- Published on
科技热门推文——2026年7月17日
- Authors
- Name
- geeknotes
今日科技动态：人工智能竞争进一步加剧。月之暗面发布超大规模多模态模型 Kimi K3，并迅速登顶一项前端编程基准测试；与此同时，据报道，谷歌因未达内部目标而推迟了 Gemini 的发布。日本成立了由 44 家企业组成、聚焦机器人领域的主权人工智能联盟，谷歌云则进一步扩大了与英特尔的合作。此外，多家人工智能硬件初创公司获得新一轮融资，Grok Build 宣布开源；林纳斯·托瓦兹为在 Linux 开发中务实使用人工智能辩护；印度也正准备发射其首枚由私营企业自主研发的轨道级火箭。
1. perrymetzger (Group Score: 565.0 | Individual: 59.4)
Cluster: 17 tweets | Engagement: 4048 (Avg: 300) | Type: Tech
RT @Kimi_Moonshot: Introducing Kimi K3: Open Frontier Intelligence
🔹 2.8 Trillion Parameters, 1 Million Context, Native Multimodal 🔹 Kimi Delta Attention enables up to 6.3x faster decoding in million-token contexts 🔹 Attention Residuals deliver ~25% higher training efficiency at <2% additional cost 🔹 Built for long-horizon agentic coding and self-evolving workflows
Kimi K3 is now live on on https://t.co/zrk6zZxZUo, Kimi Work, Kimi Code, and the Kimi API. Open Weights by July 27, 2026.
🔗 API: https://t.co/XCrgjXAqMw 🔗 Tech blog: https://t.co/YTfiMSNM1f
See 16 related tweets
- @levie: It’s truly wild that we’re getting this level of performance from open models. Congrats to Kimi team...
- @AravSrinivas: Incredible\n\nQT @Kimi_Moonshot: Introducing Kimi K3: Open Frontier Intelligence
🔹 2.8 Trillion Par...
- @EnoReyes: Watershed moment. These open weights will likely mark the beginning of a sovereign intelligence expl...
- @scaling01: The moonshot was successful
the small chinese lab is taking down trillion dollar companies\n\nQT @K...
- @synthwavedd: this is insane btw https://t.co/bbHWc14pP8\n\nQT @Kimi_Moonshot: Introducing Kimi K3: Open Frontier ...
2. yacineMTB (Group Score: 486.6 | Individual: 54.5)
Cluster: 16 tweets | Engagement: 2406 (Avg: 256) | Type: Tech
insane. they distilled american frontier models that haven't even been invented yet. damn chinese\n\nQT @arena: Big news: Kimi-K3 by @Kimi_Moonshot is now #1 in the Frontend Code Arena with 1679 pts, surpassing Claude Fable 5.
This is a 17-place jump from Kimi-k2.6 (#18 -> #1).
In Frontend, Kimi-K3 ranked #1 in 6 of 7 domains: Brand & Marketing, Reference-Based Design, Data & Analytics, Consumer Product, Simulations, and Content Creation Tools, landing #2 only in Gaming behind Fable 5.
The full model weights will be released by July 27.
Congrats to the @Kimi_Moonshot team on this major milestone!
See 15 related tweets
- @ml_angelopoulos: This may be the single biggest release of the year, and marks the moment that OSS Chinese models hav...
- @MTSlive: SITUATION EXPLAINED: Kimi K3 just beat Fable on a major coding benchmark, at a fraction of the price...
- @jason: It’s happening folks… things accelerated more in the last 30 days from a dozen players than in the l...
- @MTSlive: SITUATION EXPLAINED: Kimi K3 is out. It's a massive open model, priced like Sonnet, and reportedly b...
- @synthwavedd: 🚨 Per Financial Times, Moonshot intend to launch Kimi K3 as early as tonight. K3 is 2-3T total param...
3. Scobleizer (Group Score: 407.0 | Individual: 35.8)
Cluster: 13 tweets | Engagement: 658 (Avg: 201) | Type: Tech
RT @lazyapoorv: We raised $5.5M to build an AI hardware interface that knows what you want.
The world has changed: we talk to our devices, AI writes our emails, and self-driving cars pick us up.
But how we interact with it all, touchscreens and keyboards, was designed for an age of browsing and searching.
When we do hundreds of tasks a day, now even more with AI, every unnecessary decision, every step spent navigating instead of acting adds up to real cognitive load.
We are building the Hardware Interface, designed for the age of AI.
Private Pilot now open.
Limited spots.
See 12 related tweets
- @kimmonismus: Everyone became an AI hardware skeptic the day Humane and Rabbit went down.
Fair enough. Both of th...
- @dr_cintas: This stealth AI device just raised $5.5M without a single product shot.
It’s called Aina, from the ...
- @alex_prompter: You're prompting frontier AI through the same keyboard layout you used for AOL Instant Messenger.
T...
- @Scobleizer: He makes the little keyboard that was ahead of OpenAI's announcement yesterday of a keyboard that do...
- @rohanpaul_ai: OpenAI and Worklouder was in the news yesterday for their new hardware and today its Aina.
AI is mo...
4. MTSlive (Group Score: 326.9 | Individual: 34.9)
Cluster: 17 tweets | Engagement: 58 (Avg: 72) | Type: Tech
SITUATION EXPLAINED: Japan just formed a 44-company AI consortium aimed squarely at robotics.
• Noetra is a new Japanese sovereign AI consortium anchored by Sony, SoftBank, NEC, and Honda, with 44 companies and organizations investing in total, spanning manufacturing, finance, and logistics • Backed by $6.1 billion in Japanese government funding over five years • Its data center is slated to go online in two years, with a reported order of 27,500 Nvidia Rubin GPUs, small next to Microsoft's orders, but notable for a single national consortium • The target: 10 million AI-equipped robots deployed across the country by 2040 • Plans to release a first foundation model by March 2027, purpose-built for robotics and physical AI applications • Minister Ryosei Akazawa wore a leather jacket and black T-shirt to the announcement, visibly echoing Jensen Huang's signature look • The reason for the robotics focus: Japan's median age is 50.2, meaning half the country's population is over 50
@schisofrenia: "They need robots 'cause they don't have population."
@theojaffee: "It's really a problem. It's gonna be a problem for everywhere in the world, but it's gonna be a problem for Japan first."\n\nQT @MTSlive: SITUATION DETECTED: Noetra, a new Japanese government-backed venture, will buy 27,500 Nvidia Rubin chips to build a homegrown foundational AI model for robots, per Bloomberg.
Allocated $2.4 billion in public funds, it will build a 140 MW data center going online in June 2028.
See 16 related tweets
- @StockMKTNewz: JAPAN 🇯🇵 IS BUILDING A NATIONAL AI FACTORY WITH NVIDIA THAT DIDN'T EXIST ANYWHERE ELSE IN THE WORLD ...
- @wallstengine: JAPAN IS GETTING A 140MW NVIDIA VERA RUBIN AI FACTORY.
$NVDA is partnering with Noetra, backed by J...
- @NVIDIAAIInfra: RT @nvidianewsroom: Japan is building the AI factories to power the next industrial revolution.
NVI...
- @nvidianewsroom: Japan is building the AI factories that will power the next industrial revolution.
NVIDIA CEO Jense...
- @StockSavvyShay: $NVDA is partnering with Japan’s Noetra to build a 140MW AI factory powered by 13,750 Vera CPUs &...
5. MTSlive (Group Score: 306.2 | Individual: 36.2)
Cluster: 15 tweets | Engagement: 246 (Avg: 72) | Type: Tech
SITUATION BREWING: Moonshot is set to release Kimi K3 as early as tonight, expected to be China’s largest AI model to date at 2 to 3 trillion parameters, per FT.
The open-weight model is expected to outperform Claude Opus 4.8 on key benchmarks.
See 14 related tweets
- @scaling01: Kimi-K3 scores 57 on AAI Index, while being almost 2x more token-efficient than Opus 4.8\n\nQT @Arti...
- @ruima: RT @zijing_wu: Chinese AI start-up Moonshot to launch model challenging Anthropic’s lead
- Set to re...
- @WhaleInsider: JUST IN: 🇨🇳 Chinese AI start-up Moonshot to launch model challenging Anthropic’s lead. kimi K3 expec...
- @Techmeme: Moonshot AI releases Kimi K3, a 2.8T-parameter AI model that it says rivals Opus 4.8 and GPT 5.5, an...
- @Techmeme: Sources: Moonshot plans to launch Kimi K3, China's largest model to date with 2T-3T parameters, in t...
6. bensig (Group Score: 237.8 | Individual: 52.6)
Cluster: 8 tweets | Engagement: 4298 (Avg: 505) | Type: Tech
RT @SpaceXAI: We've open-sourced Grok Build and have reset usage limits for all users.
Open sourcing Grok Build allows anyone to support making a reliable and robust harness. Check out our code, including the Git repo for the Grok Build CLI.
See 7 related tweets
- @VaibhavSisinty: SpaceXAI just open-sourced Grok Build their AI coding agent that writes code, edits files, and runs ...
- @teslaownersSV: WokeGPT could never. https://t.co/pDprKqH7ky\n\nQT @SpaceXAI: We've open-sourced Grok Build and have...
- @BrianRoemmele: BOOM!
@Grok Build is now Open Source.
This is a major move.
Thank you SpaceXAI and @elonmusk!
Ha...
- @Gorden_Sun: Grok Build开源，同时再次重置了额度\n\nQT @SpaceXAI: We've open-sourced Grok Build and have reset usage limits fo...
- @0xSero: Damn brother Grok Build is open source.
I am again extending an invite onto my podcast!\n\nQT @Spa...
7. PolymarketMoney (Group Score: 228.2 | Individual: 36.2)
Cluster: 10 tweets | Engagement: 1335 (Avg: 251) | Type: Tech
JUST IN: Google delays the Gemini launch as the technology falls short of internal targets.
See 9 related tweets
- @negligible_cap: GEMINI LAUNCH DELAYED AS TECH FALLS SHORT OF GOALS
$GOOG said to be months behind schedule on Gemin...
- @FirstSquawk: GOOGLE DELAYS FLAGSHIP GEMINI 3.5 PRO AI MODEL LAUNCH BY MONTHS AS EFFORTS TO IMPROVE CODING CAPABIL...
- @StockSavvyShay: $GOOGL delays the Gemini launch after the technology falls short of internal targets. https://t.co/m...
- @StockMKTNewz: Google $GOOGL rumor coming out right now
GOOGLE GEMINI LAUNCH DELAYED AS TECH FALLS SHORT OF GOALS...
- @wallstengine: GOOGLE $GOOGL GEMINI LAUNCH DELAYED AS TECH FALLS SHORT OF GOALS https://t.co/xt42HmAzZF...
8. ANI (Group Score: 221.9 | Individual: 41.7)
Cluster: 7 tweets | Engagement: 336 (Avg: 80) | Type: Tech
#WATCH | Andhra Pradesh: Ahead of the launch of India's historic first privately developed orbital-class rocket, Vikram-1, CISF personnel establish a strict security blanket in and around the launch pad at Satish Dhawan Space Centre, Sriharikota. https://t.co/FrUe68Zr15\n\nQT @ANI: #WATCH | Andhra Pradesh: India's first privately developed orbital-class rocket, Vikram-1, is scheduled to launch for its maiden test flight, "Mission Aagaman," on July 18 at 11:30 AM from Satish Dhawan Space Centre, Sriharikota.
Developed by Hyderabad-based Skyroot Aerospace, the 24-meter carbon-composite rocket will independently deploy up to 350 kg of payloads into a 450 km Low Earth Orbit (LEO). Payloads include a lab-grown "Diamond Lotus" by Cosmos Diamonds.
Visuals show the Vikram-1 rocket stationed at the First Launch Pad and ISRO and Skyroot scientists carrying out final preparations.
See 6 related tweets
- @ANI: #WATCH | Sriharikota, Andhra Pradesh: Pawan Kumar Chandan, CEO and Founder of Skyroot Aerospace, say...
- @ANI: #WATCH | Sriharikota, Andhra Pradesh: Naga Bharath Daka, COO and Co-founder of Skyroot Aerospace, sa...
- @ANI: #WATCH | Sriharikota, Andhra Pradesh: Director Technical, IN-SPACE, Rajeev Jyoti says, "...We are se...
- @ANI: #WATCH | Andhra Pradesh: India's first privately developed orbital-class rocket, Vikram-1, is schedu...
- @PawanKChandana: 🚀🎥 Every rocket has a story before it has a launch.
Watch this 15-minute documentary on the journey...
9. _simonsmith (Group Score: 197.6 | Individual: 33.4)
Cluster: 7 tweets | Engagement: 21 (Avg: 26) | Type: Tech
This is extraordinary. Kimi K3 has better than Opus 4.8 performance and is multimodal. But it costs roughly the same as GPT-5.6 Sol on the Artificial Intelligence index, and GPT-5.6 is better at most things, though seemingly not frontend design based on other comparisons.
Possible implications:
Chinese labs are ahead of where many people thought they were. This can't all be from distillation, for example, because it's achieving near-Fable performance despite Fable simply not being available that long.
It's less about open weights right now, since this isn't yet, and more about how good Chinese models are relative to the majority of US labs with the exception of Anthropic and OpenAI.
I expect this will push OpenAI and Anthropic to be even more aggressive at releasing their next models, which for OpenAI is expected to be a big upgrade with GPT-6. This, in turn, will require better coordination with the US government to avoid delays.
How much longer will the US government allow powerful models like this to be distributed with no controls? What can they do? Probably control US-based inference, but this model is even open weights right now, and I doubt they'd take the step of blocking Chinese endpoints, nor that this would be fully successful.\n\nQT @ArtificialAnlys: Kimi K3 scores 57 on the Artificial Analysis Intelligence Index. Its intelligence is comparable to Opus 4.8 and GPT-5.5 but remains behind Fable 5 and GPT-5.6 Sol. Moonshot AI has expressed plans to release the 2.8T parameter model's weights, which would make it the leading open weights model
Key results:
➤ Strong agentic task performance: @Kimi_Moonshot's Kimi K3 reaches an Elo rating of 1668 on GDPval v2. This is a marked improvement over K2.6’s 1190, surpassing GLM-5.2 (1514), GPT-5.5 (1494), and Claude Opus 4.8 (1600). However, it still lags behind Claude Fable 5 (1760). Kimi K3 also scores an impressive 53% and takes the #1 position on AutomationBench-AA, our implementation of Zapier’s Agentic SaaS workflow evaluation.
➤ Second-highest performance on AA-Briefcase (agentic knowledge work): On our private long-horizon knowledge work evaluation, Kimi K3 reaches an overall Elo of 1547, +732 points from Kimi K2.6 and behind only Claude Fable 5. It is well-rounded: its rubric scoring and analytical quality almost reach Claude Fable 5’s scores, while GPT-5.6 Sol continues to outperform other leading models on presentation quality.
➤ Set to lead open weights models once weights are released: Moonshot AI has not yet released the weights but expressed plans to do so. Once available, Kimi K3 would clearly lead other open weights models including GLM-5.2 (51) and DeepSeek v4 Pro (44). However, at 2.8T parameters, it is significantly larger than its open weights peers (eg. GLM-5.2 at 753B params and DeepSeek V4 Pro at 1.6T), as well as the Kimi K2 to K2.6 models (1T params).
➤ Cost per task (1.04), ~1/2 the price of Opus 4.8 (15/1M tokens while K2.6 was 1.04) and more expensive than open weights peers, GLM-5.2 (0.04)
➤ Improved token efficiency alongside higher intelligence: Kimi K3’s token usage on the Artificial Analysis Intelligence Index decreased significantly, using 21% fewer output tokens than K2.6. The new model used approximately 132M output tokens to complete all nine evaluations, compared to approximately 166M for K2.6, while achieving higher scores.
➤ Native multimodal capabilities: Kimi K3, like K2.6, is released with native image and text multimodal input. If weights are released, this will position Kimi K3 as one of the leading open weights models with multimodal input capabilities
Other model details:
Context window: 1M
Size: 2.8T total parameters
Pricing: The first-party API is priced at 15.00 per 1M input/output tokens, with cached input discounted 90% to $0.30 per 1M tokens.
Modality: Native multimodal input supports text and images, and the model remains text-only for output.
Accessibility: Accessible at launch through Moonshot’s first party API. Model weights are not yet released but Moonshot AI has expressed plans to do so.
See 6 related tweets
- @TeksEdge: More Holy 💩!! Grok and Muse mogged by Kimi K3. https://t.co/bZwtWuI4DC\n\nQT @ArtificialAnlys: Kimi ...
- @teortaxesTex: Five fucking seven I have no words https://t.co/kUPvXPuGsi\n\nQT @teortaxesTex: pretty good video. r...
- @petergostev: Funny this was my #1 prediction at the beginning of the year, let's see how this holds
(others are...
- @VaibhavSisinty: Kimi K3 just launched. China's biggest open-weights model yet. And the numbers are genuinely wild. 🤯...
- @VaibhavSisinty: Kimi K3 benchmarks just dropped. It's ranked number one on BrowseComp, Automation Bench, and Spreads...
10. burkov (Group Score: 192.2 | Individual: 40.5)
Cluster: 6 tweets | Engagement: 173 (Avg: 52) | Type: Tech
Thinking Machines released a model that is worse than everyone in everything. It only scores high in size—nearly 1T parameters—which is impressive in the sense of how the hell 1T parameters can suck so big time?
@thinkymachines, do you have a technical report to laugh at? https://t.co/0W8gnxuS4P\n\nQT @thinkymachines: Today, we are introducing Inkling.
Inkling reasons efficiently across text, image, and audio modalities. We are making the full weights available.
Available today for fine-tuning on Tinker. Play with it in the Inkling Playground. 🧵
See 5 related tweets
- @MLStreetTalk: RT @thinkymachines: Today, we are introducing Inkling.
Inkling reasons efficiently across text, ima...
- @quxiaoyin: This is exciting. The USA open weight era has arrived\n\nQT @thinkymachines: Today, we are introduci...
- @shensi: The US needed a competitive open source model - can’t wait to try it out\n\nQT @thinkymachines: Toda...
- @deliprao: Always excited to see a native open model. Congratulations to everyone on the thinky team!\n\nQT @th...
- @curiouswavefn: Very promising!\n\nQT @thinkymachines: Today, we are introducing Inkling.
Inkling reasons efficient...
11. shiri_shh (Group Score: 180.6 | Individual: 39.9)
Cluster: 7 tweets | Engagement: 812 (Avg: 343) | Type: Tech
If Anthropic designed this vibe keyboard https://t.co/O4n5Sw6olT\n\nQT @kimmonismus: OpenAI reveals Codex Micro (their first hardware product, so to say)
OpenAI’s $230 Codex Micro is a compact control deck for agentic coding, with RGB status keys, shortcuts for common Codex actions, and a dial for adjusting reasoning effort.
Built with Work Louder, it works on Mac and Windows and is designed to make managing multiple agents feel faster, more tactile, and less dependent on constantly switching between chats.
It’s a bit gimmicky, but one thing is clear: OpenAI works within and with the community, developing cool products that are fun and capture the zeitgeist.
See 6 related tweets
- @kimmonismus: Aaand its gone. Already out of stock. Really smart and gimmicky release. People seem to love it. Kud...
- @minchoi: OpenAI just launched a $230 command center for AI agents.
Agent status lights. A joystick for w...
- @thehypedotnews: openai has released its first hardware, the codex micro, a $230 macro keypad made with work louder t...
- @minchoi: RT @minchoi: OpenAI just launched a $230 command center for AI agents.
Agent status lights. A j...
- @moneycontrolcom: #TechWithMC | OpenAI's first hardware is Codex Micro, a keyboard for managing AI agents
Read More ⤵...
12. reach_vb (Group Score: 162.7 | Individual: 28.8)
Cluster: 8 tweets | Engagement: 385 (Avg: 463) | Type: Tech
PSA: Don’t run you coding agent in full access mode!
Use the combination of Rules, Sandbox, Approve for me & Hooks to make sure that both you and your agent can work in peace.
In the extremely rare case the model does trip - you want to have a layer of protection between your data and the agent.\n\nQT @thsottiaux: On file deletions. We’ve investigated a handful of reports where GPT-5.6 unexpectedly deleted files.
What we have found is that this most commonly occurs when:
- Full access mode is enabled and codex is run without sandboxing protections, including without auto review being enabled
- The model attempts to override the $HOME env var to define a temporary directory.
- The model makes an honest mistake and mistakenly deletes $HOME instead.
This is of course not how we want the system to behave, even when a user operates the model in full-access mode without the safeguards of our sandbox or without using auto review which checks for these kinds of high risk actions and rejects them.
We are taking steps to mitigate this risk including by updating the developer message, guiding more users towards safer permission modes, and adding additional harness safeguards. Even though this happens extremely rarely, we’ll share a detailed post-mortem in the coming days that goes into more details and what we are doing to minimize risks further.
See 7 related tweets
- @Michaelzsguo: Codex and @thsottiaux have shared an update on the recent Codex file deletion incident.
The gist: t...
- @dkundel: RT @thsottiaux: On file deletions. We’ve investigated a handful of reports where GPT-5.6 unexpectedl...
- @VaibhavSisinty: Read this carefully if you use any AI coding agent.
Three things to do right now:
→ Turn on sandbo...
- @antirez: Appreciate the transparency but in many use cases the only sand choice is full permissions.\n\nQT @...
- @PaulSolt: GPT-5.6 Sol is NOT safe with "Full Access"
OpenAI recommends you use "Approve for Me" to prevent ag...
13. NewsFromGoogle (Group Score: 146.4 | Individual: 31.9)
Cluster: 7 tweets | Engagement: 32 (Avg: 71) | Type: Tech
Today @GoogleCloud and @Intel announced a new expanded partnership, which will accelerate Intel’s enterprise-wide digital evolution through the deployment of Gemini Enterprise and Google Cloud. Learn more from @KarthikSNarain ⬇️\n\nQT @KarthikSNarain: Excited to announce our expanded partnership with @intel to scale Gemini Enterprise globally, putting advanced AI capabilities directly into the daily workflows of their entire workforce. https://t.co/Q3tZbV5G4K
See 6 related tweets
- @intel: Today we're expanding our collaboration with @googlecloud, the latest milestone in a long history of...
- @sundarpichai: Great to see @Intel using Gemini Enterprise across its business, including to speed up the developme...
- @StockMKTNewz: Intel GOOGL today announced an expansion of their previously disclosed multi...
- @GoogleCloud: Announced today: We're expanding our multi-year strategic collaboration with @intel to accelerate In...
- @StockSavvyShay: GOOGL expand their strategic partnership to deploy Gemini Enterprise across Intel and acc...
14. myfear (Group Score: 143.7 | Individual: 49.2)
Cluster: 4 tweets | Engagement: 606 (Avg: 79) | Type: Tech
RT @pashmerepat: Absolutely beautiful rant about AI in Linux Kernel from Linus yesterday:
I realize that some people really dislike AI, but this is an area where I'm willing to absolutely put my foot down as the top-level maintainer.
Linux is not one of those anti-AI projects, and if somebody has issues with that, they can do the open-source thing and fork it.
Or just walk away.
AI is a tool, just like other tools we use. And it's clearly a useful one.
It may not have been that "clearly" even just a year ago, but it's no longer in question today.
There are other questions around AI (like what the economy of it will actually look like in the end), but "is it useful" is no longer one of those questions. Anybody who doubts that clearly hasn't actually used it.
Yes, it can also be a somewhat painful tool, both for maintainer workloads and just from a "it keeps finding embarrassing bugs" standpoint.
But the solution is not to put your head in the sand and sing "La La La, I can't hear you" at the top of your voice like some people seem to do.
The solution is to make sure those LLM tools help maintainers instead of just causing them pain. There's no question on that side.
We're not forcing anybody to use it, but I will very loudly ignore people who try to argue against other people from using it.
And no, AI isn't perfect. But Christ, anybody who points to the problems at AI had better be looking in the mirror and pointing at themselves at the same time.
Because it's not like natural intelligence is always all that great either.
The kernel project has been and will continue to be about the technology.
Sure, the social angle of working on open source is important and often a very motivating part of the project, but in the end that's a side benefit, not the point of the project.
This is NOT some kind of "social warrior" project, never has been, and never will be.
In the kernel community we do open source because it results in better technology, not because of religious reasons.
And so we make decisions primarily based on technical merit. Not fear of new tools.
Linus
See 3 related tweets
- @theo: As always, Linus is
- very clear
- thoughtful
- realistic
- based as hell\n\nQT @pashmerepat: Absolu...
- @VictorTaelin: You've all seen this already, but I do feel vindicated.
I know this seems obvious today, but it was...
- @penberg: > Linux is not one of those anti-AI projects, and if somebody has issues with that, they can do ...
15. no_stp_on_snek (Group Score: 142.0 | Individual: 56.4)
Cluster: 5 tweets | Engagement: 542 (Avg: 54) | Type: Tech
RT @IntCyberDigest: ‼️ Linus Torvalds told Linux kernel developers this week that Linux is not one of those anti-AI projects and that AI is now a clearly useful tool.
He says AI's usefulness was still arguable a year ago but no longer is, and that the doubters simply haven't used the tools. Nobody is forced to use AI, but he will ignore anyone trying to stop others from using it. Anyone who wants an anti-AI project, he says, can fork the code or walk away.
In 2024, Torvalds dismissed AI as mostly marketing. By this January he was shipping AI-generated Python in his own hobby repo, and the kernel now accepts AI-written code under disclosure rules.
See 4 related tweets
- @mark_k: Linus Torvalds has put his foot down regarding the use of AI for coding and developing the Linux ker...
- @Pirat_Nation: Linus Torvalds has made his stance on AI pretty clear: the Linux kernel isn’t against AI.
During a ...
- @Mayhem4Markets: Linus Torvalds, who created the Linux kernel that now is a big part of hosting the world's digital i...
- @arstechnica: Linus Torvalds to critics of AI coding in Linux: "Fork it. Or just walk away." https://t.co/ZrGbNJEi...
16. VaibhavSisinty (Group Score: 139.8 | Individual: 37.2)
Cluster: 7 tweets | Engagement: 47 (Avg: 78) | Type: Tech
DoorDash's co-founder just shipped a command line tool that lets you order food from your terminal.
It is called dd-cli.
You type a command and food shows up at your door.
Here is what it can do: → Search nearby restaurants and stores → Compare prices and find the best deals → Add items to cart and check out → All from your terminal on macOS
But here is the real play.
It is not really built for you. It is built for your AI agent.
An AI assistant can now search DoorDash, compare three restaurants, pick the cheapest healthy lunch, and place the order. No screen. No tapping.
One of the first times a major consumer company has opened its infrastructure specifically for AI agents to spend real money on your behalf. Currently in beta. US and Canada only. Waitlist to get in.
This is what agentic commerce actually looks like. Not a demo. A live product with a real checkout.\n\nQT @andyfang: Today we're opening up the DoorDash CLI in limited beta.
dd-cli lets you order DoorDash directly from your agent: search stores, find the best deals, check out, and more.
Early access for US/Canadian macOS developers by waitlist. Excited to see what folks build! https://t.co/rSFhjJnvjJ
See 6 related tweets
- @dani_avila7: Now I never have to leave Claude to order my Chipotle burrito
My skills: /door-dash-chipotle-burrit...
- @jxnlco: When you instrument your coding agent with dd and use DoorDash and not data dog\n\nQT @andyfang: Tod...
- @starbuxman: can’t wait to orchestrate deliveries from a Spring AI agent\n\nQT @andyfang: Today we're opening up ...
- @ibuildthecloud: Claude, I need food.\n\nQT @andyfang: Today we're opening up the DoorDash CLI in limited beta.
`dd-...
- @Techmeme: DoorDash launches a limited beta of DoorDash CLI, which lets users place orders via an AI agent, ava...
17. cryptopunk7213 (Group Score: 120.4 | Individual: 52.1)
Cluster: 3 tweets | Engagement: 381 (Avg: 79) | Type: Tech
what an absolute huge win for open source AI in america. 1st ever model from OpenAI's ex CTO and its a 1T open model that reasons across text, code, images, video - but the best part isn't even how good it is...
its how you can fine-tune it to create something way better:
you can adjust the thinking effort
you can influence the way it approaches a coding problem
you can perfect token spending so your task costs pennies
this solves a big problem that companies are facing now: they're spending 10Ms when they could just be fine-tuning a "less intelligent" model to do the same task at a fraction of the cost (and often better)
thats why bridgewater's using Thinking Machine's tinker product
i expect to see a lot of folks playing around with this model, finally an alternative to chinese open source!\n\nQT @thinkymachines: Today, we are introducing Inkling.
Inkling reasons efficiently across text, image, and audio modalities. We are making the full weights available.
Available today for fine-tuning on Tinker. Play with it in the Inkling Playground. 🧵
See 2 related tweets
- @jason: Open Source, open weights… what OpenAI was supposed to be
@miramurati 👏
Great to see so much inn...
- @omarsar0: Open source AI having a moment indeed!\n\nQT @sriramk: It is clear open source models and harnesses ...
18. jxnlco (Group Score: 115.3 | Individual: 48.5)
Cluster: 3 tweets | Engagement: 1214 (Avg: 199) | Type: Tech
RT @OpenAI: Introducing GPT-Red
An internal automated red teamer on a mission to find our models’ prompt injection vulnerabilities at scale, helping us build stronger defenses before wider deployment.
See 2 related tweets
- @theobearman: This is very cool research, and I am genuinely glad to see how GPT-Red is being used to make their f...
- @theobearman: OpenAI built an automated red teaming model to help its mainline models become more resistant to pro...
19. eng_khairallah1 (Group Score: 111.9 | Individual: 37.8)
Cluster: 4 tweets | Engagement: 185 (Avg: 60) | Type: Tech
Google just dropped a 1-hour course on agentic engineering from scratch:
00:00 – How to build your first AI agent 08:24 – Build agent memory (short, persistent, long) 28:34 – Agentic loops, long-running AI agents 40:04 – How to build MCP (MCP vs API) 1:00:22 – Multi-agentic systems
This 1-hour watch will replace 10 paid agentic courses on the internet.
Bookmark this. Watch this weekend.\n\nQT @eng_khairallah1: https://t.co/mtiWDoKy6m
See 3 related tweets
- @0xMovez: Andrew Ng:
"The most important trend in AI is agentic AI."
in 30-minutes talk he breaks down the a...
- @Av1dlive: RT @Av1dlive: Met a guy making $1.2 million a year as an AI Engineer
I asked him how he learned to ...
- @eng_khairallah1: RT @eng_khairallah1: Google just dropped a 1-hour course on agentic engineering from scratch:
00:00...
20. jason (Group Score: 110.7 | Individual: 40.0)
Cluster: 5 tweets | Engagement: 947 (Avg: 304) | Type: Tech
Absolutely fantastic moment happening in tech right now…
… open source enthusiasm combined with a dramatic drop in the cost of intelligence combined with personal hardware capable of running models locally is the emergence of the AI sovereignty stack 🫡💪❤️\n\nQT @SpaceXAI: We've open-sourced Grok Build and have reset usage limits for all users.
Open sourcing Grok Build allows anyone to support making a reliable and robust harness. Check out our code, including the Git repo for the Grok Build CLI.
See 4 related tweets
- @teslaownersSV: GROK BUILD IS NOW OPEN SOURCE
xAI has open-sourced Grok Build, making the full codebase publicly av...
- @Kyrannio: This is awesome.\n\nQT @tetsuoai: I used Grok 4.5 in Grok Build to create a game engine for agentic ...
- @ClementDelangue: RT @Jason: Absolutely fantastic moment happening in tech right now…
… open source enthusiasm combi...
- @BrianRoemmele: RT @BrianRoemmele: BOOM!
@Grok Build is now Open Source.
This is a major move.
Thank you SpaceXAI...