix by making the canvas machine-readable and machine-writable at the same time. OpenAI's design lead at Codex already confirmed they use it internally.

If you build software, here's what to actually do with it.

Two tools matter. get_design_context pulls layout data, tokens, and variables from any Figma file into your agent. use_figma writes back. That round-trip kills the screenshots-to-code workflow: pull a design, generate code, push running UI back to the canvas as editable layers, iterate with your team, commit to code.

Now the part Figma buried in the docs. They shipped "skills," packaged instructions that tell agents how to sequence tool calls against your specific design system. figma-implement-design turns a Figma selection into working code using your project's existing components. figma-code-connect-components links published components to their matching code so Dev Mode shows the actual import path. You can write custom skills for your team's own patterns. This is where the compounding happens.

One thing that will save you hours of debugging: use the remote server at https://t.co/S10T1FG0bl. Desktop is read-only for most operations. Multiple developers have already reported needing to disconnect all previous Figma MCP instances and reconnect fresh before write tools even appear. If use_figma is missing from your tool list, that's why.

Free during beta means now is when you build the muscle memory.\n\nQT @figma: Now you can use AI agents to design directly on the Figma canvas, with our new use_figma MCP tool and skills to teach them. Open beta starts today. https://t.co/AQZsFWvvXQ

See 4 related tweets

@realsanketp: RT @figma: Now you can use AI agents to design directly on the Figma canvas, with our new use_figma ...
@rohanpaul_ai: You can now use AI agents / Claude Code to design directly in Figma with full access to your design ...
@minchoi: Figma just gave AI agents the canvas.

Claude Code can now design directly in Figma via MCP + Skills...

@minchoi: RT @minchoi: Figma just gave AI agents the canvas.

Claude Code can now design directly in Figma via...

7. GenAI_is_real (Group Score: 130.5 | Individual: 31.6)

Cluster: 5 tweets | Engagement: 8 (Avg: 149) | Type: Tech

sora was technically incredible but the math never worked. generating video is absurdly compute-intensive and openai decided to refocus those GPU resources on coding and core models. sad but predictable. in AI right now, capability is abundant but compute is finite. the products that survive are the ones where revenue per GPU-hour exceeds cost per GPU-hour. sora couldnt clear that bar. a $1B disney deal couldnt save it. inference economics is the ultimate product filter @soraofficialapp\n\nQT @soraofficialapp: We’re saying goodbye to the Sora app. To everyone who created with Sora, shared it, and built community around it: thank you. What you made with Sora mattered, and we know this news is disappointing.

We’ll share more soon, including timelines for the app and API and details on preserving your work. – The Sora Team

See 4 related tweets

@cryptopunk7213: openai just killed Sora, the AI video app that hit 1 million downloads in <5 days of launching

it a...

@BrianRoemmele: FOR FREE I advised in public OpenAI to not do Sora and the impact of desensitization of AI video.

M...

@waitin4agi_: AI video models probably very unprofitable or even when useful held back by copyright/censoring issu...
@burkov: I never tried Sora. Should I while I still have time?

I predict that soon, only Elon will keep trai...

8. bcherny (Group Score: 126.3 | Individual: 39.8)

Cluster: 4 tweets | Engagement: 3784 (Avg: 1408) | Type: Tech

Little known fact, the Anthropic Labs team (the team I joined Anthropic to be on) shipped:

MCP
Skills
Claude Desktop app
Claude Code

It was just a few of us, shipping fast, trying to keep pace with what the model was capable of.

Those early Desktop computer use prototypes, back in the Sonnet 3.6 days, felt clunky and slow. But it was easy to squint and imagine all the ways people might use it once it got really good.

Fast forward to today. I am so excited to release full computer use in Cowork and Dispatch. Really excited to see what you do with it!\n\nQT @claudeai: You can now enable Claude to use your computer to complete tasks.

It opens your apps, navigates your browser, fills in spreadsheets—anything you'd do sitting at your desk.

Research preview in Claude Cowork and Claude Code, macOS only. https://t.co/sVymgmtEMI

See 3 related tweets

@aakashgupta: It only took Anthropic 4 weeks from acquisition to ship computer use https://t.co/AGQjyl4jqv\n\nQT @...
@gregisenberg: Just a "few" people built MCP, Claude Skills, Claude Desktop App and Claude Code 🤯

It's inspiration...

@garrytan: RT @bcherny: Little known fact, the Anthropic Labs team (the team I joined Anthropic to be on) shipp...

9. hwchase17 (Group Score: 120.5 | Individual: 56.8)

Cluster: 3 tweets | Engagement: 1102 (Avg: 49) | Type: Tech

RT @anvisha: We raised $7.5M to kill AI slop.

Introducing Moda: the world's first design agent with taste.

RT+ comment “Moda” and we’ll design your brand for FREE. https://t.co/5xiqxiTXER

See 2 related tweets

@hwchase17: congrats to Moda on the launch of their design agent!

it's built on deep agents - all agents going ...

@LangChain: Congrats @anvisha and the @trymoda team on the launch.

Moda built a design platform that turns non-...

10. myfear (Group Score: 115.2 | Individual: 38.1)

Cluster: 4 tweets | Engagement: 204 (Avg: 38) | Type: Tech

RT @svpino: Every large company will eventually ban vibe-coding.

Vibe-coding is now generating as much technical debt as 10 regular developers in half the time.

Vibe-coding is awesome for a first draft, but you can't expect to push AI slop to production and not destroy your software over time.

Producing code is no longer a bottleneck.

Testing that code, debugging it, monitoring it in production, and fixing it when it breaks is where everyone is spending their time.

We've 10x'd the speed of writing code, but we are still in the Stone Age with everything that happens after the code is written.

Here is a very cool tool tackling this:

You can build "AI Production Engineers" using PlayerZero and make them work for you.

These are agents that do this:

• Simulate how your code will work in production • Diagnose issues when they happen • Learn from every incident so it doesn't happen again

This is pretty awesome!

These agents simulate code behavior against real production data. They use actual customer behavior, historical incidents, and edge cases without writing a single test script.

When something breaks, the agent traces the issue to the exact line of code and PR, generates the fix, and routes it to the right engineer.

And every bug these agents solve serves as training data to improve the system.

Here is a link to check them out: https://t.co/JWsbNp0hTo

Thanks to the Player Zero team for partnering with me on this post.

See 3 related tweets

@dejavucoder: here's a more generalised version of dan shipper's post on pirate (vibe coder) and architect (slop c...
@amasad: RT @etnshow: Michele Catasta (@pirroh) President of Replit (@Replit) says "AGI for vibe-coding is co...
@burkov: I wish I could see some of those millions of companies making lots of money out of vibe-coded soluti...

11. dr_cintas (Group Score: 108.1 | Individual: 38.0)

Cluster: 4 tweets | Engagement: 255 (Avg: 224) | Type: Tech

Alibaba just launched Accio Work 🤯

Zero setup. Zero code. A full squad of AI agents runs your operations from day one.

Market analysis, supplier negotiations, compliance, inventory. All in parallel.

Solo founders just got an ops team.\n\nQT @Accio_official: Running a business shouldn’t mean doing 8 jobs alone. Meet Accio Work: your desktop business agent team that actually connects with your Shopify and gets it done. 10 million users has used Accio Work to run business, it’s your turn. Try free: https://t.co/dsJNxeGUl6 https://t.co/kBXzO4pqzf

See 3 related tweets

@dr_cintas: RT @dr_cintas: Alibaba just launched Accio Work 🤯

Zero setup. Zero code. A full squad of AI agents ...

@dr_cintas: 🚨 Alibaba just turned Accio into a full agent workforce.

Accio Work deploys a "squad" of specialize...

@heyshrutimishra: RT @heyshrutimishra: This AI tool is like Claude Cowork but built for E-Commerce founders.

Accio ju...

12. WesRoth (Group Score: 104.9 | Individual: 35.4)

Cluster: 3 tweets | Engagement: 9 (Avg: 54) | Type: Tech

PlayerZero, an AI production engineering startup, announced a $20 million funding round to scale its autonomous debugging and testing platform.

The round was led by Foundation Capital, with heavy-hitting angel investments from tech founders and executives at Databricks, Workday, Vercel, Figma, and Dropbox.\n\nQT @akoratana: Introducing: PlayerZero

The world's first Engineering World Model that puts debugging, fixing, and testing your code on autopilot.

We've raised $20M from Foundation Capital, @matei_zaharia (Databricks), @pbailis (Workday), @rauchg (Vercel), @zoink (Figma), @drewhouston (Dropbox), and more

PlayerZero frees up 30% of your engineering bandwidth by: 1.⁠ ⁠Finding the root cause for bugs & incidents in minutes that engineering teams take days to identify. 2.⁠ ⁠Predicting in minutes, edge case issues that a 300-person QA team would take weeks to find.

Here's why this matters:

No one in your org has a complete picture of how your production software actually behaves.

Support sees tickets. SRE sees infra. Dev sees code. Each team builds their own fragmented view - and none of these systems talk to each other. When something breaks, everyone scrambles to stitch the picture together by hand.

PlayerZero connects all of it into a single context graph -

→ The Slack thread where your lead said "we went with X because Y fell apart in prod last time" → The PR review where an engineer explained the tradeoff → The lifetime history of your CI/CD pipeline, observability stack, incidents, and support tickets

So you can trace any problem to its root cause across every silo.

And it compounds. Every incident diagnosed teaches the model something new. The longer it runs, the deeper it understands - which code paths are high-risk, which configurations are fragile, which changes tend to break which customer flows.

So when you sit down to debug a live issue, you have your entire org's collective reasoning and production memory behind you - instantly.

Zuora, Georgia-Pacific, and Nylas have reduced resolution time by 90% and caught 95% of breaking changes and freeing an average of $30M in engineering bandwidth.

Our guarantee:

If we can't increase your engineering bandwidth by at least 20% within one week, we'll donate $10,000 to an open-source project of your choice.

Book a demo - https://t.co/dH1dulIwSS

See 2 related tweets

@nalinrajput23: Raised $20M to catch AI-era bugs faster simulate production, trace root causes, and reclaim 30% engi...
@chhddavid: NO WAYYY...

IBM has 100,000 consultants whose entire job is investigating production bugs.

PlayerZ...

13. testingcatalog (Group Score: 101.7 | Individual: 37.0)

Cluster: 4 tweets | Engagement: 323 (Avg: 245) | Type: Tech

OpenAI is rolling out an upgraded shopping experience on ChatGPT, with more visual enhancements, better product coverage, and more customization options for merchants. https://t.co/1S5spGidGE\n\nQT @btibor91: OpenAI is making shopping in ChatGPT richer and more visual with side-by-side comparison, image uploads and improved speed, relevance and product coverage, powered by extending the Agentic Commerce Protocol (ACP) for product discovery, rolling out to all free, Go, Plus and Pro users this week

OpenAI found that the initial version of Instant Checkout did not offer the level of flexibility they aspire to provide, so they are allowing merchants to use their own checkout experiences while focusing efforts on product discovery
Target, Sephora, Nordstrom, Lowe's, Best Buy, The Home Depot and Wayfair are already integrated into ACP for discovery, with third-party providers like Salesforce and Stripe supported so merchants can use systems they already have
Shopify merchants are automatically in through Shopify Catalog without extra work, and Walmart is introducing an in-ChatGPT app with account linking, loyalty and Walmart payments

See 3 related tweets

@rohanpaul_ai: OpenAI is killing Sora as a standalone app, its developer version, and video inside ChatGPT.

Shows...

@TechCrunch: OpenAI’s plans to make ChatGPT more like Amazon aren’t going so well https://t.co/Mgul11VLJ5...
@CNBC: OpenAI revamps shopping experience in ChatGPT after struggling with Instant Checkout offering https:...

14. kimmonismus (Group Score: 95.4 | Individual: 35.8)

Cluster: 4 tweets | Engagement: 343 (Avg: 303) | Type: Tech

Either OpenAI officially achieved AGI or this is the biggest troll move ever:

they rename product organization to "AGI Deployment"
Altman says the next LLM is a "very strong model"
it very much accelerate the economy

Quote:

"Altman also said that the company would be renaming senior executive Fidji Simo’s product organization to “AGI Deployment,” a reference to artificial general intelligence, or AI that’s roughly on par with humans."

However, Altman says "Spud is very strong model" in “a few weeks” that the team believes “can really accelerate the economy.”\n\nQT @kimmonismus: OpenAI finished the initial developement of its next major LLM: codenamed Spud (GPT-5.5 / 6.0)

Sam Altman however is "raising capital, supply chains and “building datacenters at unprecedented scale,”

See 3 related tweets

@kimmonismus: OpenAI's Sora team is now working on world-models
they prioritize longer-term world simulation re...
@Miles_Brundage: Probably nothing\n\nQT @deredleritt3r: New "very strong" OpenAI model, with ability to "accelerate t...
@FirstSquawk: OPENAI CEO ALTMAN SHIFTS RESPONSIBILITIES, PREPS NEW 'SPUD' AI MODEL - THE INFORMATION...

15. minchoi (Group Score: 94.1 | Individual: 29.0)

Cluster: 4 tweets | Engagement: 174 (Avg: 389) | Type: Tech

Claude Code just made setup way easier.

The new /init now interviews you and helps configure CLAUDE.md, hooks, skills, and more🤯 https://t.co/vBskK699NZ\n\nQT @minchoi: This Claude Code project setup guide literally makes Claude Code 10x better🤯 https://t.co/zcggKFV7qN

See 3 related tweets

@akshay_pachaar: Claude Code just made setup 10x easier.

The new /init now interviews you and helps configure CLAUDE...

@minchoi: RT @minchoi: Claude Code just made setup way easier.

The new /init now interviews you and helps con...

@minchoi: RT @minchoi: This Claude Code project setup guide literally makes Claude Code 10x better🤯 https://t....

16. natolambert (Group Score: 81.4 | Individual: 29.7)

Cluster: 3 tweets | Engagement: 82 (Avg: 239) | Type: Tech

Really excited for the things @rosstaylor90 has been building to go out into the world. He's been one of the people I can always rely on to have non-cope takes on what we need to do to make the open ecosystem great.

What a great time in RL.\n\nQT @GenReasoning: Introducing OpenReward.

🌍 330+ RL environments through one API ⚡ Autoscaled sandbox compute 🍒 4.5M+ unique RL tasks 🚂 Works like magic with Tinker, Miles, Slime

Link and thread below. https://t.co/4fIlVKUkOF

See 2 related tweets

@omarsar0: OpenReward with 330+ RL environments.\n\nQT @GenReasoning: Introducing OpenReward.

🌍 330+ RL enviro...

@andersonbcdefg: RT @GenReasoning: Introducing OpenReward.

🌍 330+ RL environments through one API ⚡ Autoscaled sandb...

17. chddaniel (Group Score: 79.7 | Individual: 44.3)

Cluster: 2 tweets | Engagement: 172 (Avg: 27) | Type: Tech

Today, we've ended of vibe coding.

Introducing Shipper - the world's first autonomous AI company builder.

Think of it as a co-founder for your business, available today in production as a bot you're talking to.

What makes Shipper great: → Brings any business ideas to life → Creates mobile apps, web apps, websites, extensions & more → Emails you update with what it's done → Self-maintains the business for you

Under the hood: Claude Code Opus 4.6 designs, codes, monetizes, launches, translates your app and even does email marketing. Infinite ideas trained on +1,000 successful startups all over the world.

Talk to the AI Business Builder on our homepage.

Companies and indie hackers are already building companies that are: → Fully built and maintained by Shipper → Used by customers of their own → Translated in +68 languages from a single click

People don’t want to build apps, they want the success. Shipper handles the build and the success for you, 24/7.

To celebrate the launch, we're randomly giving away free credits.

Retweet and comment "SHIPPER" - Siri randomly selects the winners.

See 1 related tweets

@chddaniel: Let AI agents take over your computer.

Introducing Shipper - the world's first autonomous AI compan...

18. CSProfKGD (Group Score: 79.5 | Individual: 42.5)

Cluster: 3 tweets | Engagement: 4740 (Avg: 671) | Type: Tech

RT @lexfridman: Here's my conversation with Jensen Huang, CEO of NVIDIA, the most valuable & one of the most influential companies in the history of human civilization. It is the engine powering the AI revolution.

This was a fascinating & inspiring conversation, in parts super-technical on engineering of every part of the AI stack, memory, power, supply chain (TSMC, ASML, etc), in parts about leadership & psychology, and in parts personal & philosophical about life, consciousness, mortality, and human nature.

It's here on X in full and is up everywhere else (see comment).

Timestamps: 0:00 - Introduction 0:33 - Extreme co-design and rack-scale engineering 3:18 - How Jensen runs NVIDIA 22:40 - AI scaling laws 37:40 - Biggest blockers to AI scaling laws 39:23 - Supply chain 41:18 - Memory 47:24 - Power 52:43 - Elon and Colossus 56:11 - Jensen's approach to engineering and leadership 1:01:37 - China 1:09:50 - TSMC and Taiwan 1:15:04 - NVIDIA's moat 1:20:41 - AI data centers in space 1:24:30 - Will NVIDIA be worth $10 trillion? 1:34:39 - Leadership under pressure 1:48:25 - Video games 1:55:16 - AGI timeline 1:57:29 - Future of programming 2:11:01 - Consciousness 2:17:22 - Mortality

See 2 related tweets

@niccruzpatane: NVIDIA CEO Jensen Huang thinks AGI has already been achieved.

Lex: "Some agents, some humans, all ...

@BusinessInsider: Nvidia CEO Jensen Huang discussed AI's impact on jobs, encouraging workers from farmers to carpenter...

19. MSBIntel (Group Score: 77.1 | Individual: 31.9)

Cluster: 3 tweets | Engagement: 3 (Avg: 14) | Type: Tech

The most expensive breakup in tech history is happening in public filings.

OpenAI just told its pre-IPO investors that Microsoft is a risk to their business. Not a partner. A risk.

The numbers:

Microsoft invested ~ $13 billion in OpenAI. Holds 27% equity (~$ 135 billion). Provides the majority of OpenAI's compute through Azure. Added OpenAI to its official competitor list in 2024.

OpenAI's response: signed a $50 billion deal with Amazon.$ 30 billion with Oracle. $11.9 billion with CoreWeave. Routed workloads through Google Cloud.

45% of Azure's contracted backlog is tied to OpenAI API revenue.

So OpenAI is Microsoft's largest customer, largest investment, and now its largest competitive threat. Simultaneously.

Microsoft built Copilot Cowork on Anthropic's Claude. Not on OpenAI. Despite owning 27% of OpenAI. That tells you everything about where this relationship is heading.

OpenAI valued at up to $1 trillion for IPO. Microsoft's stake:$ 135 billion. The question is whether that stake is an asset or a liability when your biggest investment is also your biggest competitor.

See 2 related tweets

@ns123abc: 🚨 BREAKING: OPENAI PRE-IPO FILING LEAKED

OpenAI just told investors Microsoft is a risk to their bu...

@elonmusk: RT @ns123abc: 🚨 BREAKING: OPENAI PRE-IPO FILING LEAKED

OpenAI just told investors Microsoft is a ri...

20. kimmonismus (Group Score: 75.9 | Individual: 32.8)

Cluster: 3 tweets | Engagement: 216 (Avg: 303) | Type: Tech

Apples Siri gets a huge AI rework this year, biggest rework ever. The summary tl;dr

•Apple is rebuilding Siri into a system-wide AI agent, not just a voice assistant

•iOS 27 introduces a conversational, chat-like Siri (text + voice)

•New standalone Siri app with chat history, file uploads, and persistent interactions

•“Ask Siri” button integrates AI across apps (contextual actions on selected content)

•Siri gains deep access to personal data (messages, emails, notes) for task execution

•Replaces Spotlight with a unified AI search + assistant interface

•Can perform in-app actions and browse the web with Apple-built models

•“Write with Siri” adds system-wide AI writing/editing tools

•Powered by Apple Foundation Models + Google Gemini partnership

•Many advanced features delayed, expected rollout continues into late 2026\n\nQT @markgurman: BREAKING: Apple’s AI reboot this year detailed — Dedicated Siri app to rival ChatGPT; Overhauled Siri interface in the Dynamic Island with chatbot; Unified Siri and Spotlight Search; and “Ask Siri” & “Write with Siri” features. https://t.co/LG4k4U5CGB

See 2 related tweets

@markgurman: BREAKING: Apple’s AI reboot this year detailed — Dedicated Siri app to rival ChatGPT; Overhauled Sir...
@business: Apple is testing a standalone app for its Siri voice assistant alongside a new “Ask Siri” feature th...