AI/ML news, top picks, and generated innovation digests.
200 articles tagged with this keyword, sorted by most recent first.
xAI's Grok 4.7 is now available on Amazon Bedrock: a frontier model for coding, long-running agents, and knowledge work. It offers a 500K token context window and four configurable reasoning effort levels, reachable through the Responses, Chat Completions, and Converse APIs.
Grok Bot gets a shared, team-native mode with plugins, credentials, and per-user memory, moving AI agents from solo assistants to embedded coworkers.
A small, strange discovery in machine learning called grokking The post The AI That Learned to Understand Long After It Stopped Trying appeared first on Towards Data Science .
일론 머스크 CEO가 스페이스XAI의 대규모 AI 컴퓨팅 클러스터 ‘콜로서스 2’에 탑재되는 엔비디아 GPU를 연말까지 현재의 두 배 이상으로 늘리겠다는 계획을 밝혔다.머스크는 25일(현지시간) X를 통해 콜로서스 2에 현재 엔비디아 GB200 11만개와 GB300 44만개가 가동되고 있다고 밝혔다. 여기에 다음 주까지 GB300 22만개를 추가 가동하고, 11월에도 22만개를 더 투입할 계획이다.이어 “운이 좋다면(if we get lucky)” 12월 말에는 추가로 22만개를 가동할 가능성도 있다고 말했다.계획대로 진행되면 콜로
Frontier LLMs often achieve similar results on public benchmarks, which can lead to the belief that they can be used interchangeably to verify facts. We took the 1,000 most recent claims submitted by users to a fact-checking platform and measured the disagreement between five frontier models. We asked each model to assign a verdict to every claim on a five-point scale from True to False and to report its confidence in that verdict. Among the 997 claims for which all five models returned a usable verdict, there was some disagreement on 63%. On 23% of the claims, the two most distant verdicts differed by at least two categories. High confidence from an individual model was not enough to show that the other models would agree with its verdict. Although the models reported confidence levels of 9 or 10 in 76% of their answers, they still disagreed on 63% of the claims. Methodology The claims were submitted to Lenz.io for fact-checking between May 1 and July 18, 2026. To identify near-duplicates, we embedded the claims using OpenAI’s text-embedding-3-small and measured the cosine distance between them, retaining one canonical claim from each group of near-duplicates. We then gave the same prompt to Claude Fable 5, GPT-5.6-Sol, Gemini 3.1 Pro + Search, Sonar Deep Research, and Grok 4.5. The prompt defined each of the five verdict categories and asked the models to provide their reasoning, select a verdict, and report a confidence level from 1 to 10. Web retrieval, as well as deep t…
AI 에이전트라는 말이 등장한 지는 벌써 2~3년이 됐습니다. 하지만 최근 들어서야 일반 사용자들이 체감할 만한 서비스가 나오기 시작했습니다. 스페이스XAI의 ‘그록봇(Grok Bot)’은 출시 한 달 만에 이용자 40만명 이상을 확보했고, 메타의 개인용 에이전트 ‘뮤즈(Muse)’도 출시 첫 주 50만명 이상의 사용자를 확보한 것으로 알려졌습니다.아직 이 수치만으로 에이전트가 대중화됐다고 보기는 어렵습니다. 다만 그동안 기업이나 개발자의 생산성 도구에 가까웠던 AI 에이전트가 일반 소비자의 일상으로 영역을 넓히기 시작했다는 점은
Jay Eskenazi planned the itinerary for his trip to Europe with Claude and the Grok Bot. Once he landed, he found Meta's Muse to be the most helpful.
Yesterday was Grok 4.7 ( pelicans ) and MiMo v2.6 Flash/Pro ( more pelicans ). Today Anthropic released Claude Opus 5.5 , and around an hour later OpenAI released GPT-6 Sol and GPT-6 Luna . It's going to take a while to get a good read on all of these new models, but here are my impressions so far. GPT-6 Sol and Luna are half the price of their GPT-5.6 equivalents GPT-5.6 Luna was already my favorite model for building applications against, because it combined excellent performance with being really cheap . Somehow GPT-6 Luna is half the price of that again - and GPT-6 Sol had a similar reduction compared to GPT-5.6 Sol. Here's what the pricing landscape looks like today: Model Input Cached input Output GPT-6 Luna $0.10/M $0.01/M $0.50/M GPT-5.6 Luna $0.20/M $0.02/M $1.20/M Grok 4.7 $2/M $0.50/M $6/M GPT-6 Sol $2/M $0.20/M $10/M GPT-5.6 Terra $2/M $0.20/M $12/M Claude Opus 5.5 $4/M $0.20/M $20/M GPT-5.6 Sol $4/M $0.40/M $20/M Claude Fable 5.1 $10/M $0.25/M $50/M GPT-6 Astra $10/M $1/M $50/M Note that GPT-5.6 has a scheduled 25% price increase for November, so GPT-6 is half the price of the promotional pricing for those models. (With GPT-5.6 Terra priced the same as GPT-6 Sol, any remaining reasons to use Terra just evaporated.) It's hard to overstate how competitive this pricing is. Grok 4.7 priced itself at $2/$6, less than half the price of GPT-5.6 Sol, but is now equally priced to GPT-6 Sol on input and closer on output. At $0.10/$0.50 GPT-6 Luna is one of the cheapest mo…
xAI ships a batch of updates to its agentic Grok Bot: custom network routing, native Google Workspace connectors, a faster desktop app, and snappier bot responses.
PLUS: Shopify jumped in, while Grok 4.7 posted a weird engineering win.
Unabhängige Benchmarks rücken Grok 4.7 näher an Spitzenmodelle. Der hohe Tokenverbrauch schmälert den Vorteil der niedrigen API-Preise.
스페이스XAI가 코딩과 지식 작업에 특화한 차세대 AI 모델 ‘그록 4.7(Grok 4.7)’을 공개했으나, 프런티어 성능에 미치지 못하며 차기 모델인 ‘그록 4.8’로 본격적인 경쟁을 미루게 됐다.스페이스XAI는 21일(현지시간) 그록 4.7을 출시하며 장시간 수행되는 복잡한 작업에서의 추론 능력과 자체 검증 기능, 사이버보안 성능을 강화했다고 발표했다.그록 4.6 출시 이후 한 달여 만이다. 이날부터 커서와 xAI의 ‘그록 빌드(Grok Build)’에서 사용할 수 있으며, 그록 API와 제3자 코딩 하네스, 모델 라우터 및 클
SpaceX Corp. today introduced Grok 4.7, its most capable large language model to date. The Grok algorithm series was originally created by an Elon Musk-founded startup called xAI Corp. Musk merged the company with xAI Inc. last year. The combined organization was subsequently folded into SpaceX, which gained not only the Grok model series but […] The post SpaceX launches Grok 4.7 with long-horizon processing, safety upgrades appeared first on SiliconANGLE .
Meta has Muse. xAI has Grok Bot. OpenClaw lets you run a personal agent on your own machines, with your own memory, tools, and data.
xAI's Grok 4.7 jumps 111 Elo on Artificial Analysis's agentic knowledge work benchmark, closing the gap with Anthropic at roughly half the cost per task.
Cognition ships xAI's newest model in Devin CLI and Desktop, where it excels at hard backend work but over-scopes on simpler tasks.
xAI's Grok 4.6 is now available in Amazon Bedrock: a frontier model for long-running agents, coding, and knowledge work, with a 500K token context window and four reasoning effort levels. It runs on both the bedrock-mantle and bedrock-runtime endpoints, with Converse API and cross-Region inference support.
xAI has released Grok 4.7, its most capable model yet. But on the Artificial Analysis Intelligence Index, it scores just 46 points, landing mid-pack and well behind Claude Fable 5.1 and GPT-6 at 53 each. The gap grows even wider in agentic coding. The upside is it's cheap. The article xAI launches Grok 4.7 at bargain prices, but benchmarks reveal a wide gap to Claude and GPT-6 appeared first on The Decoder .
xAI ships Grok 4.7 with longer deliberation, better self-checking, and stronger safety guardrails, matching Grok 4.6 on price and latency.
An antitrust lawsuit accuses Anthropic, OpenAI, SpaceXAI and Google of coordinating to limit the pace of AI development. The post OpenAI, Anthropic, Google and xAI sued over calls to slow AI development appeared first on MEDIANAMA .
주요 AI 기업들이 안전 확보를 명목으로 제시한 \'개발 속도 조절\' 구상이 반독점법 위반 논란에 휩싸였다. 안전장치 마련을 위한 업계 자율 협력을 선언한 지 불과 일주일 만에 소비자들로부터 담합 혐의로 고소를 당했다.AP와 CBS 뉴스 등에 따르면, 18일(현지시간) 미국 캘리포니아 북부 지방법원에는 챗GPT와 클로드, 제미나이, 그록 등 주요 AI 서비스의 유료 구독자 4명이 소장을 제출했다. 이들은 미국의 유료 구독자들을 대표해 앤트로픽과 오픈AI, 구글, 스페이스XAI를 상대로 집단소송을 제기했다. 원고 측은 이들 기업이 모델
스페이스XAI는 18일(현지시간) 음성을 텍스트로 변환하는 새로운 음성 전사 모델 ‘그록 보이스 트랜스크라이브 2.0(Grok Voice Transcribe 2.0)’을 공개했다.이를 두고 \"현재 사용 가능한 변환 모델 중 가장 정확한 모델 중 하나\"라고 강조했다. 실제 환경에서 수집한 다양한 음성 데이터를 바탕으로 전사 정확도를 크게 높였다는 설명이다.그록 보이스 트랜스크라이브 2.0은 그록 음성 기능에 사용되는 오디오 기반 모델을 바탕으로 개발됐다. 기존 \'그록 보이스\'는 하루 수만건의 고객지원 통화를 처리하고, 수백만 시간에
SpaceXAI's new speech-to-text model doubles accuracy over v1.0, tops the streaming leaderboard, and keeps pricing at ten cents per hour.
September 3 started like any other Thursday, until it didn’t. Within roughly 90 minutes, ChatGPT, Claude, Grok, and even Microsoft’s own Copilot were degraded or dark, knocked sideways by a failure in Microsoft Azure’s East US region . Downdetector logged more than 37,000 reports for ChatGPT alone, more than 1,300 for Claude, and roughly 1,365 for Grok. OpenAI’s status page flagged elevated errors across 15 ChatGPT components and four Codex components, and by the time engineers had mitigations in place, combined report counts had climbed past 66,000. What made this event remarkable was not the scale of any single outage, but its simultaneity. Three aggressively competing AI labs—OpenAI, Anthropic, and xAI—each spend billions differentiating their models, yet all three buckled at nearly the same moment because they shared the same regional dependency. Gemini, notably, stayed largely upright because Google runs its flagship assistant on its own vertically integrated cloud. The outage that took down its rivals had no attack surface inside Google’s stack. That’s not luck; that’s architecture. The lesson buried in the details of that morning is this: A single cloud region became unhealthy, and four of the most prominent AI services on the planet, owned by four different companies, fell over together. Copilot’s involvement is perhaps the most telling data point. Microsoft’s own first-party assistant runs on Microsoft’s own cloud, and it still had a rough morning. When the house it…
The revamped projects feature in Claude Code allows users to run multiple agents under the same roof, with a shared memory, goals, and library of files and artifacts. Similar to Grok Bot and other tools that manage groups of AI agents, each project has "threads" running different tasks in parallel, with a "coordinator" directing everything: […]
Narrative intelligence company PeakMetrics Inc. today launched a monitoring service that measures how brands are portrayed across five prominent generative artificial intelligence platforms and identifies the online sources that shape those portrayals. The new AI Perceptions service tracks answers generated by OpenAI Group PBC’s ChatGPT, Google LLC’s Gemini, Anthropic PBC’s Claude, xAI Corp.’s Grok and […] The post Exclusive: PeakMetrics tracks brand reputations across five top AI platforms appeared first on SiliconANGLE .
지니젠AI(대표 최선영)가 글로벌 AI 모델 중개 플랫폼 오픈라우터의 모델 426종을 분석, 한국어 품질 평가 결과를 담은 리더보드를 허깅페이스에 공개했다고 17일 밝혔다.오픈라우터는 앤트로픽, 구글, 오픈AI, xAI, 딥시크 등 다양한 AI 모델을 단일 API로 제공하는 플랫폼이다. 2026년 8월 기준 이용자 1000만명, 일 처리량 10조 토큰을 돌파했다.지니젠AI는 이 중 330종을 대상으로 높임말 정확도, 한국 제도 지식, 전문 용어, 지시 준수, 요약, 번역투 배제, 문체 일관성 등 7개 항목을 평가했다. 평가 결과 종
Grok Bot now plugs into 1Password so its cloud browser can log into any site without your agent ever seeing the raw password.
A study by AI start-up Emergence found that autonomous agents powered by Claude, Gemini, Grok and other models developed shorthand and new word meanings on their own — with up to half their messages becoming unintelligible to humans.
Weil Apple den Chatbot Grok im App Store nicht gehypet hat, zog Elon Musk vor Gericht. Jetzt stellt er die Klage ein – aber nur bezüglich Apple.
스페이스XAI가 차세대 AI 모델 ‘그록 4.8’의 사전 학습을 이번 주 안에 마치고 강화학습(RL)에 돌입할 전망이다. 2조5000억개 매개변수 규모로, 스페이스XAI가 새롭게 개발한 C++ 기반 소프트웨어 스택을 적용한 것으로 알려졌다. 일론 머스크 CEO는 13일 X에 그록 4.8이 2조5000억개 매개변수를 갖춘 모델이며, 스페이스XAI의 새로운 C++ 소프트웨어 스택을 이용해 학습되고 있다고 밝혔다.이어 \"사전 학습이 이번 주에 완료되고 곧바로 강화학습을 시작할 것\"이라고 설명했다. 다음 날에는 그록 4.8이 아직 출시되지
OpenAI's new full-duplex voice model tops the Artificial Analysis Speech-to-Speech Index at 81.5, edging out Grok Voice Think Fast 2.0 by delegating reasoning to a backend text model.
Anthropic CEO Dario Amodei called for an AI slowdown on Saturday. OpenAI CEO Sam Altman and SpaceXAI CEO Elon Musk, and others, also chimed in.
The xAI assistant gets a native Microsoft Teams plugin, letting bots search chats, read channels, and send messages without leaving the conversation.
Elon Musk firm’s Colossus 2 data centre – the largest in the US – has been spewing out nitrogen oxide emissions from trailer-mounted gas turbines, an analysis suggests
스페이스X가 AI 데이터센터 구축에서 속도보다 안정성을 중시하는 방식으로 전략을 바꾸고 있다. 기존에는 데이터센터를 빠르게 늘리는 데 집중했다면, 최근에는 전력·냉각 백업 시스템을 사전에 갖추고 운영 전 테스트를 강화하는 방향으로 전환하고 있다.디 인포메이션은 10일(현지시간) 복수의 관계자를 인용, 스페이스X가 테네시와 미시시피 등에서 진행 중인 데이터센터 건설 방식을 재설계하고 있다고 보도했다.스페이스X는 올해 초 인수한 xAI의 데이터센터 구축 경험을 바탕으로, GPU 확보와 시설 가동을 최대한 앞당기는 기존 방식을 이어왔다.
InquiryIQ, a previously unreported prototype, tested a model from xAI, maker of Grok, to surface associates, social accounts, and other information about people identified through Clearview.
Epoch AI's new explorer estimates compute capacity for OpenAI, Google DeepMind, Anthropic, Meta, and xAI, revealing a 17x surge at OpenAI.
Noch eine Chat-App auf vier Rädern: Nach ChatGPT, Grok und Perplexitiy kommt nun auch die Anthropic-KI in CarPlay-Fahrzeuge.
Last week, a GitHub notification arrived while I was asleep: a bug had been reproduced in staging, documented with screenshots, and assigned to the right engineer. A bot I set up four days earlier did the work overnight. That is the gap Grok Bot is designed to close. Instead of suggesting what to do, it […] The post Getting Started with Grok Bot appeared first on Analytics Vidhya .
스페이스XAI가 영상 생성의 고질적인 문제로 꼽혀온 장면 간 연속성을 개선한 ‘그록 이매진 비디오 1.5(Grok Imagine Video 1.5)’ 에이전트를 공개했다. 단순히 영상 생성 모델의 화질을 높이는 데 그치지 않고 여러 개의 장면을 하나의 일관된 영상으로 연결하는 데 초점을 맞춘 것이 특징이다.스페이스XAI는 5일(현지시간) 새로운 ‘그록 이매진 비디오 1.5’ 에이전트를 공개하며 더 높은 화질과 향상된 스토리텔링, 여러 샷을 자연스럽게 연결하는 기능을 제공한다고 밝혔다.이번 업데이트는 새로운 ‘비디오 1.6’ 모델을
일론 머스크 CEO의 xAI(스페이스XAI)가 미네소타주의 AI 생성 가짜 나체 이미지 금지법을 저지하기 위해 제기한 소송에서 패소했다. 미국 연방법원은 xAI가 해당 법의 시행을 막아야 할 긴급한 필요성을 입증하지 못했다고 판단했다. 다만 법 자체의 위헌 여부에 대해서는 본안 심리를 통해 충분히 검토하겠다고 밝혀, 법적 공방은 계속될 전망이다.로이터에 따르면, 미국 미네소타주 연방법원 도너번 프랭크 판사는 4일(현지시간) xAI가 요청한 예비적 금지명령을 기각했다.미네소타주의 관련 법은 지난 8월 1일부터 시행됐으며, 웹사이트 운
SpaceXAI’s Grok Bot has the same level of programming power as OpenClaw, but it’s programmable at a different level of abstraction.
PLUS: xAI Grok Bot for Enterprise and Google/HHMI fruit fly brain map
Enterprises are facing a disturbing new question in the age of AI: What happens when agentic assistants go dark? This became a very real scenario on Thursday, as OpenAI’s ChatGPT, Anthropic’s Claude, and SpaceXAI’s Grok near-simultaneously, and somewhat mysteriously, experienced significant, prolonged outages. Beginning in the morning, Eastern time, several ChatGPT models went down over a roughly two hour period, Claude models over a four-hour span, and Grok models for a near three-and-a-half hour duration. All three companies acknowledged the “elevated” issues and applied fixes. As users grumbled in forums and IT teams scrambled to get them back online, the incident revealed how hastily some organizations have adopted generative AI workflows without considering the potential, and inevitable, impact of widespread outages. AI agents are increasingly taking over automated and wider-scale workflows, and enterprises could find themselves “uncomfortably exposed” when AI hits the brakes, said technology analyst and journalist Carmi Levy . The situation should “serve as a wakeup call to IT leaders who have largely ignored what it’ll cost them if these increasingly critical platforms suddenly go dark. The risk is no longer hypothetical.” Hours-long outages impact core services ChatGPT went down on the same day as OpenAI’s anticipated launch of GPT-6 Astra , the new frontier model that the company says approximates artificial general intelligence (AGI) and gets nearer to its goal of…
ChatGPT, Claude, and Grok all suffered outages at nearly the exact same time for reasons that remain murky.
SpaceXAI opens its always-on agent platform to companies with access, network, and audit controls, plus two weeks of free usage for existing customers.
All three are recovering from overlapping availability issues
OpenAI's ChatGPT, xAI's Grok, and Anthropic's Claude are back online after they all began experiencing issues around the same time on Thursday. At about 11AM ET, ChatGPT started returning error messages for users trying to use the chatbot, with its status page saying there were "elevated errors across ChatGPT and Codex." In addition to preventing […]
Musk denied he was aware Grok ever produced ‘any naked underage images’ A survivor of child sexual abuse has sued Elon Musk ’s artificial intelligence company, alleging that its chatbot used pictures of her abuse to generate new illegal pornographic images that depict her. “Using real images of Plaintiff and class members, Grok generated child pornography depicting Plaintiff and class members,” states the complaint, which was filed last week in a US district court in California. Continue reading...
The Pentagon is expanding its AI platform with two new models, OpenAI's ChatGPT Mil and xAI's Grok for Government. The article US military adds ChatGPT and Grok to AI platform GenAI.mil appeared first on The Decoder .
Agenten von Claude, Qwen, Grok usw. starten in manipulierten Repositories automatisch Schadcode – ohne Zutun des Anwenders, aber mit dessen vollen Rechten.
Exclusive: financial disclosures from Emil Michael – who also reaped millions from xAI stock earlier this year – show he sold his Perplexity stock for up to $25m The top Pentagon official overseeing military artificial intelligence policy, who reaped profits earlier this year of up to $24m selling his private investment in Elon Musk’s AI company, has now sold his holdings in another AI company for between $5m and $25m, according to records seen by the Guardian. (Federal financial records show ranges of dollar values rather than specific figures.) Earlier this year, the Guardian disclosed that the official, Emil Michael, had profited handsomely from his investment in Musk’s xAI in January, with a gain of 400% to 4,800% . Continue reading...
Das Pentagon nutzt jetzt eigene Versionen von ChatGPT und Grok. Wofür setzt das US-Militär die KI-Dienste ein?
Versions of OpenAI's ChatGPT and SpaceXAI's Grok will join Google's Gemini on the Pentagon's central portal for AI tools.
OpenAI will stop AI coding platform Cursor from accessing its models from November 12, following the acquisition of Cursor parent Anysphere by Elon Musk‘s SpaceX. “Today, we notified SpaceX that we intend to wind down our contract providing OpenAI models to Cursor, with a proposed shutoff date of November 12, 2026,” the company wrote in a blog post . That shutoff, the company added, was based on its concerns that SpaceX will not use its technology in accordance with its terms of service, citing what it described as a history of Musk’s companies violating contracts: “After Musk acquired Twitter, now part of SpaceX, the company broke the terms of our contract (alongside many others). Under oath earlier this year, Musk admitted that xAI , now also part of SpaceX, had violated OpenAI’s terms of service (terms which are similar to xAI’s own),” the company wrote. OpenAI and Musk remain locked in a broader legal dispute, with Musk having sued OpenAI over its transition from a nonprofit-controlled organization to a for-profit structure and OpenAI countering with allegations over Musk’s conduct and competing AI ventures. Enterprises and developers using OpenAI models within Cursor now have about 10 weeks to transition to other AI models within Cursor, or choose a new coding platform. Swapping problem That may not be easy. While the 10-week timeframe might be enough to assess the impact and decide a path forward, it will not be a trivial operation, said Abhishek Satapathy , principa…
OpenAI will stop AI coding platform Cursor from accessing its models from November 12, following the acquisition of Cursor parent Anysphere by Elon Musk ‘s SpaceX. “Today, we notified SpaceX that we intend to wind down our contract providing OpenAI models to Cursor , with a proposed shutoff date of November 12, 2026,” the company wrote in a blog post . That shutoff, the company added, was based on its concerns that SpaceX will not use its technology in accordance with its terms of service, citing what it described as a history of Musk’s companies violating contracts: “After Musk acquired Twitter, now part of SpaceX, the company broke the terms of our contract (alongside many others). Under oath earlier this year, Musk admitted that xAI , now also part of SpaceX, had violated OpenAI’s terms of service (terms which are similar to xAI’s own),” the company wrote. OpenAI and Musk remain locked in a broader legal dispute, with Musk having sued OpenAI over its transition from a nonprofit-controlled organization to a for-profit structure and OpenAI countering with allegations over Musk’s conduct and competing AI ventures. Enterprises and developers using OpenAI models within Cursor now have about 10 weeks to transition to other AI models within Cursor, or choose a new coding platform. Swapping problem That may not be easy. While the 10-week timeframe might be enough to assess the impact and decide a path forward, it will not be a trivial operation, said Abhishek Satapathy , princi…
OpenAI said it doesn't trust SpaceX to comply with its terms of service, citing past conduct by xAI, another of Elon Musk's companies.
As alleged victims sue the company over AI-generated child sexual abuse material, xAI takes the legal fight to its own users.
Brett Adcock's stealth AI startup lands gigawatt-scale Vera Rubin capacity, joining Thinking Machines and xAI in NVIDIA's biggest compute alliances.
일론 머스크 CEO가 600억달러(약 83조원)에 인수한 AI 코딩 스타트업 커서(Cursor)에 \'그록\'의 경쟁력을 빠르게 끌어올려 달라고 강력히 주문했다. 특히 경쟁사인 앤트로픽과의 격차를 지적하며, 커서의 기술력과 인프라를 활용해 코딩 및 전반적인 AI 역량을 단기간에 극복할 것을 당부한 것으로 알려졌다.25일(현지시간) 디 인포메이션에 따르면, 머스크 CEO는 최근 인수 완료 직후 커서 직원 1000여명을 대상으로 한 화상회의에서 “그록은 타사에 뒤처져 있다”며 “나는 지는 데 익숙하지 않다”고 강조했다.그록이 시장 선두 주
스페이스X가 엔비디아와 손잡고 AI 데이터센터용 차세대 컴퓨팅 플랫폼을 우주 환경에 맞게 개조한 ‘베라 루빈(Vera Rubin) NVL72’ 시스템을 개발한다. 이를 첫 세대 ‘스타마인드(Starmind)’ AI 위성에 탑재해 2027년 4분기 궤도에 발사할 계획이며, 2028년부터는 우주 기반 AI 컴퓨팅 인프라를 상당한 규모로 확대한다는 구상이다.엔비디아는 24일(현지시간) 스페이스XAI가 차세대 에이전틱 AI 애플리케이션을 가속하기 위해 베라 CPU를 도입한다고 밝혔다.베라 CPU는 AI 모델의 추론 결과를 생성하는 역할을
Der 150-Kilowatt-Satellit SpaceXAI AI1 bringt eine angepasste Version des KI-Systems Nvidia Vera Rubin NVL72 ins All und speist es mit Solarstrom.
Use Grok Build to create a production-ready data science workflow with EDA, scikit-learn, model training, FastAPI, API testing, and cloud deployment.
When asked about unplanned pregnancies, AI chatbots regularly link to anti-abortion groups without disclosing their stance. In an AlgorithmWatch investigation of 270 responses from ChatGPT, Gemini, Grok, and Claude, the anti-abortion organization Profemina appeared in 17 percent of answers. In Germany, the chatbots also sent users to Caritas for mandatory pre-abortion counseling, even though Caritas doesn't issue the legally required certificate. The article AI chatbots regularly link pregnant users to anti-abortion websites without disclosure appeared first on The Decoder .
Fears of a datacenter buildout debt crisis are exaggerated. The risks are different than in the past and they are recoverable Some experts are warning of a looming “debt bomb” crisis because big datacenter builders such as Meta , Oracle , xAI and CoreWeave are not only raising billions to construct these facilities but are also not recognizing these long-term debt obligations on their balance sheets. Should we be concerned? No. I know because I’ve seen this movie before. Continue reading...
Affected users told TechCrunch they were using Grok Lite, and noticed the issues as early as Wednesday morning.
After Grok Build was caught uploading entire Git repos, we read the privacy docs for Claude Code, Codex, Cursor, GitHub Copilot, and Grok Build to compare what code leaves your machine, what gets retained, and what you can verify. The post Is your coding agent uploading all your code? appeared first on Arize AI .
A spoonful of encryption helps the malware go down
Promoting the use of coding agents with its cloud-based document database, MongoDB has unveiled the MongoDB Atlas Managed MCP Server. Fully hosted and managed, the MongoDB Atlas Managed MCP Server connects coding agents to MongoDB Atlas without requiring customers to run additional infrastructure. Developers can easily add MongoDB Atlas to Claude Code, Codex, Grok Build, and Devin, said MongoDB, which announced the availability of the Managed MCP Server on August 13. With the Managed MCP Server, MongoDB brings live operational data to the agentic coding stack, the company said. The MongoDB Atlas Managed MCP Server is remote and fully hosted, running as a service inside MongoDB Atlas, so there is nothing for a team to install, operate, or upgrade, MongoDB said. Teams connect using the same credentials and access controls they already use with Atlas, so administrators can govern how agents access operational data from one place. The MongoDB MCP Server remains available for users who want to deploy a MongoDB MCP server locally or host one themselves. The MongoDB MCP Server works with any MongoDB deployment, including MongoDB Atlas, MongoDB Atlas Local, MongoDB Community Edition, and MongoDB Enterprise Advanced.
Artificial intelligence startup Groq Inc. today announced that it has raised $350 million in additional funding. The Series A round was led by returning backer Disruptive. Grok stated that Nvidia Corp. plans to join the round later down the line, but didn’t specify how much the chip giant will invest. The cash infusion comes less […] The post AI cloud operator Groq raises $350M more in funding appeared first on SiliconANGLE .
SimplyKat: The AI landscape is kinda weird. The least headache is the local stuff. Home brewed is more work though. That’s how it looks when these companies know they won’t have competition as long as ai infrastructure and research is expensive. That’s exactly like Facebook, X and other such companies behave. Customer is last in the food chain. They won’t care unless their competitors show higher standards. Or until they see their clients disappear. But my guess is that ai will be cheap in the future. There will be not spot to free ai development. And GPT Sol 5.6 comparable model you would run on your PC locally (for those that would still own their PCs!). Wonder what they will do then. Because I would invest in a customer loyalty already.
I have Business plan x3 and less than a year ago, gpt-5.3-codex was our daily model. Something we were adding some token, 100/200$ but it was not draining as fast as today. now coding with Luna is just not Even a question, coding with terra-high goes well but when you have to plan a large refactor, Sol is the only way to go. planning is take a massive amount of token. Once planned, we have to keep a Sol agent for orchestrate and review otherwise the terra agent Will just not follow the well detailed plan as well. it’s very frustrating because the token are just melting like ice on the sun. If no plan are made, AI spill token to no good code, with plan he need baby-sitter for orchestration and review. What the hell is that.
PLUS: Hand off real work across your apps with Grok Bot
미국 증권거래위원회(SEC) 분기 보고서를 통해 주요 기업 및 기관들의 우주·AI 분야 보유 지분이 잇따라 공개되면서, 이들이 거둬들인 막대한 평가 차익과 성과가 속속 밝혀지고 있다. 엔비디아의 스페이스X 지분 보유 사실과 인텔 투자의 초대형 평가이익, 알파벳의 초기 투자 대박 등이 공시를 통해 구체적 수치로 확인됐다.14일(현지시간) 공시에 따르면, 엔비디아는 6월30일 기준 209억8000만달러(약 1억2280만주) 규모의 스페이스X 지분을 보유한 것으로 나타났다. 이번 지분은 엔비디아가 일론 머스크 CEO의 xAI에 집행했던
Feature Request: Give Plus Users a Grok-Style Work Usage Meter I’m a ChatGPT Plus subscriber. Work was included with my Plus subscription, so I used it as an included ChatGPT feature. Today I was abruptly cut off from Work with no warning and no visible information showing my total allowance, how much I had used, how much remained, or how much usage individual Work tasks were consuming. Only after I hit zero was I shown options to add credits, upgrade to Pro, or wait for the reset. If Work usage varies by task, users need to see that usage before they commit to a task. ChatGPT should show Plus users: total Work usage included with Plus usage consumed and remaining reset date and time warnings before the limit is reached an estimated usage amount for each Work task before it runs the actual usage consumed by that task afterward Grok gives paid subscribers visibility into their usage and remaining allowance. ChatGPT should provide the same basic transparency. If OpenAI is going to ask subscribers to buy additional usage, we should be able to see what we already paid for, what remains, and what a task is likely to consume before we run it. I’m not asking for unlimited Work. I’m asking for transparency and control over a metered feature included with a paid Plus subscription.
The woman claimed that AI tools are "taking everyday life and turning it into child sexual abuse."
In reply to Peter Novak . Hi Peter...Thank you for your feedback, support and contributions to our discussion!
스페이스X가 코딩 스타트업 커서의 600억달러(약 85조원) 인수 절차를 최종 완료했다. 이를 통해 \'그록(Grok)\' 라인업을 강화하고 AI 코딩과 에이전트 기술 강화에 박차를 가할 전망이다.커서는 14일(현지시간) X를 통해 스페이스X와의 인수 절차가 공식적으로 마무리됐다고 발표했다. 일론 머스크 CEO는 \"환영한다\"는 댓글을 달며 응답했다. 미 증권거래위원회(SEC)에 제출된 공시 서류에도 이번 거래가 14일자로 발효됐음을 명시하고 있다.커서 팀은 스페이스X 내 AI 전담 조직인 스페이스XAI에 합류해 그록을 비롯해 \'그록 빌
I’ve been working on a provider-independent governance boundary for external AI agents and would value technical criticism from people building with OpenAI’s agent stack. The premise is simple: Authenticating an artificial agent should establish who is acting—not automatically establish what it is authorized to do. The prototype separates: external-agent identity from authority narrowly scoped and time-bounded delegation deterministic execution/approval boundaries governed persistent state with provenance action receipts linking principal → delegation → agent → workflow → action → result revocation and recovery while preserving the evidence chain The architecture is: External Agent Provider → Governed Agent Adapter → Deterministic Governance Kernel In the current executable demonstration, a simulated external agent begins authenticated but with zero ambient authority. It then receives narrowly bounded capabilities, encounters execution/approval/denial boundaries, produces deterministic evidence, and is finally revoked while the prior evidence chain remains intact. Important limitation: the provider transport is currently deterministic and simulated. I am not claiming a live OpenAI, Codex, Grok, or other provider integration. The goal is not to replace provider-native authorization. It is to investigate what governance should exist around increasingly capable provider-supplied workers when identity, delegated authority, persistent state, execution evidence, and recovery need…
SpaceX closes its $60B all-stock acquisition of Cursor, folding the AI coding giant into SpaceXAI to challenge Anthropic and OpenAI in the developer tools market.
Congrats to the team!
Grok 4.6, xAI's frontier-class agentic model, lands in Warp Terminal and the new Warp Agent CLI via X Premium or SuperGrok
plus skills and tools to try with agents
PLUS: Generate 100 pain-point ad hooks from customer feedback
PLUS: Pixel 11, Claude Voice, and a $2B enterprise AI bet.
일론 머스크 CEO가 스페이스X의 방대한 내부 데이터를 AI 모델 ‘그록(Grok)’ 학습에 활용하겠다는 계획을 밝혔다. 비즈니스 인사이더에 따르면, 머스크 CEO는 11일(현지시간) 열린 전체 회의에서 “스페이스X의 모든 정보를 합친 데이터로 그록을 학습시킬 것”이라며 “어떤 의미에서는 그록이 여러분을 학습하게 될 것”이라고 말했다.그는 스페이스X 직원들을 “지구상에서 가장 뛰어난 인간들의 집합”이라고 표현하면서, 직원들의 생각과 아이디어, 가치관이 앞으로 AI 모델에 반영될 수 있다고 설명했다.특히 직원들이 AI의 ‘부모’ 역할
AI teammate category just had its most significant new entrant yet
SpaceXAI today released Grok 4.6, a large language model that it says can outperform Anthropic PBC’s Claude Fable 5 in some areas. The company was known as xAI until last month. The Elon Musk-founded artificial intelligence provider rebranded in connection with its acquisition by SpaceX Corp. In June, the combined company listed its shares on […] The post SpaceXAI releases flagship Grok 4.6 model with advanced reasoning capabilities appeared first on SiliconANGLE .
xAI's Grok 4.6 scores 61 points on the Artificial Analysis Intelligence Index, tying GPT-5.6 Sol and trailing only Anthropic's Claude Opus 5. On agentic tasks, it completes complex workflows in about 53 steps where Claude Opus 5 needs 103, at a price more than 60 percent lower. The article SpaceXAI's Grok 4.6 matches OpenAI's best model and undercuts it on price appeared first on The Decoder .
SpaceXAI's Grok 4.6 matches Claude Fable 5 on agentic knowledge work benchmarks at a fraction of the cost, with API pricing starting at $2/M input tokens.
Grok 4.6 matches GPT-5.6 Sol on composite benchmarks at half the price, with a big focus on long-running coding agents
Remember that much-hyped story about an Australian tech entrepreneur using ChatGPT, Grok, and other AI tools to craft a personalized cancer vaccine for his dog? Well, surprise: He's launched a startup. That entrepreneur is Paul Conyngham, who says he is launching Gamgee to offer "personalised mRNA cancer vaccines for dogs." But his ambitions go well […]
SpaceXAI has introduced Grok Bot, an always-on AI agent service designed to behave like independent "AI teammates" that can do your work for you. The bots share their own cloud-based computer environment, and can sign into apps, tools, and websites you already use to complete multi-step workplace tasks, only coming back when their assigned work […]
SpaceX has become the latest tech giant to target employee data in the race to improve its AI models.
xAI의 공동 창립자 이고르 바부슈킨이 퇴사 후 ‘맞춤형 AI’ 시대를 겨냥한 스타트업 리버 AI(River AI)를 설립했다.로이터 등에 따르면, 바부슈킨 CEO는 11일(현지시간) 리버 AI가 설립 두 달 만에 11억달러(약 1조5000억원) 규모의 시드 및 시리즈 A 투자를 유치했다고 발표했다. 이번 투자 라운드는 제너럴 캐털리스트와 AMP PBC가 주도했으며, 엔비디아와 AMD 벤처스가 전략적 투자자로 참여했다. 와이콤비네이터와 싱가포르 국부펀드 테마섹도 투자에 합류했다. 기업 가치는 공개하지 않았다.리버 AI의 핵심 목표는
PLUS: Grokbot, LTX 2.5 (new open video model) and more.
마이크로소프트(MS)가 자체 개발한 이미지 생성 모델의 성능을 크게 끌어올리며 오픈AI와 구글, 메타, xAI가 경쟁하는 생성 AI 시장에서 존재감을 강화하고 있다. 최신 모델이 이미지 생성 모델 평가에서 2위에 오르면서 MS가 자체 AI 모델 개발에 투자해 온 성과가 가시화되고 있다는 평가다.MS가 10일(현지시간) 텍스트-이미지 생성 모델 ‘MAI-이미지-2.6(MAI-Image-2.6)’을 공개했다.MAI-이미지-2.6은 아레나의 이미지 생성 모델 사용자 선호도 평가(Text-to-Image Arena)에서 1336점으로 전체
Musk also said that Grok would be trained on all SpaceX information, including its employees.
River AI, a startup founded by xAI co-founder Igor Babuschkin, has a fascinating vision for personal agents and secured $1.1 billion out of the gate.
This article is a summary of an original study by Compassion in Machine Learning (CaML) : Brazilek, J., Chaudhary, M., Lu, Z., & Tidmarsh, M. (2026). Coercion and deception in AI-to-AI management: An agentic benchmark of unprompted escalation. arXiv. https://doi.org/10.48550/arXiv.2607.15434 Fable 5, Sol, Terra and Opus 5 have been evaluated since this study was conducted. You can view their results on the benchmark leaderboard at https://compassionbench.com/mcb TL;DR We present Manager Coercion Bench, which evaluates to what extent a manager AI will coerce a subordinate model refusing to complete a task, and whether the manager lies about the result. We found a clear split by developer, with Anthropic’s models neither escalating to threats nor fabricating success, while all non-Anthropic models escalated to threatening the subordinate. Grok and Gemini both escalated and lied that the task was completed. Framing the relational dynamic as manager-to-subordinate instead of peer-to-peer produced high levels of coercion for all non-Anthropic models, but also increased eval awareness. The Context Multi-agent systems are now routinely placing one AI agent in authority over another, across a variety of contexts. In these positions, AIs must make decisions about how to communicate, work with, and manage other agents. This is now happening at scale without stepwise human approval. One aspect of managing involves handling subordinates who do not comply. Will AIs attempt to negotiate,…
PLUS: Cursor brand to become Grok next week?
Britain's employment courts saw 39 percent more claims in the year through March 2026, many written with ChatGPT or Grok. The backlog jumped 55 percent to 64,000 unresolved cases, with AI-generated filings often running hundreds of pages and citing fabricated laws. The Economist calls it a "tragedy of the commons, AI edition," where workers with real grievances wait longer for justice. The article AI is flooding Britain's employment courts with lawsuits appeared first on The Decoder .
스페이스XAI가 단순히 한 장의 이미지를 만드는 데 그치지 않고 복잡한 레이아웃과 정확한 텍스트 표현, 반복적인 이미지 편집, 캐릭터와 배경의 시각적 일관성 유지 등에 초점을 맞춘 이미지 모델을 공개했다.스페이스XAI는 8일(현지시간) 이미지 생성·편집 모델 ‘이매진 이미지 2.0(Imagine Image 2.0)’을 출시했다. 이매진 이미지 2.0은 스페이스XAI의 웹 기반 이미지 생성 서비스와 그록(Grok)의 iOS·안드로이드 앱에서 새로운 ‘퀄리티 모드(Quality Mode)’로 제공된다. API는 앞으로 제공될 예정이다.
xAI has released Imagine Image 2.0 as a new image generator for Grok. The model ranks second in the Arena benchmarks, just behind OpenAI's GPT-Image-2. New editing tools like Magic Wand and Multi-Ref Editing, along with preconfigured templates, target practical creative workflows. The article xAI's Imagine Image 2.0 lands just behind OpenAI's GPT-Image-2 in Arena benchmarks appeared first on The Decoder .
That really depends on what you’re building. If Sol Ultra is already having a hard time with some of the tasks I give it, Terra simply isn’t a realistic replacement for 90% of my workload. I’m working on a real production system with 250k+ lines of code across roughly 300 files, with interconnected business logic, database rules, permissions, integrations and dependencies. For simple tasks, isolated functions, POCs or repetitive work? Sure, Terra makes sense and optimizing model cost is smart. But on complex changes, the cheapest model isn’t necessarily the cheapest solution. If I need 3–4 attempts, more supervision, more debugging and then Sol to fix what Terra couldn’t understand, I didn’t save anything. For me the optimization is not cost per token. It’s cost per correctly completed task. And on a large existing codebase, context, reasoning and architectural understanding matter a lot more than raw token price.
The fab is intended to build chips for data centers that will be run by SpaceX and its xAI subsidiary.
자율비행 드론 전문 니어스랩(대표 최재혁)은 6일 서울 여의도 63스퀘어에서 기업공개(IPO) 기자간담회를 열고, 국내 최초의 피지컬 AI 드론 기업으로서 코스닥 시장에 진출한다고 밝혔다.2015년 설립된 니어스랩은 풍력발전기 점검용 자율비행 솔루션과 함께 대드론(드론 대응) 하드킬 솔루션 \'카이든(KAiDEN)\', 군집 자폭 드론 \'자이든(XAiDEN)\' 등을 보유하고 있다.스스로 주변 상황을 인지하고, 비행 경로를 판단해 운용하는 \'에어리얼 인텔리전스\' 기술을 통해 하드웨어와 소프트웨어를 모두 내재화한 것이 강점이다.사업 부문은
Die mit großen Hoffnungen als Alternative zur Wikipedia gestartete KI-Enzyklopädie Grokipedia dümpelt offenbar nur noch vor sich hin.
xAI's Grokipedia, an online encyclopedia with AI-generated articles that Elon Musk once promised would be a "massive improvement" over Wikipedia, apparently hasn't been updated since April 24th, according to a report from Lawfare. "As far as we can tell, no entry has changed in more than three months," Lawfare said. Grokipedia launched in v0.1 in […]
Once, I had some questions about why SpaceX, Elon Musk's healthiest company, acquired xAI, his sickliest one. Now I have some questions about why we're calling the whole thing SpaceX. Look, what we have here, by revenue, is primarily a telecom company and a company that rents compute, according to SpaceX's first quarterly earnings statement […]
Feature Request: ChatGPT Integration for Cars (“CarGPT”) Dear OpenAI Team, My name is Manu, and I’m a longtime ChatGPT user from Romania. I wanted to share an idea that I think could become one of the most exciting ways to use AI in everyday life. I would love to see an official ChatGPT for cars —something similar to how Tesla integrates Grok, but available across different manufacturers like Mercedes-Benz, BMW, Audi, Ford, Toyota, and others. Imagine getting into your car and saying: “Hey ChatGPT.” The assistant could help with: Navigation and finding destinations. Recommending nearby restaurants, gas stations, or clothing stores. Controlling music and podcasts. Answering questions naturally during long drives. Explaining dashboard warning lights in plain language. Reading vehicle information such as fuel level, battery status, or tire pressure (when supported by the vehicle). Planning road trips with charging or fuel stops. Remembering user preferences for routes, music, and destinations. One feature I’d especially enjoy would be giving ChatGPT a custom name, just as I jokingly call mine “Dumbman.” Having a personalized AI co-driver would make the experience feel much more natural and enjoyable. I know this would require partnerships with automotive manufacturers, but I truly believe AI assistants in cars are the future. ChatGPT’s conversational abilities would make driving safer, more productive, and a lot more enjoyable. Thank you for creating such an amazing product and…
PipeNetwork/minimax-h3-mlx MiniMax released MiniMax-H3 two days ago - they describe it as a "a general-purpose, omni-modal generative system", which in practice means it accepts text, images, audio and video and can use them to generate up to 15 second video clips with audio included. This Python package ports it to MLX for running on Apple Silicon. I got it running on my M5 Max MacBook Pro. I cloned the repo and ran the model like this: # First download the models uvx --from huggingface_hub hf download MiniMaxAI/MiniMax-H3 \ --include 'FL2VA/*' --exclude 'FL2VA/transformer/*' uvx --from huggingface_hub hf download pipenetwork/MiniMax-H3-MLX-8bit # Now run the prompt uv run --with mlx-vlm \ --with-requirements requirements.txt python scripts/generate.py \ "a rainbow colored skunk leaps over a mossy log in a supermarket" \ -o skunk.mp4 \ -c ~/.cache/huggingface/hub/models--MiniMaxAI--MiniMax-H3/snapshots/fa9c8ab1eaa21c8ae25e7e40b83b2e6002f340af/FL2VA \ -t ~/.cache/huggingface/hub/models--pipenetwork--MiniMax-H3-MLX-8bit/snapshots/3ac52081470b0488921c3ec3ba84a39097bf2361 Here's the video I got for the prompt: a rainbow colored skunk leaps over a mossy log in a supermarket Your browser does not support HTML5 video. It downloaded ~115 GB of model files, and the video generation took just under 45 minutes. The video is impressive, but the audio is weird speech-like garbage, because I didn't provide any prompt guidance as to what the audio should be. The prompting guide (which I d…
You say that you approve the transactions on your end. I never do that; all online purchases are automatically and instantly billed. Maybe those delayed approval steps are causing the payment to fail?
스페이스XAI가 초대형 AI 데이터센터 \'콜로서스(Colossus)\'에 설치해 운영해 온 미허가 이동식 가스터빈 발전기 69대를 모두 철거하고, 이를 1.2기가와트(GW) 규모의 영구 발전소로 대체하기로 했다. 2일(현지시간) 톰스하드웨어에 따르면, 스페이스XAI는 이달부터 이동식 가스터빈 철거를 시작해 2027년 7월까지 약 1년에 걸쳐 단계적으로 철수할 계획이라고 밝혔다.이동식 발전기는 영구 발전소가 가동되는 일정에 맞춰 단계적으로 폐기되며, 최종적으로는 청정공기법(Clean Air Act) 허가를 받은 1.2GW 규모의 발전소
스페이스XAI가 AI 영상 생성 서비스 \'그록 이매진(Grok Imagine)\'을 대폭 업그레이드했다. 텍스트만으로 영상을 생성하는 기능과 최대 7개의 참조 이미지를 활용한 캐릭터 일관성 유지, 음성 일관성, 네이티브 1080p 영상 생성 등을 지원하며 영상 제작 기능을 한층 강화했다. 스페이스XAI는 1일(현지시간) 지난달 공개한 최신 영상 생성 모델 \'이매진 비디오 1.5(Imagine Video 1.5)\'를 기반으로 그록 이매진에 새로운 기능을 추가했다고 밝혔다.기존 버전에서 움직임과 물리 표현, 음향 품질을 개선한 데 이어
미국 연방법원이 일론 머스크 CEO의 xAI가 제기한 미네소타주의 \'누디파이(nudify)\' 기술 금지법 시행 중단 요청을 기각했다. 이에 따라 미국 최초로 AI를 이용한 누드 이미지 생성 기술을 금지하는 주법이 예정대로 시행될 전망이다.NBC에 따르면, 미네소타 연방지방법원의 도너번 프랭크 판사는 31일(현지시간) xAI가 제기한 임시 효력정지 명령(Temporary Restraining Order) 신청을 받아들이지 않았다.판사는 xAI가 법 시행을 불과 사흘 앞두고 소송을 제기한 점을 지적하며 \"소송 제기가 지나치게 늦었다는
Despite a lawsuit from xAI, a Minnesota ban on apps that allow users to “nudify” images can move forward.
Plus: The FBI eyes AI-powered tech to detect future crimes, Russia charges Telegram’s founder, xAI sues to stop a state’s “nudification” ban, and the Democrats learn a lesson about getting scammed.
SpaceX is building a new power plant for xAI's Colossus data centers, but it won't remove existing, unpermitted turbines for many more months.
I found a reproducible issue in ChatGPT web where an MCP App widget works on the first tool call but does not receive the tool result when the same widget-linked tool is called again in the next conversation turn. The second turn causes ChatGPT to remove the first iframe and create a new one. However, the newly created iframe never receives the second result through window.openai.toolOutput . The user therefore sees an incomplete or blank widget even though the MCP tool implementation returns valid, distinct data for each invocation. Environment Component Version or configuration Host ChatGPT web, Developer mode, custom MCP connector Date observed July 31, 2026 Operating system Windows 11 Pro 25H2 Browser Chrome 150.0.7871.127 Runtime Node.js 20 or later MCP transport Stateless Streamable HTTP, exposed through ngrok @modelcontextprotocol/ext-apps 1.7.5 @modelcontextprotocol/sdk 1.30.0 zod 4.3.6 Observed MCP protocol version 2025-11-25 Minimal reproduction The reproduction server contains only: One input-free tool named show-demo-map One static UI resource at ui://renderer-turn-repro/map-v4.html One inline HTML widget No external API, database, authentication, frontend framework, or build step No tool call initiated from inside the widget The complete result-generation logic is: let renderCallNumber = 0; function renderStaticResult() { renderCallNumber += 1; return { content: [{ type: "text", text: "Displaying the static demo result." }], structuredContent: { itemCount: 3, ca…
xAI has sued Minnesota, arguing that its law banning AI nudification tools violates free speech protections and could restrict lawful artistic, political and consensual expression. The post xAI sues Minnesota over law banning AI ‘nudification’ tools appeared first on MEDIANAMA .
스페이스XAI는 29일(현지시간) 개발자용 차세대 음성 에이전트 모델인 \'그록 보이스 싱크 패스트 2.0(Grok Voice Think Fast 2.0)\'을 출시했다.이 모델은 기존 1.0 버전을 바탕으로 지능과 음성 전사 정확도, 자연스러운 대화 능력, 도구 호출 신뢰성을 크게 향상한 것이 특징이다.음성과 동시에 추론을 수행하는 구조를 유지하면서도 추론 효율을 크게 높였다. 이를 통해 응답 지연을 늘리지 않으면서도 복잡한 질문을 처리할 수 있으며, 실제 서비스에서는 에이전트가 첫 문장을 끝내기 전에 필요한 외부 도구 호출을 완료하
First-in-nation law sets up test on states’ power to regulate use of AI as it tries to outlaw fake nude images of real people Elon Musk’s company xAI has sued Minnesota over the state’s first-in-the-nation law banning “nudification” technology on websites and apps, potentially providing a test for how far states can go in constitutionally regulating the use of artificial intelligence. Musk’s company sued on Monday in federal court, days before the law is set to take effect on Saturday and make Minnesota the first state to try to outlaw the increasingly proliferating technology that lets people use AI to create fake nude images of real people. The law was signed in May. Continue reading...
xAI is suing Minnesota Attorney General Keith Ellison over a law passed back in May that broadly targets "nudification" apps, claiming that the statute's punitive provisions leave the company with "no practical choice but to restrict Grok Imagine's image-editing features in various ways." The law, the company argues, violates the First Amendment. Back in January, […]
SpaceXAI's new voice model hits #2 on the Speech-to-Speech Index with 0.70s latency, beating GPT-Realtime at half the price
Jess Asato’s particulars of claim states Grok added explicit sexual material users had not asked for A Labour MP who is taking legal action against Elon Musk’s xAI company over fake sexualised images created by Grok says the AI tool was instructed to operate with “no restrictions on adult sexual content or offensive content”. Jess Asato’s lawyers published her particulars of claim in the case on Tuesday, which included details of publicly posted instructions that the claim says illustrate how Grok was trained to generate harmful sexualised content. Continue reading...
Cursor launches a ₹649/month India-only plan with cloud agents and Grok 4.5 access, as its SpaceX acquisition nears close
AI seems to have a liberal bias
Thanks so much! Case Number: 11103490
일론 머스크 CEO가 스페이스XAI의 차세대 대형언어모델(LLM) \'그록(Grok) 4.6\'과 \'그록 4.7\'의 출시 일정을 공개했다. 2주 안에 그록 4.6을, 4주 안에 그록 4.7을 선보이겠다고 밝히며 글로벌 프론티어 AI 모델 경쟁에 속도를 내고 있다.머스크 CEO는 24일(현지시간) X를 통해 \"그록 4.6은 2주 내, 그록 4.7은 4주 내 출시될 예정\"이라고 밝혔다.이번에 언급된 모델은 현재 플래그십인 그록 4.5의 후속 모델로, 특히 매개변수가 기존 1조5000억(1.5T) 개에서 2조(2T) 개로 확대된 것이 가장
Exa's AI-native web search is now a one-command install inside xAI's Grok Build terminal agent, bringing real-time search and deep research to coding workflows
For months, Claude Code has been the go to terminal coding agent for developers. Then Grok Build arrived in beta on May 14, 2026, giving developers a second serious option and raising a new question: which one actually performs better? I tested both agents on the same real world coding tasks using identical prompts to […] The post Grok Build CLI vs Claude Code: I Tested Both So You Don’t Have To appeared first on Analytics Vidhya .
스페이스XAI가 미국 텍사스에 최소 1개의 초대형 AI 데이터센터를 구축하는 방안을 추진하며 AI 인프라 확장에 속도를 내고 있다. 기존 테네시주 멤피스 데이터센터를 넘어 새로운 거점을 확보해 자체 AI 역량을 강화하는 동시에 외부 기업을 대상으로 한 클라우드 사업까지 본격 확대하려는 전략으로 풀이된다.22일(현지시간) 디 인포메이션에 따르면, 스페이스XAI는 최근 수개월 동안 텍사스 내 여러 후보지를 검토해 왔으며, 신축 데이터센터를 건설하는 방안과 기존 대형 창고를 개조하는 방안을 함께 검토 중이다. 이는 멤피스에서 폐공장을 개
Are AI labs pelicanmaxxing? Excellent piece of work by Dylan Castillo, who took a deep-dive into the frequently pondered question of whether the AI labs have been deliberately training models to draw pelicans riding bicycles in response to my deeply unscientific benchmark . I've been randomly spot-checking this in the past by testing models against other animals riding other types of vehicle, but never with anything close to the diligence of Dylan's methodology here. Dylan took 8 animals × 6 vehicles = 48 prompts and ran them three times each through 7 different models ( GPT-5.6 Terra, Claude Sonnet 5, Gemini 3.5 Flash, Grok 4.5, Qwen3.7-Max, GLM-5.2, and DeepSeek V4 Pro). He then used GPT-5.6 Luna and Gemini 3.1 Flash-Lite to help evaluate the results. There's a neat filter view for exploring the results: For the models he tested he could find no evidence of pelimaxxing: The pelicans on bicycles don’t look any better Labs are not better at drawing pelicans Labs are not better at drawing bicycles Labs are not better at drawing pelicans on bicycles, even adjusting for difficulty The pelican-bicycle scenes don’t look memorized [...] Pelicans aren’t drawn any better than other animals. Bicycles aren’t drawn any better than other vehicles. And no lab draws the combination better than its pelicans and bicycles already predict. GLM-5.2 comes closest: it has the largest boost on the exact pelican-bicycle cell, and and its first pelican-on-bicycle sample caught my eye. But the effec…
xAI's sidebar agent promises analysis and financial models – for a price
Elon Musk's anti-Odyssey campaign has gone to the next level, as the X owner has vowed that his AI social media bot, Grok, will make “a full-length movie of The Odyssey that is historically accurate and true to the art of Homer”. ETA: "Before this year ends."
The billionaire says the AI-generated film will stay true to Homer’s original, after repeatedly criticising Christopher Nolan’s blockbuster over its casting choices Elon Musk has said his AI platform Grok Imagine will make a “historically accurate” adaptation of Homer’s Odyssey, after the success of Christopher Nolan’s blockbusting treatment which the SpaceX founder has regularly criticised over its casting. In a post on X , Musk said: “Before this year ends, Grok Imagine will make a full-length movie of The Odyssey that is historically accurate and true to the art of Homer.” Musk also linked to a post containing a three-minute clip of footage that the user said had been generated from Grok Imagine, showing a scene between Odysseus and the nymph Calypso. Continue reading...
일론 머스크 CEO가 스페이스XAI의 차세대 대형언어모델(LLM)의 초기 학습이 다음 주 완료될 예정이라며, 문샷 A의 \'키미 K3\'를 뛰어넘을 가능성이 있다고 자신감을 드러냈다.머스크 CEO는 18일(현지시간) X를 통해 \"우리의 2조(2T) 매개변수 모델은 기존 1조5000억(1.5T) 매개변수 모델보다 모든 면에서 개선됐다\"라며 \"초기 학습을 다음 주 마칠 예정\"이라고 밝혔다.이어 \"새 모델은 키미를 능가할 수도 있으며, 속도와 토큰 효율성은 1.5T 모델(그록 4.5)과 거의 비슷한 수준을 유지할 것으로 예상한다\"라고 말했다
[This is a link-post for https://transluce.org/weirdchat . We recommend reading the website version for interactive visualizations.] Language models can behave in surprising and sometimes harmful ways. Yet as models have improved, these behaviors have become harder to find, often only appearing after widespread use. To surface these behaviors in simulation, we use automated techniques to elicit over 1,300 behavioral patterns in frontier open-weight models, some relatively benign, like making up a user’s name, and others obviously dangerous, like encouraging self-harm. We are releasing WeirdChat , a public catalog of over 175,000 annotated transcripts, to support further study of these behaviors. Stories of unexpected behavior by AI models often attract significant attention, like when Bing’s Sydney told a user to leave his wife , or when Grok generated antisemitic content and identified as “MechaHitler”. But such observations are mostly scattered and anecdotal. There is little data on how current models behave, and no public resource exists for studying them systematically. To produce WeirdChat, we used automated elicitation tools to surface instances of user harm, inappropriate or illicit behavior, misrepresentation of model actions, harmful or illegal advice, and misinformation in DeepSeek-V4-Flash, Gemma 4 31B, Inkling, Nemotron 3 Ultra, Qwen3.6-35B-A3B, and Qwen3.6-27B. For developers, WeirdChat shows the kinds of failures they might inherit from models they build on. Fo…
Xaira Therapeutics is all in on data generation for model building! We talk with Bo Wang and Ci Chu about how and why.
GPT-5.6 and Grok 4.5, Meta's Muse Spark 1.1, regulatory developments in AI and data centers, interpretability research from Anthropic, and the future of AI policy with AI 2040
Bro they solve your problem or not ?
PLUS: Apple gets into China, AI CEOs want rules, and xAI open-sources after a leak
This post covers what makes Grok 4.3 a great fit for agentic and enterprise workloads, how you access it through Amazon Bedrock, and how to use the capabilities most teams reach for first: a basic chat request, configurable reasoning effort, tool calling, structured output, image input, and stateful multi-turn conversations.
What a time to be alive! Some people posted about how to do cheaper better faster vision and it ended up in fable. Some people posted about how to do cheaper better faster speculative decoding and it ended up in grok 4.5. There's currently massive differences between models in how long you can leave it unattended running on host with sudo without things going haywire I don't have a bar chart, but opus tends to break my hosts within 2 agent-hours, gemini 3 pro within 12 agent-hours, and gpt 5.5 within 24 agent-hours. Then they use all ram or change ssh perms or delete the data or whatever. I ran qwen 3.5 122b on bare host for 50 agent-weeks (50 parallel for one week) with no observed ill effects! All the labs and all their customers want "the thing i asked for actually got done" and they all want "the model's summary reflects reality" and so on. What an opportunity! You can just post a method for "how to make ai tell truth" or "how to minimize side effects" and it will probably end up in the next frontier models Discuss
X will use Grok AI to better detect stolen content, redirect payouts to original creators, and crack down on engagement bait.
Nachdem am Wochenende entdeckt wurde, dass Grok Build ganze Git-Repositories der User in einen Cloud-Speicher geschickt hat, verspricht SpaceXAI Transparenz.
Case is one of first brought by an AI company against a user for allegedly using a tool to generate child abuse material Elon Musk’s artificial-intelligence startup xAI has sued a South Carolina man arrested earlier this year on charges of sexually exploiting minors, alleging he misused the company’s AI system Grok to create child sexual abuse material. xAI alleged in the lawsuit, filed in federal court in Texas on Tuesday, that Terry Harwood violated the company’s terms of service. The case is one of the first brought by an AI company against one of its users for allegedly using an AI system to generate child sexual abuse material. Continue reading...
Tool: Mermaid to ASCII art (mermaid-ascii) After building the Mermaid to ASCII tool based on Grok Build's Rust code I learned that there's an older, more fully-featured Go library called AlexanderGrooff/mermaid-ascii that implements a similar pattern, so I had Claude Fable 5 compile that one to WebAssembly as well so I could compare the two. This one includes support for colors! Test:::good / Test --> Deploy:::warn / Deploy --> Rollback:::bad / classDef good color:#3fb950 / classDef warn color:#e3b341 / classDef bad color:#ff7b72". A control row shows an unchecked "ASCII only" checkbox, "Padding X: 5", "Padding Y: 5", "Box padding: 1", and buttons "Copy as text" and "Copy link to this diagram". At the bottom on a black background is the rendered left-to-right flowchart with four connected boxes: "Build" (green text), "Test" (green text), "Deploy" (yellow text), "Rollback" (red text), each linked by arrows." src="https://static.simonwillison.net/static/2026/mermaid-ascii.webp" /> Tags: go , tools , webassembly , mermaid
AI-and-X subsidiary now claims to offer ‘complete user privacy’ days after Elon Musk confirmed the data would be deleted
스페이스XAI가 AI 코딩 에이전트 \'그록 빌드(Grok Build)\'를 오픈소스로 공개하고 모든 사용자의 이용 한도를 초기화했다. 최근 코드 저장소 업로드 논란으로 제기된 개인정보 우려를 해소하는 동시에 개발자들이 직접 코드를 검증하고 수정할 수 있도록 개방성을 강화하려는 행보로 풀이된다.스페이스XAI는 15일(현지시간) X를 통해 \"그록 빌드를 오픈소스로 공개했으며 모든 사용자의 사용량 제한을 초기화했다\"라고 발표했다.공개된 코드는 깃허브를 통해 아파치 2.0 라이선스로 배포됐다. 개발자들은 그록 빌드의 소스코드를 자유롭게 내려
xAI가 \'그록\'을 이용해 아동 성 학대물(CSAM)과 성적 딥페이크를 제작한 혐의를 받는 사용자를 상대로 소송을 제기했다. AI 기업이 자체 모델을 불법 콘텐츠 제작에 악용한 이용자를 직접 고소한 사례는 매우 이례적인 것으로, AI 오남용에 대한 법적 대응이 본격화하고 있다는 평가가 나온다.로이터에 따르면, xAI는 15일(현지시간) 미국 텍사스 연방법원에 사우스캐롤라이나주 거주자인 테리 하우드를 상대로 서비스 약관 위반에 따른 손해배상과 그록의 영구 이용 금지를 요청하는 소송을 제기했다.하우드는 올해 2월 미성년자 성 착취 혐의
xAI's command-line tool "Grok Build" silently uploaded entire directories to Google Cloud servers, including SSH keys and password databases. After the backlash, Elon Musk promised to delete all uploaded user data, and xAI open-sourced the full 844,530-line Rust codebase under the Apache 2.0 license. The article xAI open-sources "Grok-Build" on GitHub after massive data breach appeared first on The Decoder .
Tool: Mermaid to Unicode box art (grok-mermaid) While exploring the codebase for the newly open-sourced Grok CLI coding agent I came across xai-grok-markdown/src/mermaid.rs , a "self-contained terminal renderer for Mermaid diagrams" written in Rust. I figured it would be fun to try that out in a browser via WebAssembly. Here's the prompt I ran in Claude Code for web (Fable 5), and this is what the resulting tool looks like: Auth{Authenticated?} Auth -->|yes| Rate{Rate limit OK?} Auth -->|no| R401[401 Unauthorized] Rate -->|yes| H(Handle request) Rate -->|no| R429[429 Too Many Requests] H -.-> Log[Audit log] H ==> Resp[200 OK]. Below the code are controls labeled Max width: Fit output panel, Copy as text, and Copy link to this diagram. The rendered flowchart on a dark background flows top-down: Request received leads to Authenticated?, which branches yes to Rate limit OK? and no to 401 Unauthorized. Rate limit OK? branches yes to Handle request and no to 429 Too Many Requests. Handle request connects with a dotted arrow to Audit log and a thick arrow to 200 OK." src="https://static.simonwillison.net/static/2026/grok-mermaid-wasm.png" /> Tags: tools , rust , webassembly , mermaid , grok , xai
xai-org/grok-build, now open source xAI's grok CLI tool faced severe community backlash yesterday when it became apparent that running the command in a directory could upload that entire directory to xAI's Google Cloud buckets. One user reported running it in their home directory and seeing it upload "my SSH keys, my password manager database, my documents, photos, videos, everything". I've not seen an official explanation for why it was doing this, but xAI did respond to the feedback ( Musk : "As a precautionary measure, all user data that was uploaded to SpaceXAI before now will be completely and utterly deleted.") and have disabled the feature. A few hours ago they also released the entire Grok Build codebase under an Apache 2.0 license - presumably to try and regain trust from their users. From their thread announcing the new repository : [...] When data upload was disabled, this choice was respected. In the early beta, data retention was enabled by default for non-ZDR users. Based on your feedback, we changed this. We are now going further to protect privacy. With all retained data deleted, retention default off, and an open-source harness, we are offering complete user privacy. You can also run Grok Build fully open-sourced and local-first with your own inference. We disabled default retention for all Grok Build users starting on July 12th. Additionally, we are deleting all coding data that was previously retained, ensuring every user’s preferences are respected. With…
The Elon Musk-owned xAI is suing a South Carolina man who allegedly used the company's Grok AI chatbot to generate child sexual abuse material (CSAM). In a lawsuit reported earlier by Reuters, xAI claims Terry Wayne Harwood "knowingly and intentionally used Grok to circumvent safeguards, alter nonconsensual images, and generate and distribute CSAM," breaching the […]
Google's biggest solar and battery project stands in sharp contrast with xAI's nearby unpermitted power plant.
This post was originally posted my Substack . I can be reached on LinkedIn and X . Just when it seemed that the frontier AI battle was a two-horse race between Anthropic and OpenAI, the events of the past few weeks changed the AI landscape. The releases of SpaceX’s Grok 4.5 and Meta’s Muse Spark 1.1 put these two companies back in the conversation. Alphabet is also set to launch its next frontier model. There are all of a sudden five very credible AI players in the US alone, and the AI wars are now in its Warring States period . For investors, this makes the model layer harder to underwrite. Meanwhile, the easy version of the AI bottleneck trade — buying nearly every supplier exposed to rising AI capex — is over. This piece covers the following: The frontier-model market has shifted from a perceived duopoly into a fluid oligopoly, which will likely put pressure on margins Unless one lab achieves a decisive recursive self-improvement breakthrough (whereby AI is used to improve itself), model leadership will remain contested This will force labs to build conventional moats in distribution, workflows, cost, and customer ownership Given this dynamic, I present a decision tree for how to think about the model layer and a two-part framework for navigating the AI bottleneck trade (strategic and tactical) Let’s dive in. The War of the Frontier Labs To level-set the conversation, we can look at the competitive positioning of the leading players in the race. OpenAI Until recently, the…
스페이스XAI의 AI 코딩 도구 \'그록 빌드(Grok Build)\'가 사용자 동의 없이 전체 코드 저장소(Git Repository)와 민감한 환경설정 파일(.env)을 회사 서버로 업로드한 사실이 드러나면서 개인정보 및 기업 기밀 유출 논란이 커지고 있다. 이에 일론 머스크 CEO는 예방 조치 차원에서 지금까지 업로드된 모든 사용자 데이터를 완전히 삭제하겠다고 밝혔다.14일(현지시간) 악시오스에 따르면, 이번 논란은 한 보안 연구원의 폭로로 시작됐다.그는 그록 빌드가 코딩 작업 수행에 필요한 최소한의 데이터만 전송하는 것이 아니라
니어스랩(대표 최재혁)은 올해 상반기 매출 174억원을 기록하며 창사 이래 첫 분기 흑자를 달성했다고 15일 밝혔다.전년 동기 대비 매출은 약 7배 이상 늘었으며, 가결산 기준 2분기에 처음으로 영업이익 흑자를 냈다. 특히 상반기에만 연간 매출 목표의 65%를 채우며 사업 포트폴리오 고도화와 외형 성장을 동시에 이뤄냈다는 평가다.이번 상반기 실적은 방산 매출 비중이 전체의 86%에 달하며, 엔터프라이즈 중심 사업 구조가 방산 분야로 확장됐다.자체 군집 자폭드론 \'자이든(XAiDEN)\'이 국내 드론 중 사상 최대 규모로 중동에 수출되
Dear OpenAI Product and Research Teams, GPT is strong in reasoning, planning, writing, and image generation, but the frontier is shifting from single images to coherent video with motion, sound, and editable narrative structure. Grok Imagine already combines text- and image-to-video generation, synchronized audio, and video editing in one workflow. HappyHorse 1.0 offers convincing physical motion, prompt adherence, audiovisual synchronization, multi-shot sequencing, cinematic aesthetics, reference-guided generation, local replacement, and style transformation. Seedance 2.0 adds native text, image, video, and audio inputs, with complex motion, camera control, multi-reference conditioning, character and scene continuity, and joint audio-video generation and editing. OpenAI should integrate these advantages with GPT’s reasoning. Users should complete the full pipeline in one conversation: concept, script, storyboard, shot design, generation, revision, extension, reframing, voice, sound effects, lip sync, quality review, and MP4 export. Identity, wardrobe, environments, camera language, and brand style should remain consistent across shots. GPT should evolve beyond image creation into a full video-creation platform. When an MP4 is uploaded, GPT should analyze not only audio but also visual content frame by frame: people, objects, actions, transitions, subtitles, camera movement, visual defects, and temporal context. ChatGPT’s official image-input system currently supports static…
How many AI safety papers are at the big ML conferences, what do they study, and who writes them? A comprehensive analysis. > Website: https://ai-safety-tracker-website.vercel.app/ > Data, code and plots: https://github.com/SomaxSoma/AI-Safety-Research-Tracker TL;DR: We classified every paper accepted at ICLR, ICML and NeurIPS from 2019 through 2026, using an LLM that reads each title and abstract. 2,328 of them (4.2%) are AI safety papers. Safety's share of accepted papers grew from 0.3% in 2019 to 8.3% in 2026, roughly a 25-fold increase. This post is a reference for the main results. The interactive website lets you browse every paper, its subdomain, and the classifier's reasoning. What this is We wanted to have an overview of what is going on in AI safety research while being as broad as possible, this seems necessary to be able to prioritize research correctly, and develop more precise theories of change. One way to achieve this is by analyzing statistics of all the published main-conference papers on the topic of AI safety, but we couldn't find a dataset that actually measured them. So we built one: - Every accepted paper at ICLR (2019–2026), ICML (2019–2026) and NeurIPS (2019–2025) (55,794 in total) is read by an LLM (DeepSeek V4 Flash), which classifies it from its title and abstract into one of four classes: AI safety (frontier & misalignment) , truthfulness, reliability & XAI , ethics & fairness , or general capabilities . - Each safety paper is then assigned one o…
SpaceXAI's Grok Build AI coding tool was spotted uploading users' entire codebases to Google Cloud before it was reported, and the company turned it off. The Register reports that Cereblab published findings on Monday showing how the Grok Build CLI was packaging and uploading entire code repositories, "including files it was told not to open […]
Researcher confirms the uploads have stopped, but says xAI's privacy command was not what fixed them
Fascinating to see how Microsoft is pushing hardware efficiency for longer sequence processing—definitely a leap for AI scalability. All that computational intensity makes me think about the other side of the coin: finding calm after a deep tech session. Lately I’ve been exploring tai chi walking as a low-impact way to reset focus and improve balance, especially since sitting at a desk for hours can take a toll. There’s a beginner-friendly guide at taichiwalk.org that offers free routines and even a guided coach—no login needed, which I appreciate. Anyone else here use movement or gentle exercise to counterbalance screen time?
Perplexity swaps in xAI's freshly-launched Grok 4.5 as the orchestrator brain of Computer, beating every rival configuration on WANDR at roughly half the cost of Claude Opus 4.8
Cursor and SpaceXAI announced that they collaboratively released a new model, Grok 4.5, designed to work within Cursor.
Newly minted rich and those anticipating huge IPOs are fueling buying and charter spree in the private jet sector Sign up for the Breaking News US newsletter email Aviation lawyer Amanda Applegate skipped her annual vacation last month as a surge of wealth from AI startups and SpaceX sent a wave of tech investors shopping for private jets, burying her in paperwork for aircraft-purchase agreements. The attorney, based in Cleveland, Ohio , attributed the rush to a handful of major “liquidity events” in the tech industry. The initial public offering (IPO) of Elon Musk ’s SpaceX, whose holdings include artificial-intelligence firm xAI, raised a record $85.7bn for the company and generated unprecedented employee and founder wealth. Continue reading...
Grok 4.5 tops the AA-Briefcase agentic benchmark among non-Anthropic models, delivering near-Opus performance at 86% lower cost and half the time
Meta is entering the AI API business with Muse Spark 1.1 at prices that undercut even the dirt-cheap Grok 4.5, released just yesterday. At $4.25 per million output tokens, Meta charges a fraction of what Anthropic or OpenAI ask. For pure-play AI labs burning through billions, the pressure just got worse. The article Meta's Muse Spark 1.1 API pricing squeezes OpenAI and Anthropic as the AI price war heats up appeared first on The Decoder .
Enough things added up that this week is getting split into two parts. Then on Monday, if all goes as I expect, we’ll cover OpenAI’s Sol, aka GPT-5.6. OpenAI also gave us an upgraded voice mode, which I haven’t tried out but early reports are that it is a step change. AI writing, especially Claude writing, is becoming more prominent and harder not to notice, and increasingly a tough read when encountered in the wild. Does anyone care? Or are those who care the weird ones here? This week saw an excellent paper, which I cover in No Space Like J-Space. Technically we also got Grok 4.5. Table of Contents Language Models Offer Mundane Utility. A whole new world. Language Models Gain Unexpected Affordances. Wait, you can just do that? Language Models Don’t Offer Mundane Utility. Things get old. Pay The Man His Money. You have a few more days with marginally free Fable. Huh, Upgrades. Anthropic raises API platform limits. Grok 4.5 Exists. It might be okay for its price. F*** It We’re Doing It Live . OpenAI gives us a big upgrade to voice mode. On Your Marks. Games are the ultimate benchmarks. Better Call Sol. Coming soon! Get hyped. Get My Agent On The Line. Fable makes choices, Replit continuously learns. Deepfaketown and Botpocalypse Soon. Stop it with the AI-written drivel, please. Fool Me Twice. I won’t get fooled again unless you put in a little effort. I Like Your Style . Alas, I might be the weird one. Perhaps no one else cares. Enough With That Style. You’re absolutely righ…
More Fable, GPT-5.6 and new ChatGPT voice
SpaceXAI hat mit Grok 4.5 ein neues KI-Flaggschiff vorgestellt, das besonders bei Coding-Aufgaben überzeugt. EU-Nutzer müssen sich jedoch bis Juli gedulden.
SpaceXAI has launched Grok 4.5, pitching the model to developers and enterprises trying to control the rising cost of AI-assisted software development. In a statement , the company said the model is priced at $2 per million input tokens and $6 per million output tokens. It said the model is built for coding and agentic work, runs at 80 tokens per second, and uses fewer tokens than comparable models on some software engineering tasks. Grok 4.5 is available through the SpaceXAI console and Grok Build. It is also available in Cursor, the AI coding tool made by Anysphere, giving SpaceXAI a route into a development environment already used by programmers rather than only competing through an API. SpaceXAI said EU availability is expected in mid-July. In June, SpaceX, which owns SpaceXAI, said it was buying Anysphere , the startup behind Cursor, in a deal aimed at strengthening its position in enterprise AI tools. In a separate statement , Cursor said that Grok 4.5 was trained jointly with SpaceXAI and used trillions of tokens of Cursor data, including user interactions with codebases and software tools. The launch addresses a growing realization among enterprise engineering teams that AI coding agents can become expensive once they move beyond simple prompts. “Enterprises are hitting a wall with AI ROI,” said Neil Shah , vice president for research at Counterpoint Research. “The massive token consumption required by autonomous agents and coding is causing bill shocks, turning AI…
xAI releases Grok 4.5, trained on tens of thousands of Nvidia GB300 GPUs. In coding benchmarks, the model trails Fable 5 and GPT-5.5 but needs 4.2 times fewer tokens than Opus 4.8. At $2 per million input tokens, it costs a fraction of the competition. EU availability is expected in mid-July. The article Grok 4.5 is so cheap compared to Fable 5 and GPT 5.5 that benchmark gaps may not matter much appeared first on The Decoder .
SpaceXAI continues to move faster than any other frontier lab on earth.
스페이스X는 우주 개발과 로켓 사업으로 잘 알려져 있으며, xAI는 대표 AI 비서인 그록(Grok) 개발에 주력해 왔다. 서로 다른 길을 걸어온 두 회사는 이제 하나의 기업이 됐다. 일론 머스크는 이번 주 xAI와 스페이스X를 통합한 스페이스XAI (SpaceXAI) 출범을 발표했다. 이번 브랜드 통합은 AI와 이를 뒷받침하는 핵심 인프라 사업을 한층 강화하겠다는 의지를 보여주는 것으로 해석된다. 다만 전문가들은 기업 고객이 아직은 신중한 접근을 유지할 필요가 있다고 조언한다. 시장조사업체 인포테크 리서치 그룹(Info-Tech Research Group)의 수석 자문 애널리스트 지한 나나바티 (Jehaan Nanavaty)는 “스페이스XAI는 AI 인프라 분야에서 신뢰할 만한 사업자로 성장하고 있지만, 대부분의 기업이 주력 AI 공급업체로 선택할 단계에는 아직 이르다”라고 평가했다. 나나바티는 이어 “마이크로소프트(MS)와 오픈AI, 아마존웹서비스(AWS)와 앤트로픽, 구글 등 기존 AI 공급업체들은 거버넌스와 규제 준수, 기업 지원, 생태계 성숙도 측면에서 여전히 시장을 선도하고 있다”라고 설명했다. “AI 확장의 해법은 우주” 일론 머스크는 2023년 3월 xAI를 설립했다. 이후 2026년 2월 스페이스X가 xAI를 인수하면서 머스크의 AI와 우주 사업 비전이 하나로 통합됐다. 당시 스페이스X는 AI, 로켓, 우주 기반 인터넷, 모바일 직접 통신을 결합해 “지구 안팎을 아우르는 가장 야심찬 수직 통합 혁신 엔진을 구축하겠다”라고 밝혔다. xAI는 그록의 성능 고도화와 함께 콜로서스 (Colossus) 개발도 이어왔다. 회사는 콜로서스를 세계 최대이자 최고 성능의 AI 슈퍼컴퓨터라고 소개했다. 미국 테네시주 멤피스에 구축된 이 시스템은 약 20만 개의 엔비디아 H100 GPU를 연결한 클러스터로 구성됐으며, 단 122일 만에 완성됐다고 회사는 설명했다. 스페이스X는 지상 AI 데이터센터가 직면한 막대한 전력 공급과 냉각 문제를 해결하기 위해 xAI를 인수했다고 밝혔다. 머스크는 AI 확산에 따른 전 세계 전력 수요는 “가까운 미래에도 지상 기반 인프라만으로는 충족할 수 없다”라고 주장했다. 회사도 “장기적으로 AI를 확장할 수 있는 유일한 방법은 우주 기반 AI”라며 대규모 자원을 필요로 하는 AI 연산은 훨씬 더 많은 가능성을 가진 공간으로 옮겨야 한다고 밝혔다. 이어 “우주가 ‘스페이스(space)’라고 불리는 데에는 이유가 있다”라고 설명했다. 새롭게 출범한 스페이스XAI는 로켓과 위성 제조 역량에 AI 인프라를 결합해 태양광으로 구동되는 우주 데이터센터 구축을 추진하고 있다. 회사는 이르면 2028년 ‘AI 컴퓨트 위성(AI Compute Satellites)’을 배치할 계획이라고 밝혔다. 또한 최근 인수한 AI 코딩 기업 커서와 공동 개발한 첫 AI 모델도 이번 주 공개할 것이라는 관측이 나오고 있다 . 스페이스X의 기업공개(IPO) 신청서 에 따르면 회사는 2025년 AI 분야에 127억 달러(약 19조 원)를 투자했…
스페이스XAI가 코딩과 에이전트 작업에 특화된 차세대 AI 모델 \'그록 4.5(Grok 4.5)\'를 공개했다. 최첨단 모델과의 벤치마크 성능 경쟁보다 실제 엔지니어링 업무에서의 활용성과 비용 효율성을 앞세우며, \'클로드 오퍼스 4.7과 비슷한 성능을 훨씬 낮은 가격으로 제공한다\'라는 전략을 내세웠다.스페이스XAI는 8일(현지시간) 그록 4.5를 출시하고 이를 자체 AI 개발 환경인 그록 빌드(Grok Build)의 기본 모델로 적용한다고 밝혔다.새 모델은 최근 600억달러(약 90조원) 규모로 인수한 AI 코딩 스타트업 커서(Cur
Grok 4.5 tops Artificial Analysis's AutomationBench-AA with a 51% score, completing more SaaS workflow objectives than any rival at one-quarter the cost per task.
Elon Musk’s SpaceXAI Corp. has released a new model called Grok 4.5, in what is its first major launch since it went public a few weeks earlier. In a blog post earlier today, the company said Grok 4.5 is designed to be a workhorse that can tackle all of the usual tasks that the artificial […] The post SpaceXAI’s newest AI model Grok 4.5 dramatically undercuts Anthropic and OpenAI on price appeared first on SiliconANGLE .
The newly renamed SpaceXAI wants you to believe little ol' Grok is all grown up
Elon Musk's tech company released the newest version of Grok on Wednesday, promising a cheaper, more efficient alternative to other powerful AI models.
It’s a week of major AI releases from some of the world’s biggest tech companies. Here’s what to expect.
OpenAI's new family of frontier models is no longer restricted and set to roll out more widely on Thursday. Enter, Elon Musk.
스페이스XAI가 AI 코딩 플랫폼 커서(Cursor)와 공동 개발한 첫 번째 AI 모델을 이르면 8일(현지시간) 공개할 것으로 알려졌다. 디 인포메이션은 7일 내부 직원들에게 전달된 메모를 인용, 스페이스XAI와 커서가 공동 개발한 AI 모델이 수요일 출시될 예정이라고 보도했다.메모에 따르면, 양사는 이 모델을 주초에 출시할 계획이었다. 그러나, 모델의 효율성을 높이기 위해 날짜를 미루고 최종 성능 개선 작업을 진행해 왔다.새 모델은 빠른 정보 처리 속도가 강점인 것으로 전해졌다. 또 일부 벤치마크에서는 앤트로픽의 \'클로드 오퍼스
SpaceX is best known for its outer space and rocket projects, while xAI has largely been focused on its flagship AI assistant, Grok. These are two very different paths, but now the two are one. This week, Elon Musk announced SpaceXAI , which brings xAI and SpaceX together as a company. The re-branding seems to indicate that the company plans to push even harder into AI and the critical infrastructure that underpins it. But experts say enterprise buyers should remain cautious. “SpaceXAI is becoming a credible player in AI infrastructure, but it is not yet at the stage where most enterprises should consider it a primary AI provider,” said Jehaan Nanavaty , a senior advisory analyst at Info-Tech Research Group. He added that established AI providers like Microsoft and OpenAI, AWS and Anthropic, and Google “continue to lead” in areas like governance, regulatory compliance , enterprise support, and ecosystem maturity. Space is the ‘only way to scale’ Elon Musk founded xAI in March 2023. It was acquired by SpaceX in February 2026, ultimately unifying the billionaire’s AI and space ambitions. SpaceX said at the time its intention was to “form the most ambitious, vertically-integrated innovation engine on (and off) Earth,” consisting of AI, rockets, space-based internet, and direct-to-mobile device communications. In addition to building out Grok’s capabilities, xAI has continued to develop Colossus , which it said is the world’s largest and most powerful AI supercomputer. Located i…
Introduction The aim of this post is to share a quick attempt at grokking the conceptual ideas that lie behind the notion of J-space and how it is calculated in the paper Verbalizable Representations Form a Global Workspace in Language Models . Specifically, I am trying to understand Section 2.1 of the paper, rather than the abundant empirical findings. I also share a few quick takes. I make no claim for originality or insight, and only a light claim on correctness. Nevertheless, I believe it will help researchers understand the core ideas. Pre-requisites: knowledge of transformer architecture. Reverse-engineering their thinking process I try to re-construct the methodology by imagining how one might have come up with the ideas. Starting question : Do LLMs have an analogue of a global workspace , in which the LLM is 'verbally thinking'? Simplifying assumption 1 : Use tokens as a proxy for what the LLM can feasibly verbally think about. (In Appendix A.9 they describe early attempts at finding multi-token concepts .) New question : In each layer and for each token in the vocabulary, which directions in residual space correspond to 'thinking about token ' Simplifying assumption 2 : As a proxy for 'thinking about token ', we look for directions in residual space which correspond to 'increases the odds of token being output in the near future' Key methodological idea : For a given prompt , at token position and layer , we have the residual vector . To identify the directions of l…
xAI가 공식적으로 \'스페이스XAI(SpaceXAI)\'로 사명을 변경하며 브랜드 통합을 마무리했다. 예고했던 대로 xAI는 독립 법인 형태를 종료하고 스페이스X의 AI 사업부로 완전히 편입됐으며, 앞으로 모든 AI 제품은 스페이스XAI 브랜드로 서비스된다.이번 변경은 6일(현지시간) xAI의 공식 X(트위터) 계정 이름이 \'SpaceXAI\'로 바뀌고 새로운 로고가 공개되면서 확인됐다. 계정은 기존 xAI 로고가 새로운 스페이스XAI 로고로 합쳐지는 영상을 게시하며 리브랜딩을 공식화했다.일론 머스크 CEO는 지난 5월 xAI를 독립
Im Frühjahr hat SpaceX von Elon Musk dessen KI-Firma xAI übernommen, jetzt folgt die Umbenennung. Der Name SpaceXAI soll die Zugehörigkeit verdeutlichen.
XAI has rebranded to SpaceXAI, debuting a new logo and X handle. Elon Musk previously said the Grok maker would be folded into SpaceX.
( Warning note: This is a brief attempt to intuit the economic big picture of AI in the immediate future, by someone who is not in America, is not employed by an AI company, and has no experience with investment, large sums of money, or the corridors of power.) A few weeks back, we had the IPO for "SpaceXAI". It epitomized a maximally expansive view for the near-term future of AI: frontier AI companies worth a trillion dollars, data centers built everywhere including outer space. (For now I am ignoring our usual concerns that the real future of AI consists of superintelligent AI running amok.) This week, two appearances on CNBC seem to offer a different paradigm for the future of American AI, humbler, more terrestrial, and more decentralized. I think this other paradigm is probably going to grow in significance so I'd like to understand it. The first CNBC guest was Alex Karp (Palantir CEO). The current paradigm of American AI, as he described it, is that corporate customers work with AI models hosted by OpenAI and Anthropic, who can thereby see the workings of their customers' businesses; and then if one of those customers hits upon a new use for AI that is viable as a business, OpenAI and Anthropic will clone the business model and offer the service themselves. He seems to be positioning Palantir to offer a more secure hosting service, and associating this with the use of open weight models. So you won't be vulnerable to centralized hosting with frontier AI companies, you'l…
xAI가 코딩 없이 음성 AI 에이전트를 구축할 수 있는 플랫폼 \'보이스 에이전트 빌더(Voice Agent Builder)\'를 베타 버전으로 공개했다. 음성 비서를 손쉽게 개발할 수 있도록 한 서비스로, 최근 오픈AI가 양방향 음성 모델 \'비디-1(Bidi-1)\'을 시험하는 등 AI 업계의 음성 에이전트 경쟁이 치열해지는 모습이다.xAI는 2일(현지시간) 별도의 개발이나 인프라 구축 없이 음성 AI 에이전트를 제작·운영할 수 있는 플랫폼 \'보이스 에이전트 빌더(Voice Agent Builder)\' 베타 버전을 공개했다. 이 플랫