科技推特精选 - 2026年5月12日

2026年5月12日科技每日简报

Today's top tech conversations are led by @Dimillian, whose post about 'RT @OpenAI: Today we’re launch...' garnered the highest engagement. Key themes trending across the top stories include claude, openai, models, building, gemini. The community is actively discussing recent developments in AI, engineering practices, and startup strategies.

1. Dimillian (Group Score: 614.5 | Individual: 53.2)

Cluster: 25 tweets | Engagement: 1492 (Avg: 112) | Type: Tech

RT @OpenAI: Today we’re launching the OpenAI Deployment Company to help businesses build and deploy AI.

It's majority-owned and controlled by OpenAI. It brings together 19 leading investment firms, consultancies, and system integrators to help organizations deploy frontier AI to production for business impact. https://t.co/GnyjGFaLLA

See 24 related tweets

@cryptopunk7213: ok this is a much bigger announcement than people realize

in the last 5 days BOTH anthropic and ope...

@kimmonismus: OpenAI is no longer just selling models.

With its new Deployment Company, OpenAI is moving deeper i...

@ibuildthecloud: Ah crap. OpenAI needed an enterprise play and this is it. The amount of money this is going to make ...
@cgtwts: Sam Altman after copying whatever Anthropic has been doing for the last 6 months: https://t.co/6Enbi...
@business: OpenAI will acquire a consulting company to contribute to a new private equity-backed joint venture ...

2. pitdesi (Group Score: 260.0 | Individual: 36.4)

Cluster: 12 tweets | Engagement: 1227 (Avg: 264) | Type: Tech

600 OpenAI employees sold an average of $11M each of equity in October

Anecdotally it does feel like that’s when things started getting particularly nutty in the $5-10M segment of SF real estate\n\nQT @zerohedge: *MORE THAN 600 OPENAI EMPLOYEES SOLD SHARES LAST OCT.: WSJ *OPENAI EMPLOYEES COLLECTIVELY MADE$ 6.6B IN THE SHARE SALE: WSJ

See 11 related tweets

@negligible_cap: *OPENAI EMPLOYEES COLLECTIVELY MADE $6.6B IN THE SHARE SALE: WSJ

Nothing wrong with taking some chi...

@Hesamation: it's hard to believe that 6 years ago, OpenAI was a non-profit research lab working on AI agents pla...
@WSJ: Exclusive: OpenAI allowed employees to sell up to $30 million worth of shares each https://t.co/uxYq...
@aiedge_: OpenAI employee fun facts:

< estimated 70%+ are millionaires < average comp per employee is ~...

@KobeissiLetter: BREAKING: OpenAI allowed more than 600 current and former employees to sell stock in October 2025, p...

3. svpino (Group Score: 234.7 | Individual: 32.9)

Cluster: 8 tweets | Engagement: 84 (Avg: 108) | Type: Tech

So many companies are trying to kill the SaaS model, that for the first time I think we might have a shot at creating something different.

Literally everyone wants to replace 12 apps with a single, integrated interface that does everything we need.

AI can make this happen.\n\nQT @GilboaAmitay: We got an 8-figure acquisition offer 2 days after launch.

We said no, because the problem we're solving is worth way more than that.

It’s 2026, but teams are only getting lonelier, and context is still the problem.

The issue isn’t intelligence. Your team has plenty of that.

It’s shared memory and context, the thing that makes 10 A-players feel like 1.

That’s what we’ve solved with @playdotfast, while making work more fun.

We're killing traditional SaaS, and believe you me, we're leaving no holds barred.

See 7 related tweets

@rohanpaul_ai: Two years into the AI boom, and not a single team got faster. Individuals did. Teams didn't.

That i...

@svpino: You either move fast, or you are left behind.

AI companies are ruthless... they are coming after ev...

@cgtwts: Salesforce convinced an entire generation of companies that dashboards, tabs and workflows were prod...
@markgadala: Single-player AI was a mistake.

Your team did not need 100 private genius assistants. It needed one...

@seraleev: X is the best platform to announce a startup launch. Over a million views in just a few hours after ...

4. kimmonismus (Group Score: 215.0 | Individual: 33.9)

Cluster: 9 tweets | Engagement: 696 (Avg: 323) | Type: Tech

Anthropic: “Claude Mythos is too cyber-capable to release broadly. We need tight controls. 😳”

OpenAI: “Here’s GPT-5.5-Cyber, Codex Security, Trusted Access tiers, repo scanning, patch generation, and red-team workflows. Please be verified first, but yes, go find the bugs. 😎”\n\nQT @OpenAI: Introducing Daybreak: frontier AI for cyber defenders.

Daybreak brings together the most capable OpenAI models, Codex, and our security partners to accelerate cyber defense and continuously secure software.

A step toward a future where security teams can move at the speed defense demands.

See 8 related tweets

@charliermarsh: RT @OpenAI: Introducing Daybreak: frontier AI for cyber defenders.

Daybreak brings together the mos...

@minchoi: OpenAI just dropped Daybreak.

AI agents for finding vulnerabilities, validating fixes, and burning ...

@cryptopunk7213: hell yeah! openai just released their new mythos-level security model (before anthropic did)

this s...

@chatgpt21: OpenAI releases Daybreak, GPT 5.5 cyber - with the most “permissive behavior for specialized authori...
@jxnlco: New era of cool names for OpenAI\n\nQT @OpenAI: Introducing Daybreak: frontier AI for cyber defender...

5. ying11231 (Group Score: 150.6 | Individual: 29.8)

Cluster: 7 tweets | Engagement: 422 (Avg: 78) | Type: Tech

RT @thinkymachines: People talk, listen, watch, think, and collaborate at the same time, in real time. We've designed an AI that works with people the same way.

We share our approach, early results, and a quick look at our model in action.

https://t.co/AFJZ5kH7Ku https://t.co/uxl1InS6Ay

See 6 related tweets

@soumithchintala: Thinky's secret plan:

1: Increase Human<->AI bandwidth 2: Raise ceiling of human+AI intelligence 3:...

@kimmonismus: I think this is bigger than it sounds at first glance.

Thinking Machines hasn't just unveiled "Chat...

@giffmana: Very cool announcement from Thinky!

The model looks nice (they go into some reasonable amount of d...

@testingcatalog: THINKING MACHINES 🔥: Research preview of a new family of realtime voice models have been announced! ...
@SIGKITTEN: oh wow looky at thinky - they got Her\n\nQT @thinkymachines: People talk, listen, watch, think, and ...

6. StockSavvyShay (Group Score: 146.8 | Individual: 40.3)

Cluster: 4 tweets | Engagement: 1543 (Avg: 588) | Type: Tech

$IONQ is flying past$ 55 today after SkyWater shareholders approved the company’s sale to IonQ.

The deal could make IonQ the only U.S. quantum company with its own American chip factory. https://t.co/8Xd1MjTo0j\n\nQT @StockSavvyShay: BUILDING THE INFRASTRUCTURE OF TOMORROW

$RKLB building the vertically integrated logistics layer of the space economy through launch vehicles, spacecraft buses, satellite components & propulsion systems backed by a$ 1.8B backlog & a growing government manifest.
$ASTS building the communications layer that connects directly to standard smartphones with BlueBird 7 set to launch April 19 & more than$ 1B in contracted revenue commitments already in place.
$OKLO building the “local nuclear plant” for the AI economy through its Aurora microreactor with$ META targeting up to 1.2 GW in Ohio & federal momentum around advanced nuclear continuing to build.
$ONDS building the autonomous defense stack across drone communications, counter-drone systems, robotics & secure ISR networks with FY2026 revenue guided to$ 375M.
$IONQ building quantum as the overflow compute layer for problems classical silicon cannot solve with native availability on$ AMZN AWS, $MSFT Azure &$ GOOGL Cloud plus a recent $54M AFRL contract for quantum networking research.
$PL building the real-time geospatial intelligence layer from orbit with a$ 900M growing backlog & expanding traction across NATO & U.S. government programs.
$NBIS building sovereign AI compute infra outside the hyperscaler stack with$ 46B in contracted AI infra deals, a $2B$ NVDA strategic investment & connected data center capacity guided to 900 MW by the end of 2026.

See 3 related tweets

@StockSavvyShay: $ASTS achieved a 98.9 Mbps peak download speed from space directly to an unmodified smartphone using...
@StockSavvyShay: $RKLB BREAKING ABOVE$ 110 🤯 https://t.co/6WCe0FpUjg\n\nQT @StockSavvyShay: MY 3 FAVORITE SPACE ECON...
@StockSavvyShay: RT @StockSavvyShay: MY 3 FAVORITE SPACE ECONOMY STOCKS

$RKLB (Vertically Integrated Defense Pri...

7. TeksEdge (Group Score: 103.3 | Individual: 36.3)

Cluster: 3 tweets | Engagement: 60 (Avg: 8) | Type: Tech

📰 A new coding Agent Index was just released by @ArtificialAnlys. This measures both the model and the harness. No open source harnesses included.

OpenSource for Coding is now legit.

🔥 Claude Code+GLM-5.1 (53) > Claude Code+Sonnet 4.5 🔥 Claude Code+GLM-5.1 (53) > Gemini CLI+Gemini3.1

DeepSeek V4 Pro & Kimi K2.6 also hit 50\n\nQT @ArtificialAnlys: Announcing the Artificial Analysis Coding Agent Index! Our new coding agent benchmarks measure how combinations of agent harnesses and models perform on 3 leading benchmarks, token usage, cost and more

When developers use AI to code they’re choosing a model, but also pairing it with a specific harness. It makes sense to benchmark that combination to understand and compare performance.

The Artificial Analysis Coding Agent Index includes 3 leading benchmarks that represent a broad spectrum of coding agent use: ➤ SWE-Bench-Pro-Hard-AA, 150 realistic coding tasks that frontier models struggle with, sampled from Scale AI’s SWE-Bench Pro ➤ Terminal-Bench v2, 84 agentic terminal tasks from the Laude Institute and that range from system administration and cryptography to machine learning. 5 tasks were filtered due to environment incompatibility ➤ SWE-Atlas-QnA, 124 technical questions developed by Scale AI about how code behaves, root causes of issues, and more, requiring agents to explore codebases and give text answers

Analysis of results: ➤ Opus 4.7 and GPT-5.5 lead the Index: Opus 4.7 in Cursor CLI scores 61, followed closely by GPT-5.5 in Codex and Opus 4.7 in Claude Code at 60. GPT-5.5 in Cursor CLI follows at 58.

➤ Open weights models are competitive, but still trail the leaders: GLM-5.1 in Claude Code is the top open-weight result at 53, followed by Kimi K2.6 and DeepSeek V4 Pro in Claude Code at 50. These are strong results, but still meaningfully behind the top proprietary models.

➤ Gemini 3.1 Pro in Gemini CLI underperforms: Gemini 3.1 Pro in Gemini CLI scores 43, well below where Gemini 3.1 Pro sits on our Intelligence Index, highlighting that Gemini’s performance in Gemini CLI remains a relative weak spot for Google’s offering.

➤ Cost per task (API token pricing) varies >30x: Composer 2 in Cursor CLI is cheapest at $0.07/task, followed by DeepSeek V4 Pro in Claude Code at$ 0.35/task and Kimi K2.6 in Claude Code at $0.76/task. At the high end, GPT-5.5 in Codex costs$ 2.21/task, while GLM-5.1 in Claude Code costs $2.26/task. For both models this was contributed to by high token usage, and in GPT-5.5’s case by a relatively higher per token cost.

➤ Token usage varies >3x: GLM-5.1 in Claude Code uses the most tokens at 4.8M/task, followed by Kimi K2.6 at 3.7M/task and DeepSeek V4 Pro at 3.5M/task. GPT-5.5 in Codex uses 2.8M tokens/task, substantially more than Opus 4.7 in Claude Code at 1.7M/task. In GLM-5.1’s case, higher token usage, cost and execution time were partly driven by the model entering loops on some tasks.

➤ Cache hit rates remain high but vary materially: Cache hit rates range from 80% to 96% across combinations. Provider routing, harness prompt structure and cache behavior can materially change the economics of running the same model given cached inputs are typically <50% the API price of regular input tokens.

➤ Time per task varies >7x: Opus 4.7 in Claude Code is fastest at ~6 minutes/task, while Kimi K2.6 in Claude Code is slowest at ~40 minutes/task. This is contributed to by differences in average turns per task, token usage and API serving speed. Opus 4.7 had materially lower amount of turns to complete a task than all other models while Kimi K2.6 had the most.

➤ Cursor made real progress with Composer 2: Composer 2 in Cursor CLI scores 48, near the leading open-weight model results, while being the cheapest combination measured at $0.07/task. Cursor has stated Composer 2 is built from Kimi K2.5, showcasing they have made substantial post-training gains.

This is just the start. We are planning to add additional agents (both harnesses and models). Let us know what you would like to see added next.

See 2 related tweets

@teortaxesTex: > Composer 2 in Cursor CLI is cheapest at $0.07/task, followed by DeepSeek V4 Pro in Claude Code ...
@srush_nlp: RT @ArtificialAnlys: Announcing the Artificial Analysis Coding Agent Index! Our new coding agent ben...

8. wallstengine (Group Score: 101.4 | Individual: 38.3)

Cluster: 4 tweets | Engagement: 549 (Avg: 102) | Type: Tech

CEREBRAS UPSIZES IPO TO $4.8B AFTER 20X DEMAND

$CRBS filed to sell 30M shares at$ 150- $160 each, up from 28M shares at$ 115-$125.

At the top of the new range, the AI chipmaker would raise roughly $4.8B, compared with$ 3.5B under the prior terms.

Reuters says the IPO drew orders for more than 20x the shares available.

The deal is set to price May 13.

Cerebras makes AI inference chips and lists Amazon and OpenAI among customers.

This is the strongest public-market demand signal yet for a pure-play AI chip IPO.

See 3 related tweets

@StockMKTNewz: Cerebras is now expecting to go public between $150 to$ 160 per share above its original $115 to$ 12...
@StockMKTNewz: Polymarket currently says there is a 91% chance that Cerebras ends its first day of trading above a ...
@business: Cerebras increased the size of its initial public offering, now seeking to raise as much as $4.8 bil...

9. business (Group Score: 101.0 | Individual: 54.4)

Cluster: 4 tweets | Engagement: 810 (Avg: 77) | Type: Tech

OpenAI co-founder and former chief scientist Ilya Sutskever said his stake in the ChatGPT maker is worth roughly $7 billion, making him one of the largest individual shareholders in the artificial intelligence startup. https://t.co/CxqZfspKwx

See 3 related tweets

@StockMKTNewz: OpenAI co-founder and former chief scientist Ilya Sutskever said his stake in the ChatGPT maker is w...
@WSJ: OpenAI allowed employees to sell up to $30 million worth of shares each in a recent financing, makin...
@axios: OpenAI launches AI consulting arm valued at $14 billion https://t.co/SQXw2CFJMw...

10. Dagnum_PI (Group Score: 91.4 | Individual: 48.4)

Cluster: 3 tweets | Engagement: 544 (Avg: 66) | Type: Tech

RT @Cointelegraph: 🚨 ALERT: Google says hackers used AI to create a zero-day exploit capable of bypassing multi-factor authentication, per Bloomberg. https://t.co/hFEZL9x18T

See 2 related tweets

@theobearman: GTIG now seeing zero day exploits likely assisted by AI.\n\nQT @AIRiskExplorer: For the first time, ...
@techdaily24: Google stopped a zero-day hack that it says was developed with AI https://t.co/jkGVfvQLzg #TechTrend...

11. cgtwts (Group Score: 89.2 | Individual: 33.3)

Cluster: 4 tweets | Engagement: 1590 (Avg: 360) | Type: Tech

Startups with massive AWS credits right now: https://t.co/vuY0s5qi5J\n\nQT @claudeai: The Claude Platform on AWS is now generally available.

AWS customers get the full set of Claude API features, with AWS authentication, billing, and commitment retirement. https://t.co/9WPrEgCpkg

See 3 related tweets

@mattsgarman: We're the first cloud provider to offer @AnthropicAI 's native Claude Platform directly through cust...
@adocomplete: Claude Platform on AWS!!!\n\nQT @claudeai: The Claude Platform on AWS is now generally available.

A...

@amazon: RT @awscloud: Claude Platform on AWS is now generally available through your AWS account.

@claudea...

12. firstadopter (Group Score: 86.7 | Individual: 33.3)

Cluster: 3 tweets | Engagement: 48 (Avg: 72) | Type: Tech

Low valuations + exponential demand ramp from new secular growth trend -> earnings acceleration + margin expansion -> surging stock prices. Classic Benton/Vinik/Lynch setup.

Lazy “stocks are up too much” mean reversion folks got decimated\n\nQT @firstadopter: AI skeptics going max bearish right into the reasoning AI models exponential compute ramp last year and then again going max bearish into the agentic AI exponential compute ramp this year tells you perhaps those skeptics are actually flying blind and clueless.

Perhaps you shouldn't listen to them on technology fundamentals. Just a thought.

See 2 related tweets

@firstadopter: Remember when the AI bears got excited about TurboQuant? https://t.co/hCD6bi3jLH\n\nQT @firstadopter...
@firstadopter: AI skeptics going max bearish right into the reasoning AI models exponential compute ramp last year ...

13. ivanfioravanti (Group Score: 85.1 | Individual: 30.6)

Cluster: 3 tweets | Engagement: 75 (Avg: 34) | Type: Tech

Hugging Face is another level! 🔥 Use this model on Hermes Agent is here! https://t.co/bUyJRCpbF4\n\nQT @mervenoyann: 🆕 Hugging Face 🤝 Hermes Agent 🔥

we added Hermes Agent to local apps: run it locally with any compatible GGUF/MLX model shipped native traces support for Hermes Agent: visualize your Hermes traces directly on the Hub

Very soon most agents will run locally and we want to accelerate things as much as we can ⚔️

See 2 related tweets

@Teknium: Check out the awesome new Hermes x HuggingFace layers HuggingFace added!\n\nQT @mervenoyann: 🆕 Huggi...
@_akhaliq: RT @mervenoyann: 🆕 Hugging Face 🤝 Hermes Agent 🔥

we added Hermes Agent to local apps: run it loca...

14. lmsysorg (Group Score: 84.4 | Individual: 31.8)

Cluster: 4 tweets | Engagement: 26 (Avg: 35) | Type: Tech

🎉 Meet MiniCPM-V 4.6 from @OpenBMB, a 1.3B edge-friendly multimodal LLM with superior efficiency. Day-0 support is now live in SGLang! ✅ Leading capability: scores 13 on Artificial Analysis Intelligence Index benchmark ✅ Strong multimodal: Matches Qwen3.5 2B-level capacity across 5 major VL benchmarks ✅ Ultra-efficient: 50%+ less visual FLOPs via LLaVA-UHD v4 and mixed 4x/16x visual token compression ✅ Mobile-ready: can be deployed across iOS, Android, and HarmonyOS Cookbook: https://t.co/CbMwdLvYf3 Run it now with SGLang!\n\nQT @OpenBMB: 1/5 MiniCPM-V 4.6 (1.3B) is now live 🚀🚀 High-res visual processing, optimized for consumer-grade and mobile hardware. We’ve leveraged the latest LLaVA-UHD v4 technique to cut vision encoding costs by 55%, enabling native edge deployment with extreme efficiency. 🔥 Beats Gemma4-E2B-it and Qwen3.5-0.8B across key multimodal and Artificial Analysis benchmarks — scoring higher than Qwen3.5-0.8B using just 2.5% of its token budget. ⚡ TTFT (75.7ms) 2.2x Faster than Qwen3.5-0.8B even with 3136² high-res images. 🏗️ ~1.5x Token Throughput compared with Qwen3.5-0.8B on a single RTX 4090. Try the model here: 🤗 Hugging Face: https://t.co/CEkwKMSBwc 💻 GitHub: https://t.co/iYDxpa52tn 🔭 Modelscope: https://t.co/CHflKPLbvK 🌐 Web Demo: https://t.co/DYUrtD0YzM 📱 App Demo: https://t.co/SL7IOhm6zv

See 3 related tweets

@OpenBMB: Huge thanks to the SGLang team for the support 🤩\n\nQT @lmsysorg: 🎉 Meet MiniCPM-V 4.6 from @OpenBMB...
@OpenBMB: Huge thanks to the SGLang team for the support 🤩 this means a lot to us！\n\nQT @lmsysorg: 🎉 Meet Min...
@BanghuaZ: RT @lmsysorg: 🎉 Meet MiniCPM-V 4.6 from @OpenBMB, a 1.3B edge-friendly multimodal LLM with superior ...

15. teortaxesTex (Group Score: 84.0 | Individual: 24.7)

Cluster: 5 tweets | Engagement: 50 (Avg: 64) | Type: Tech

This is of course not evidence that nothing else can be found by a Mythos+. But more importantly, finding just one new vuln is already a big deal – curl is 30 years old, it's benefited enormously from Linus's Law, and it had already been extensively scrutinized by AI tools. https://t.co/65EFPyAl4E\n\nQT @martenmickos: Mythos found one (1) vulnerability in curl - an open-source software product with an installed base of 20 billion instances. https://t.co/JmvvxU8WRb

See 4 related tweets

@teortaxesTex: Pretty compelling argument You don't need Mythos for SoTA vulnerability search and in fact AISLE see...
@moyix: RT @martenmickos: Mythos found one (1) vulnerability in curl - an open-source software product with ...
@Techmeme: curl founder Daniel Stenberg says Mythos identified five vulnerabilities in curl, but a manual revie...
@AndrewCurran_: RT @teortaxesTex: This is of course not evidence that nothing else can be found by a Mythos+. But mo...

16. tether (Group Score: 82.3 | Individual: 46.8)

Cluster: 2 tweets | Engagement: 315 (Avg: 85) | Type: Tech

Build what matters. Get rewarded. ⚡️

Introducing https://t.co/929JWDpTvD. We are providing the resources to scale your vision through community grants and technical bounties.

Whether you are building the next billion-user wallet with @WDK_tether, local-first AI with @QVAC, serverless P2P apps with @Pears_p2p, or mining automation with @MDK_tether - our entire stack is open-source, and we want to back you.

Shape the future of unstoppable systems. Apply today: https://t.co/61ydRJPAF1

See 1 related tweets

@paoloardoino: Build together ♥️\n\nQT @tether: Build what matters. Get rewarded. ⚡️

Introducing https://t.co/929J...

17. 0x_kaize (Group Score: 80.4 | Individual: 33.4)

Cluster: 3 tweets | Engagement: 48 (Avg: 42) | Type: Tech

This is way bigger than people realize

Anthropic is moving towards becoming a better text-to-speech model

They just turned Claude's constitution into an audiobook, read by the actual philosophers who wrote it. Sounds like a podcast drop but here's what's really happening:

Amanda Askell left OpenAI because they weren't taking safety seriously enough and made the Time 100 AI list Wall Street Journal said her job is to 'teach Claude how to be good' Constitution is 20,000 words that directly shape every Claude response

this isn't marketing. it's a 20,000-word document that directly shapes every response Claude gives you

no other AI company has published anything close to this

OpenAI ships first, explains later. Anthropic publishes the rulebook and lets you listen to the people who wrote it\n\nQT @AnthropicAI: Claude's Constitution is now an audiobook, read by two of its authors, Amanda Askell and Joe Carlsmith.

It includes a Q&A on the writing process, the philosophies that shaped the document, and how it might change as models become more capable.

Listen at https://t.co/dKMfpeOblm https://t.co/792RQ4sAxc

See 2 related tweets

@MTSlive: .@theojaffee breaks down the lore behind the authors of the Claude Constitution:
Amanda Askel...
@0x_kaize: RT @0x_kaize: This is way bigger than people realize

Anthropic is moving towards becoming a better ...

18. awnihannun (Group Score: 79.7 | Individual: 26.4)

Cluster: 5 tweets | Engagement: 129 (Avg: 45) | Type: Tech

Claude Code agent view is where I start and manage a lot of my work. This is an exceedingly useful new feature - 10/10\n\nQT @claudeai: New in Claude Code: agent view.

One list of all your sessions, available today as a research preview. https://t.co/NnbsAQjSPW

See 4 related tweets

@Prince_Canuma: I love this!

Guess I'm going back to full on terminal mode\n\nQT @claudeai: New in Claude Code: age...

@Shashikant86: Claude Code Agent view is cool. But still getting used to new normal that X post from @awnihannun i...
@nummanali: Not gonna lie, Agent View is slick\n\nQT @adocomplete: So excited for Agent view (research preview) ...
@Prince_Canuma: RT @awnihannun: Claude Code agent view is where I start and manage a lot of my work. This is an exce...

19. MarioNawfal (Group Score: 79.5 | Individual: 40.2)

Cluster: 3 tweets | Engagement: 1133 (Avg: 520) | Type: Tech

🇺🇸 Microsoft's CEO just took the stand in Elon's OpenAI lawsuit.

Satya Nadella took the stand today, the third billionaire to testify in weeks, after Elon and Brockman. Altman is up later this week. Closing arguments are Thursday.

Elon's case is straightforward. He says OpenAI abandoned its nonprofit mission and that Altman and Brockman enriched themselves at the expense of what was supposed to be a charity. He donated $38 million on that basis. Microsoft, he says, helped make it happen.

Nadella on the stand said Microsoft had no idea about any conditions Elon placed on his donations and that OpenAI kept its independence throughout.

"OpenAI had all the rights and resources they always had."

Microsoft put in $1 billion in 2019,$ 2 billion in 2021, then $10 billion after ChatGPT launched. Elon flagged it publicly in 2020, saying OpenAI had been "essentially captured by Microsoft."

Thursday is when it all lands.

Source: NBC NEWS\n\nQT @MarioNawfal: 🇺🇸 Elon's legal team just put OpenAI's second-in-command on the stand.

Greg Brockman, OpenAI's President, has been subpoenaed to testify in the trial against Sam Altman, meaning he has something damaging to say.

The core of Elon’s case: OpenAI abandoned its non-profit mission the moment Microsoft money came in. Brockman was there for all of it: the founding, the pivot, the transformation into a for-profit empire.

This testimony could get uncomfortable fast.

Source: Bloomberg, Business Insider

See 2 related tweets

@MarioNawfal: RT @MarioNawfal: 🇺🇸 Microsoft's CEO just took the stand in Elon's OpenAI lawsuit.

Satya Nadella too...

@techreview: RT @michelletomkim: Microsoft CEO Satya Nadella takes the stand in a black suit and sky blue tie. Si...

20. chandrarsrikant (Group Score: 75.6 | Individual: 35.5)

Cluster: 3 tweets | Engagement: 88 (Avg: 958) | Type: Tech

India's semiconductor startup ecosystem is inching closer to a key milestone, with Chennai-based Mindgrove Technologies preparing to begin the first commercial production of its Secure IoT chip, founder and CEO Shashwath T R told Moneycontrol.

Founded in 2021 by Shashwath T R and Sharan Srinivas J, the IIT Madras-incubated fabless semiconductor startup is in the final stages before the commercial rollout of its RISC-V-based microcontroller system-on-chip (SoC), which targets applications including secure biometrics, smart meters, industrial IoT devices, wearables, access-control systems and EV battery management systems.

Mindgrove has positioned the Secure IoT chip as India's first commercial-grade high-performance microcontroller SoC built on the open-source RISC-V architecture. The company claims the chip can lower overall device costs by nearly 30 percent compared to imported alternatives.

Commercial launch nearing after prototype deployments

“We are right on the cusp of actually having it commercially available for everybody to buy,” Shashwath said, adding that several customers are already testing prototype versions of the chip.\n\nQT @chandrarsrikant: IIT Madras incubated Mindgrove to begin commercial production of India's Secure IoT chip, eyes biometrics and industrial IoT market

By @BhavyaDKumar

https://t.co/diHioggaBs

See 2 related tweets

@chandrarsrikant: IIT Madras incubated Mindgrove to begin commercial production of India's Secure IoT chip, eyes biome...
@moneycontrolcom: 🚨 Mindgrove to begin commercial production of India’s Secure IoT chip, eyes biometrics and industria...