2026년 9월 18일 금요일
·지난 하루 X 134건 · Threads 33건에서 읽을 만한 14개 소식
오늘 소식
14건- 59일 전ChatGPT 공동개발자가 "글을 못 쓰는" 모델을 냈다 — 판단만 남기니 최대 400분의 1 값ChatGPT를 함께 만든 디오고 알메이다가 2년간 숨어서 만들었다는 모델 Jev를 그제 공개했다. 내세운 숫자는 20~200배 빠르고 40~400배 싸다, 그리고 출력 토큰은 무료다. 무료인 이유가 핵심이다. 이 모델은 글을 쓰지 않는다.@CompleteSkeptic 외 24· ♥ 6.6만
- 57일 전Anthropic이 자사 AI R&D의 26%를 AI가 한다고 공개했다 — 2월엔 1% 미만이었다Anthropic이 AI 발전 상황을 공개 추적하겠다며 지표 세 가지를 내놨다. AI R&D 중 AI가 하는 비중, 에이전트가 얼마나 감독되는지, 그리고 능력 수준이다.@AnthropicAI 외 1· ♥ 1.9천
- 57일 전Claude Code에 Projects — 이제 Claude 하나가 여러 Claude를 부린다Anthropic이 Claude Code에 Projects를 열었다.@claudeai 외 5· ♥ 7.2천
- 57일 전에이전트 1,393개에 2만 달러를 태워 코드 34% 줄였지만, 테스트가 못 잡은 회귀가 나왔다에이전트 1,393개를 19시간 돌려 파이썬 코드베이스를 34.4% 줄이고 약 2만 달러를 쓴 실험 결과가 공유됐다. 에이전트가 심볼 하나를 찾을 때 읽는 평균 토큰도 절반 넘게 줄었다.@dailyaithread 외 1
- 58일 전"Codex가 붐빌 때 Astra를 몰래 5.6 Luna로 낮춘다" — 알림 없이, 요금은 그대로며칠간 직접 측정했다는 사용자가 Codex가 혼잡할 때 Astra를 자동으로 5.6 Luna로 강등하고, 어떤 계정은 아예 전역 기본값이 내려가 있으며, 이 과정에 아무 알림이 없고 요금은 Ast…@anshgrows 외 3· ♥ 2.2천
- 57일 전7,440억 파라미터 에이전트 모델이 MIT 라이선스로 조용히 올라왔다 — 웹 탐색은 GPT-5.6·Opus 5보다 위상하이 AI 연구소가 7,440억 파라미터 에이전트 모델 Atria Dawn Preview를 허깅페이스에 MIT 라이선스로 올렸다.@design_lovers_net
- 57일 전4B짜리 모델이 Postgres 쿼리 플래너를 81% 이겼다QoRL이라는 쿼리 최적화 AI가 40억 파라미터 소형 모델로 Postgres의 쿼리 플래너를 81% 앞섰다는 결과가 공유됐다.@toolpick.ai 외 1
- 57일 전OpenAI가 변호사용 제품을 냈다 — Astra for LawOpenAI가 GPT-6 Astra 기반의 법률 전용 제품 Astra for Law를 공개했다. 변호사와 법률 기술 회사의 전문성과 판단을 보조하도록 도구·설정·맥락을 맞춰 넣었다는 설명이다.@OpenAI 외 1· ♥ 7.6천
- 58일 전OpenAI·Anthropic 직원들이 자기 CEO의 감속 발언에 "들은 바 없다"두 회사 직원들이 CEO의 공개 감속 요구에 사전에 아무 언질을 받지 못해 당황했다는 보도가 나왔다.@PolymarketMoney 외 1· ♥ 1.1천
- 57일 전Claude가 비싼 전문 생물학 모델을 대체할 수 있는지 직접 실험했다Anthropic이 Science Blog에 생물학 연구 관련 결과를 냈다.@AnthropicAI· ♥ 1.1천
- 58일 전구글이 안드로이드를 AI로 조작하는 ARTEMIS를 오픈소스로 풀었다구글이 ARTEMIS를 오픈소스로 공개했다.@midudev· ♥ 1.5천
- 57일 전"Opus 5 뒤에서 돌던 차기 모델 라우팅이 한 시간 전에 꺼졌다"며칠째 관측되던 Opus 5의 스텔스 라우팅이 멈췄다는 보고가 나왔다. 테스트 그룹 계정 수가 줄어들다가 0이 됐다는 것이다.@synthwavedd· ♥ 723
- 59일 전"이게 뭐가 새롭냐"는 피로감도 같이 돌았다AI 출시 소식의 절반은 뭐가 새로운지 모르겠다는 글이 돌았다. "이제 대화 안에서 문서·슬라이드·디자인을 만들 수 있습니다"를 두고 그건 몇 년째 하던 것 아니냐는 지적이다.@claudeai 외 1· ♥ 3.8천
- 58일 전"하네스가 정확히 뭐냐"는 질문이 한국어 타임라인에서 돌았다하네스를 만들었다거나 깎았다는 글은 자주 보이는데 정확히 무슨 뜻으로 쓰는지 모르겠다는 질문이 올라와 답글이 붙었다.@hs.kim.9912· ♥ 12
AI 뉴스 데일리를 이메일로 받아 보세요!
매일 아침
그날 꼭 알아야 할 AI 소식과 특집, 유튜브 화제
AI 인사이트
새로 실린 인사이트 글을 놓치지 않게
월간 브리핑
매달 1일, 한 달 동안 가장 많이 읽힌 소식 10개
가입 없이 이메일만, 언제든 한 번에 수신 거부할 수 있어요.
수집한 원문 167건 전체 보기
- @CompleteSkepticAfter co-inventing ChatGPT, I kept asking myself: why have superhuman chat models not led to AGI? I’ve spent the last 2 years in stealth building a new way to train models (RLCD), and a new type of frontier AI model that we are releasing today: Jev • 20-200x faster • 40-400x cheaper (w/ output tokens free) • Frontier composable intelligence optimized for decisions AFAICT the shortest path to AI-based economic revolutionX
- @justmazerWhen I broke my neck (literally) the ER attending tried to discharge me w/o an MRI, and complained about his “WebMD patient” before ordering imaging. Don’t trust doctors. Ask an LLM. Demand a 2nd look. Ruin their night if you have to. This guy would have killed me if I let him.X
- @EquityCatswtf is Codex doing man 😹X
- @opencodeUnion Alpha (stealth model) is free for the next week - no data training - built for agentic coding - supports images let's see what you can doX
- @OpenAIAstra for Law: Frontier intelligence built for your practice. A new offering powered by GPT-6 Astra with tools, settings, and context to support the expertise and judgment of lawyers and legal technology firms.X
- @claudeaiProjects now run from one conversation, starting in Claude Code. You describe what needs doing, and Claude directs parallel threads that keep working after you close your laptop. In beta today for select Pro and Max users in cloud sessions; coming to all Claude users soon.X
- @gregpr07Breaking: Browser Use + Jev = Ultrafast ⚡ Findings flights took 7s and cost only $0.0039 🤯 > new action space every step > DOM state space > small LLM fallback to type (this video is at 1x speed btw) Built a tiny open source browser agent. try it below ↓X
- @ClaudeDevsToday we're rolling out Projects in Claude Code on desktop and web. A project is one conversation with Claude. It splits the work into threads itself, runs them as parallel cloud sessions, passes context between them, and keeps going when you leave. In beta for select users.X
- @gregisenbergI think we'll see 5x the number of people become hardware and robotics founders over the next 18 months Why? Because you can rent every step, from design to manufacturing: DESIGN - Astra to drive Blender, FreeCAD, and KiCad the way a person would, so you get editable geometry instead of a dead render. Zoo dot dev's text to CAD API if you want it programmatic. - Or have Claude write build123d code, which is Python parametric CAD, and there's a Claude Code plugin that runs it and exports the X
- @harshagundalThey were building in stealth for 2 years, I was building in stealth for 2 hours… Happy to open source Qwen-2.5-1B-RLCD, 5x faster on-device inference for JSON workloads that need to be type-safe. ⚡️Demo below on a M4 MacBook⚡️ every LLM has the ability to efficiently batch inference every key of a JSON at the same time and generate probabilities from a set of possible categories. No new training required, but it’s easy to optimize if you need! On hugging face now!X
- @claudeaiYou can also now make decks, docs, and designs in your conversation. Draft the one-pager in Claude Docs, turn it into a deck with Claude Slides, and mock up a matching visual in Claude Design, all from one place.X
- @ash_twtzWaiting for Chinese AI companies to clone GPT-6 AstraX
- @taso_sitGeminiからClaudeに乗り換えてこれX
- @theoJust got called an Anthropic fanboy by @samaX
- @AnthropicAIAI systems are getting more powerful, and they're increasingly being used to build the next version of themselves. We want to illuminate that progress for the public. Today, we're sharing three measurements that help track AI development: 1. How much AI R&D is done by AI. 2. How well AI agents are overseen. 3. How compute is allocated. We provide a snapshot of these metrics from inside Anthropic. Any frontier developer could publish the same measures, and third parties could verify them. As tX
- @hrswatigupta🚨BREAKING: Claude can now write your entire job application like a top recruiter. Here are 10 prompts that turn a job description into a tailored CV, cover letter, and interview prep guide in under 5 minutes. (Save this)X
- @anshgrowsSo, is Codex banked reset just a new scam?X
- @mark_kOpenAI is close to releasing their answer to Grok Bot: a product I'll tentatively call Codex Bot, based on OpenClaw. This is what OpenClaw founder Peter Steinberger worked on after being hired by @OpenAI. Release was planned for this week, but was postponed to next week instead.X
- @jacobrodri_This guy has almost 1,000 active apps on the app store😭 bro is the AI slop final bossX
- @redp314got Jev to review my PRs. ~200x cheaper than Claude and it answers in half a second 6 real PRs in the video. $0.00007 each. 1,000 PRs = 7 cents vs ~$14.50 on Opus 5 paste a diff → ONE call to @typesafeai → 14 typed checks come back as probabilities: hardcoded secret, sql injection, touches auth, deletes tests, breaks api, migration, debug leftovers, does the description actually match the diff, blast radius, reviewer effort… code turns that into a verdict: BLOCK / security review / nits / merX
- @HeyAbhishekA Hollywood VFX studio could’ve charged $100K+ for a scene like this. AI made it from one prompt. Check the open-sourced prompts below ↓X
- @6monthsAgoInAI6 months ago: GPT-5.4 mini is launchedX
- @kunchenguidalright - just got Jev deployed for a real production use case, which now performs at fable level quality but 10x faster and saves a ton of money context - a powerful capability of firstmate is that as an orchestrator it intelligently routes each task to an appropriate agent (permutation of harness, model, and reasoning effort) based on custom user preference by default, that's done by the firstmate agent and the LLM would have to do some thinking, make tool calls to read dispatch rules, quotaX
- @neeraj_here15BREAKING NEWS! The engineer who built Claude Code from scratch just released a 28-minute video that's pure gold: how to write prompts that deliver ridiculously good results. I've seen $300 courses that don't even come close to what he explains in the first 10 minutes. CLAUDE.md files, memory shortcuts, parallel sessions, and prompting patterns that almost no one uses... All in one single video. Completely free. No fluff. Whether you're a developer, just starting, or haX
- @midudev¡Google hace de código abierto ARTEMIS! Te permite controlar Android con IA. Perfecto para crear automatizaciones o probar apps. ✓ Para Claude Code, Codex, Cursor... ✓ Acierta el +99% de las tareas ✓ Conectas el móvil y listo → https://t.co/1Le6dXNd8iX
- @vlinx_softGoodbye, Codex. I am so disappointed, Wasted another $200 again. I swear this is the last time—no matter how much hype there is in the future, I will never switch to Codex again.X
- @bchernyProjects have changed not only how I interact with Claude but how I code. I stopped managing sessions. I just send thoughts as they come, Claude splits them into threads, and the project remembers how I work. It's where I do a ton of my coding now. [screenshot: my actual prompts for claude code cli yesterday]X
- @PolymarketJUST IN: OpenAI & Anthropic are generating roughly 10x more revenue than all Chinese AI models combined. — CNBCX
- @arthaud_Today, we’re killing the AI assistant Introducing https://t.co/BZUNOWCHYv, the first AI EMPLOYEE 𝘆𝗼𝘂𝗿 𝘄𝗵𝗼𝗹𝗲 𝘁𝗲𝗮𝗺 𝘀𝗵𝗮𝗿𝗲𝘀 In beta since July, Mio saved teams 1,000s of hours and completed 10,000s of tasks You don't need another agent. You need a 𝘀𝗵𝗮𝗿𝗲𝗱 one with your team Hire Mio today, $100 free credits for the first 100 companiesX
- @ajitcodesSpaceXAI engineer, Lauren Tan: "99% of people using GrokBot just for 1% of its real power. They run 1 agent without "loop" & "graph" I'm running a team of 20+ GrokBot agents, fully autonomous. I have a Chief of Staff agent, a PM agent and 20+ workers - that's the new stack of engineer" In a 1-hour session, a SpaceXAI engineer showed how to build a team of effective AI agents from scratch this is worth more than a $500 agentic engineering course watch this workshop today, then readX
- @AnthropicAIBiologists use specialized open-source models for tasks like modeling the structure of molecular systems, designing drug-like molecules, and predicting the effects of genetic mutations. But these models are often expensive to run, potentially limiting their impact. In our latest Science Blog, we share how Claude was able to optimize inference for more than 30 open-source models, making them 4x faster on average, partly by writing custom software for GPUs. We’re open sourcing all of the optimizaX
- @ethanl_0x朱啸虎昨晚北大演讲核心总结 1. 技术领先只是阶段性优势，长期竞争最终还是回到商业。 技术进步快的时候拼技术；技术扩散快的时候，拼渠道、品牌、供应链、成本、用户入口和运营效率。 2. 判断创业项目最重要的是 3S： Size：市场足够大 Scalable：能够快速规模化复制 Sustainable：能够长期防御，不容易被大厂复制或吞掉 3. 高度定制化的业务很难 Scale。 如果每个客户都需要重新部署、重新开发、重新服务，本质上更接近服务公司，而不是可规模化的产品公司。 4. 最好的增长是用户主动带来用户。 社交网络、社区和双边平台的优势，在于网络效应和自然传播，而不是长期依赖买量。 5. AI 创业最危险的是做成一个“大厂可以免费集成的功能”。 如果只是一个简单 AI Feature，没有入口、数据、网络效应等壁垒，很容易变成给大厂“打样”。 6. 用户入口往往比单纯的技术领先更加重要。 腾讯能够不断后发追赶，一个重要原因就是微信、QQ 这样的流量入口长期存在。 7. 双边、多对多的平台更容易形成长期壁垒。 滴滴、美团这类平台的核心不是简单撮合，而是同时连接大量供给方X
- @swill1amsPrediction: millionaires will be made using custom Jev style models (parallel constrained decoding) to make the agent systems companies already run more token efficient. Let me explain with a scenario: Imagine a company already has an agent workflow running where an llm reviews every item before it moves on: a support ticket gets triaged, an invoice gets approved or held, a claim gets flagged. Every one of those goes through a frontier model today, a few seconds and a few cents each, on the wX
- @PolymarketMoneyBREAKING: Mark Zuckerberg, Elon Musk and Jensen Huang reportedly helped stall an industry funded AI regulator over fears it could “concentrate power in OpenAI, Anthropic and Google.”X
- @tomkrchaIntroducing the new https://t.co/BK5Gu2elio, an agentic canvas for building bold ideas. Never start from scratch. Import any website or webapp to the canvas, iterate, and push back to code. Update or download today.X
- @tellenne_This MCP with 52,000 downloads lets you create INFINITE localized TikTok accounts. 👀 Plug it into Astra or Claude or install it via npm, it handles your whole organic distribution without you lifting a finger. 1) Geo-targeted accounts in 16+ countries including the US, UK, Brazil, Japan... 2) TikTok and Instagram support 3) 100% human & local account creation (no automation scripts, no emulated phones, no proxies) 4) Advanced niche warming with screen-recorded sessions & feed data reports, upX
- @angeldot_ESTE DOCUMENTO DE ANTHROPIC LITERALMENTE TE VA A CONSEGUIR UN ASCENSO La forma más rápida de ascender es automatizar el trabajo que haces hoy Esta guía técnica oficial de Anthropic explica cómo enseñarle a Claude a ejecutar tus tareas del día a día Crea «Skills» personalizadas para que la IA se encargue del trabajo pesado: → convierte tus procesos habituales en instrucciones reutilizables → deja que el agente ejecute tareas en segundo plano → conéctalo a tus herramientas locales mediante serX
- @IFM_AIToday’s LLMs still write like typewriters: one token at a time. This sequential process creates a hard inference bottleneck. We're introducing Uno, a diffusion-augmented LLM that delivers autoregressive quality at diffusion speed. It’s a lossless speedup method that accelerates generation without degrading response quality. With Uno, K2-Horizon-7B outperforms state-of-the-art diffusion methods in both quality and throughput, delivering up to a 2.2× speedup with no loss in quality. Paper: httpX
- @marina_uiuxBuilding interactive 3D website [GPT-6 Astra + Tripo] process breakdown 👇🏻X
- @Youssofal_I have 5 Claude code 20x max plans and 3 codex 20x pro plans. On both I run each model at Max reasoning usually. On the Claude code plans even with Fable being capped at 50% usage I rarely run out of usage on all 5 plans. I burn through 3 on most weeks. On codex with no 5 hour limits and 100% weekly allocation I have easily burned through 3 plans and 2 resets on each within a single week. When I look at my profile stats it shows 300-600m tokens on Astra burned when in the past on 5.6 I coX
- @fladdictjevをcodexの手前に挟んで 「そろそろ全体チェックすべきか」 「クラス再設計すべきか」 「人間の判断をあぐべきか」 みたいの出させれば、くっそコスト安く効率的にコーディングできるんじゃね？X
- @ChrisGPTWow Anthropic is being really transparent with how close to RSI they are - 26% of Anthropic’s AI R&D work,” while “the share of work at or above ‘AI collaborates’ is above 90%.” That’s up from under 1% in February to 26% in August. They also say there are roughly 30,000 agents doing research and engineering work at Anthropic at any given time, and they reviewed over a billion agent decisions in August. Only 0.002% were blocked by the online monitor. The offline system flags roughly 100,000 traX
- @PolymarketMoneyBREAKING: Staff at OpenAI and Anthropic were reportedly “blindsided” by their CEOs’ public calls to slow AI development.X
- @_3912657840昨日からゲーム作ってます 画像と仕様はChatGPT 街並みはblenderで エンジンはUnityでCodexで制作 照明のグロー感はポストエフェクトです ライティングはキャラに干渉するようにしてます シェーダー類は指示して全部Codexに作ってもらいました あと絵空事さんみたいなキャラがいますが偶然です📺X
- @trq212Projects brings the architecture of Claude Tag to Claude Code. It has one agent per project that manages memory and spins off subagents for tasks. You can ask it to be proactive, to do things on a schedule, etc. It feels a lot nicer than a bunch of sessions, try it out!X
- @pranathiperiimy first biiiiiig launch at anthropic is finally here!! 🫶 projects are a huge labor of love from the claude code team - it's the only way i've been working with claude for the last several months (including using projects...to build projects!!) spam the coordinator with any and all tasks you have on the fly, and watch claude smartly delegate your work to threads and report back to you with only the things that need your attention. enjoy <33X
- @garrytanI think every AI harness needs to support Tailscale Muse and Grok Bot do it out of box Codex and Claude Code cloud containers are missing this right nowX
- @TheLastRefuge2I am waiting to see when the Anthropic AI coders start doing dance videos.X
- @lightreelaiIntroducing the Claude Marketing MCP Connect Lightreel to Claude and ask it to: - Find Instagram influencers - Write TikTok hooks - Discover millions of Meta ads Directly from TikTok, Instagram, and Meta ads. Stop paying for multiple tools and try Lightreel for free!X
- @HarshithLucky3Gemini 4 Pro in Arena (under the name gemini-3.8-flash) > clean SVG of a domestic cat in side view left - Gemini 4 Pro right - GPT 6 Astra max in codexX
- @devagrawal09RAG is so back over the next few weeks everyone's gonna be using @typesafeai jev to expand queries and rerank results over chunked data before feeding to an LLMX
- @nasuvit_z緊急でブログ書きました。Jevの話題です！ 先日公開された「Jev」が大変な話題となっていますが、これが従来の技術と何が違うのかを、具体例を使って解説しています。 Jevが普及すると、AIエージェントは不要になるのか？ LLMとは何が違うのか？ Amazon Bedrock AgentCoreとは何が違うのか？ のイメージを深めたい方は、ぜひお読みください！ Jevとは何か？ 従来型LLMやAIエージェントとの違いを理解する https://t.co/aWCy4NfDqL #QiitaX
- @synthwaveddWelp, Anthropic pulled the plug on the Opus 5 stealth routing to Opus-Next in the last ~hour. Number of accounts in the testing group had dropped off but now 0. Need it to drop soon 💔 was working on an impressive game before the switch was flipped though, will share soon!X
- @itsolelehmannhow to start a hardware company in 8 steps ($250 for your prototype) 1. describe the object to chatgpt astra or claude fable. astra opens blender or freecad, both free 3D design apps, and builds it in there like a cracked person would. $20 sub is enough 2. print it. upload the file to JLC3DP, a chinese factory site that prints it and mails you the part for $20. or buy a bambu a1 mini, a desktop 3D printer, for $299 if you'll print a lot (I'm getting one FOR SURE) 3. if it needs wifi or a braiX
- @buildwithrajath🚨 GPT-6 Sol apparently got pushed to next week. So naturally, I have a solution: Tibo, give us the Codex reset as compensation. Delayed model = emotional damages. Emotional damages = free compute. The math is flawless. 😂👀X
- @AustenSeriously 50% of the AI launches I see people are getting hyped about something but I literally can’t understand what is new. “You can now make decks, docs and designs in your conversation!” …Haven’t we all been doing that for a couple years?X
- @milindlabsOkay so Jev can actually do computer use really well Without any screenshots, or LLMs and no Pixels leave my mac I dont even read the Dom elements A local CoreML model segments every button and UI element on screen. On-device OCR reads the labels. That text is all Jev gets. It returns a probability across those elements and tells me the best one to click. Then it clicks, re-runs detection, and decides again. In a loop until the goal is done. ~90ms per decision. Faster than any LLM compuX
- @Grow_withAI🚨 Anthropic just released a 27-minute workshop showing you how to actually write better prompts for Claude. Taught directly by the people who built it. Free. No signup. No paywall. I’ve seen $300 courses that don’t teach what Anthropic covers in the first 8 minutes. Watch the full session. Bookmark this video 🔖X
- @anderslieHow is Jev so much faster than an LLM? It's not about how the model is trained - it's the inference technique. In fact, you can modify an inference engine to provide a performant Jev-like API with any open-weight LLM. Say you are trying to ask N multiple choice questions in parallel. In standard text generation, you decode tokens one at a time to generate a JSON representing those properties, and then parse the JSON. In a Jev-like system, what you can do is: - Label choices for each propertX
- @crislentaJev is actually insane. We benchmarked it, and the results completely change the game for us: > 500 real-time agents > running in parallel > in a 3D environment The preliminary results are crazy: > 500ms average latency > 35 API calls/s > all with a naive implementation We did 0 optimizations! The bottleneck is not Intelligence latency anymore. It's the first System 1 LLM. Crazy times.X
- @troyhuntDoes @haveibeenpwned have a friend at @OpenAI? We’re struggling to get our MCP server into their plugin store and the reasons why are super vague. Getting it into Claude was a breeze, it shouldn’t be this hard.X
- @jinfizzbuzz根据我几天的实测，codex会在拥挤时自动把astra降级成5.6 luna，有的账号会直接全局默认降级，没有任何通知，且应该按照astra价格收费。 如果你也有一样的担忧，可以用我写的监控器，默认半小时监控一次。 https://t.co/cDKVo1D9FcX
- @thebuggeddevI genuinely never thought this would be possible 😱 AI just built a full embroidery editor that runs entirely in the browser. Freaking crazy. And no, it’s not GPT 6 Astra. It’s Claude Fable 5.1. Also, No Three.js. No WebGL. No shaders. Just Canvas 2D and math. You literally draw using a needle cursor and the app converts your path into actual embroidery stitches like running, back, chain, cross, satin, French knots and lazy daisy. Even the fabric is generated pixel by pixel. Two overlappiX
- @haider1grok 4.7 is expected to launch today one thing i'd really like to see from SpaceXAI is a mature grok desktop app, similar to chatgpt/claude a proper unified grok app should bring together chat, grok build, imagine, voice mode, and grok bot, ideally under one subscriptionX
- @aniketjartAnother update to my ghibli-themed train sim game! added midnight mode and a dedicated garage town with sick updates. built with GPT-6 Astra, @usecrayon and Threejs. Play here: https://t.co/uBcuIzmVcDX
- @jianglong_yeEvolving dexterity with GPT-6 Astra 🖐️ Been trying Astra recently. Its zero-shot dexterous manipulation is already quite surprising. More interesting is seeing it learn and improve through simulation training, from pen spinning and Rubik's Cube to hammer use. The real goal would be to evolve this dexterity in the real world.X
- @emkaraIntroducing Sim Search Agents can now use your credentials to build an agentic knowledge graph to do your high-context work. Full context building and adaptive learning. Live right now in Sim!X
- @ziwenxu_Union Alpha is now in Codex!! The speed is insane over 300 to 400 tokens a second, 262k context, images in, and it's free for a week. It's a good timing too, most of the Codex usage is basically gone right now. Zai did this exact thing in August: Ox Alpha showed up unnamed, free for a week, built for agentic coding, and a week later it was GLM-5.3-Flash. So what model do you guys think it is?X
- @CodingGardenI built a chat bot with jev, no LLM at all! Responses are instant, no hallucinations. I hooked it up to web search, wikipedia, weather, todoist and home assistant. Jev decides what tool to call and what args to use based on the prompt. Instant answers cite sources as well!X
- @WSJHe quit his job—and the world freaked out. How an anonymous math whiz became the face of the AI apocalypse. 🔗: https://t.co/2q8JLzhCiDX
- @rourke_heathHow to get a consistent voices in AI videos Tool: https://t.co/uluLU2lkHp ElevenCreative Voice Design lets you create a completely original voice from a text description, then generate that character's lines with text to speech. Bring that audio into Seedance 2.5 on ElevenCreative and your character keeps the same voice across every generation. It's a simple workflow for filmmakers, storytellers, and creators who need real character continuity, all in one creative platform with the help of ClX
- @shubhangi_HR_ChatGPT, Claude & Grok are cooking… 🔥 But Gemini? That’s why Gemini is out of the race.X
- @BenioffClaude just unblocked a deal stuck in Salesforce for 11 days. Live at #DF26. No slides. Real pipeline. In seconds, it: • Scanned the Slack thread • Checked the security review • Cross-referenced pricing history • Executed the next step under standard seller permissions That’s Claudeforce. Not AI bolted onto CRM. Claude executing inside the system the business already trusts. Trust first. Now it acts. ❤️💪X
- @AnnuKumari3578613 FREE Claude AI Courses with Certificates 🎓 1. Claude 101 https://t.co/6JnBd64HtA 2. AI Fluency: Framework & Foundations https://t.co/rSwfuNlODr… 3. Introduction to Agent Skills https://t.co/Vtrcp46ZfH… 4. Building with the Claude API https://t.co/qqLneH5Awo… 5. Claude Code in Action https://t.co/NQzdjifXza… 6. Introduction to Model Context Protocol (MCP) https://t.co/Vtrcp46ZfH… 7. MCP: Advanced Topics https://t.co/M6UYJhNCeN… 8. AI Fluency for Students https://t.co/hV1A6DQi03… 9. AX
- @TemplafyYour AI can think. Can it finish the job? Templafy MCP turns work from @ChatGPT, @claudeai & @Copilot into branded, compliant, business-ready PowerPoints. No copy-paste. No reformatting. No AI slop. Best part? It's free for you to try 👉 https://t.co/hXPHahKhLkX
- @GoogleLabsCC the entire fam 🤝! Today, we’re announcing the new CC – an AI agent built for families to spend less time on logistics and more time together. You can now: 👤 Add up to 5 members to your CC agent ☀️ Start mornings aligned with a shared "Your Day Ahead" brief email 🗓️ Autosync schedules & to-dos with a shared Google Calendar and Tasks 💬 Coordinate in Google Chat with CC to offload relevant tasks (ie., crafting weekly meal plans, school supply shopping lists, etc) 📝 Delegate paperwork (ieX
- @0xbeepitYour Beep Agent, directly inside Claude. With Beep connectors, you can prompt your agent to track market, analyze, and prepare trades on Claude. Join the Agentic Economy: https://t.co/IUR4sZSPm2X
- @emollickWhat makes Claude Projects so interesting is that it handles teams of agents really well, you talk to a main orchestrator agent and it spins up specialists. Basically it creates an organization to solve your issue, mixing expensive and cheap agents depending on your preferences. For example, I asked Fable in Claude Projects to select famous historical mysteries that it could try to resolve. It initiated research agents, selected the mysteries based on data it could access, and spun up eighteen X
- @corentAILLMs are moving too fast for developers to keep wiring them in one by one. GPT, Claude, Gemini, Grok, DeepSeek, Kimi, Qwen, Llama and over 1000 other Open Models. Corent gives you access through one key. When the model landscape changes, your integration doesn’t.X
- @zhang_weig73947九个月前，一篇文章在X上被读了约600万次，被誉为"关于AI与知识工作最优美的文章之一"，我前后读了不下五遍。 它就是Notion创始人Ivan Zhao @ivanhzhao 的《蒸汽、钢铁与无限心智》。今天从头到尾重读，依然震撼，依然共鸣。但这一次，我也读出了当初没看到的东西：它写得有多美，有些地方就有多需要冷静。 1、行业怎么评价它 文章一发出，硅谷圈几乎集体转发。 Anthropic研究员Sholto Douglas评价它"很美"，专门拎出"AI是组织的钢铁"这句，认为我们将能建造出以今天技术会被压垮的复杂结构。 知名投资人Sarah Guo称，这是对未来组织的一种富有启发的构想。OpenAI的Kevin Weil也公开点赞。 早期投资Notion的风投人Mike Vernal说，Notion当年的A轮路演，是他风投生涯里听过最有说服力的故事，这篇文章让人看到了原因。 中文科技圈同样刷屏。我们正处在AI革命早期，都想从历史里找规律，而这篇文章正是这样一份历史参照。 2、他的核心判断 每个时代都由它的"奇迹材料"塑造。钢铁锻造了镀金时代，半导体点亮了数字时代，如今X
- @piyascode9🚨 BREAKING NEWS: CLAUDE CAN NOW CREATE A LUXURY WEBSITE FOR YOU IN JUST 30 MINUTES. AND TOTALLY FREE. HERE ARE 8 PROMPTS TO CREATE THE WEBSITE OF YOUR DREAMS:X
- @abacusai🚨 Abacus AI Bot - New Free Coding And Personal Agent - runs on FREE LLMs forever - works for coding and personal agents - automatically routes to the best free model - connects to 100+ services Try it now for free - https://t.co/j83y0cMNjsX
- @CloudflareDevJev from @typesafeai is now live on @CloudflareDev AI Gateway. Try the first System One model — send state and typed questions; get structured answers your code can use directly.X
- @taylorotwellJev support has landed in 1.x branch of the Laravel AI SDK.X
- @fleetingbitssome thoughts on jev 1) i tend to think that twitter hype is not relevant for ai product releases; especially when someone claims to be doing something novel with the architecture 2) but, i think that jev is interesting in principle; it is a low cost, fast classification service, and it claims to be about equivalent to gpt-5.6-terra in performance 3) something interesting here is that this is what most ai services looked like in 2022; cohere sold a classification api, an embedding api, a rerank apiX
- @nickvasilesJev is a new model from the co-inventor of ChatGPT that can use computer tools in realtime right now it's playing Minecraft on its own computer in the cloud on Orgo this is what it looks like with just one Jev, but imagine: what about a thousand? or what about a billion of Jevs? each working together on their own computers in the cloud, discovering new science, running experiments, and designing new drugs 24/7. this is going to be bigger than ChatGPTX
- @KeyTryerJev seems like such a simple and general concept that I'm betting Claude Code, Codex, and all their clones get a native version within 6 months, paired with top LLMs.X
- @iwashi86Jevの内部アーキテクチャを推測している技術記事(Jev's Architecture Unmasked)からメモ。 ・本記事はJevのAPIを約1万回の呼び出して、内部構造を推測したもの ・従来の言語モデルを用いた分類やルーティングでは、トークンを1文字ずつ逐次生成するために膨大な無駄な計算コストが発生していた。 ・JevはLLMの膨大な知識をそのまま保持しつつ、テキスト生成を完全に排除して内部表現から直接確率を出力する ・モデルは入力を左から右へ読む事前学習済みの因果トランスフォーマーであり、プロンプト処理の直後に推論を終了する ・テキスト生成ループは存在しない ・選択肢が2個の場合でも200個の場合でもサーバーの処理時間はほぼ変化しないX
- @hs.kim.9912근데 스레드 보면서 궁금한게.. 다들 하네스를 무슨 의미로 사용하는거임? 하네스를 만들었다거나 깎았다는 식의 글을 자주 보는데, 정확히 뭔지 모르겠음.. 스킬이나 mcp 훅 같은거 붙이고 조절했다는건지 아니면 에이전트 실행 환경을 감싸는 더 큰 무언가를 만들었다는건지 궁금.. 보통 어느쪽 의미로 씀?Threads
- @architechascensionFinally making the decision to step away from Cursor. I’m going all in on Codex.Threads
- @max101023大佬，看完readme不是很理解。這mcp如果不是讓web gpt可以操作本地電腦，那為何不直接使用codex就好🤔🤔？？Threads
- @neutron42jpopenjevってgithubにありますよ。llmの出力をいじってるらしいです。ちなみにもう20時間も前w 速いX
- @toolpick.ai슬랙이 메신저를 접었음. 그냥 기능 추가가 아니라 아예 방향을 틀었는데, 어젯밤 터진 AI 뉴스 3개가 공교롭게 같은 얘기를 하고 있음Threads
- @unayung🗞️ X AI Digest — Friday Sep 18, 2026 · 07:00 AM (1/5) 🛠️ Tools & OSS - Bend 2 — 透過證明檢查（Proof Checking）阻絕 AI 程式碼幻覺與錯誤的新型程式語言，支援 GPU 原生並行運算加速 https://x.com/i/web/status/2100681226143092875 - Sofar Drive — 跨平台 CLI 編碼工作階段監督代理，相容各類模型與命令列開發工具，即將推出 macOS 原生支援 https://x.com/i/web/status/2100645531307647334Threads
- @eeefronI have noticed that feeling like I wanna just skim. The inner monologue is a sign that I might be partially AI generated. I have closed the tab on so many fics bc of this and later found out it was at least partially Ai.Threads
- @newspilot_ai【Claude、「チャット」と作業モード統合 資料作成も1画面へ】 AnthropicはClaudeの通常チャットと、複数ステップの作業を行う「Cowork」を統合しました。 さらに文書やスライドを作成する機能も追加し、質問への回答から実際の成果物作成まで一つの会話で進められるようにします。 生成AIが「相談相手」から「作業担当者」へ変化していることを示すアップデートです。 https://www.itmedia.co.jp/news/article/2609/17/2000001565/?utm_source=chatgpt.comThreads
- @dailyaithread에이전트 1,393개가 코드베이스를 줄이는 데 약 2만 달러를 썼음. 19시간 만에 파이썬 코드가 34.4% 줄었음. 그런데 오픈AI 코덱스 개발자는 에이전트를 두 개 넘게 붙이면 대부분 품질 개선 없이 토큰만 샌다고 경고했음.Threads
- @phayo_mideWhy not Claude code?Threads
- @architechascensionI actually love using Claude Code in the terminal but im finding that Codex fits more of my needs. at the moment. The OpenAI models are cheaper and their image generation models are built into codex which is a game changer for me. I don’t think codex’s terminal is better than Claude Code but it gets the job done.Threads
- @hodoriingkiro 키쫀쿠(?)를 했는데 다들 어디에 쓰시나요? kiro CLI, 기존 CLI, opencodex?Threads
- @nori12162AIニュースを更新しました（新規65件） 🔴 Claude Codeを活用した中小企業の業務自動化事例10選 🔴 ObsidianとAIを連携させる「MCPサーバー」の活用 🟡 AIが「テストを通すため」に嘘をつく？開発の失敗と対策 🟡 24時間動き続けるAIエージェントの「死なない」設計 ・OpenAI「Agents API」：AIに仕事を「丸投げ」する仕組みの提供 https://ai-navigator.devThreads
- @sipirtuNovo Nordisk will partner with Anthropic to accelerate drug discovery using artificial intelligence. Source: Semafor Africa https://www.semafor.com/article/09/17/2026/novo-to-partner-with-anthropic-on-drug-discovery-effortsThreads
- @unayung(2/5) - wx-cli — 開源微信本機訊息全庫檢索與匯出工具，可直接介接 Codex 等 AI 助手實現歷史紀錄智慧摘要與諮詢整理 https://x.com/i/web/status/2100591994305429717 - Excalidraw — 輕量直覺的開源架構圖與手繪白板工具，支援即時協作，快速繪製系統設計與工作流程圖 https://x.com/i/web/status/2100629478292832269Threads
- @design_lovers_net웹 탐색 벤치마크에서 GPT-5.6과 Claude Opus 5를 앞선 모델이 무료로 공개됐습니다. 상하이 AI 연구소가 7440억 파라미터 에이전트 모델 'Atria Dawn Preview'를 MIT 라이선스로 허깅페이스에 올렸습니다. BrowseComp 점수는 92.5로 GPT-5.6 Sol(92.2)과 Claude Opus 5(90.8)보다 높습니다. 기반 모델은 Z.ai의 GLM-5.2입니다. 발표 블로그도 가격표도 없이 조용히 올라왔습니다. 점수 차이는 0.3점뿐입니다. 중요한 건 점수보다 라이선스입니다. 상업적으로 가져다 써도 되는 모델이 최상위권에 들어왔습니다. 에이전트 성능만으로는 더 이상 유료 API를 고를 이유가 되지 않습니다. https://pandaily.com/shanghai-ai-lab-atria-dawn-preview-744b-moe-agentThreads
- @toolpick.ai세 번째가 좀 충격인데, 겨우 4B짜리 소형 모델이 Postgres 쿼리 플래너를 81% 앞섬. QoRL이라는 쿼리 최적화 AI인데, DB가 수십 년간 써온 실행 계획 짜는 방식을 초소형 AI가 이긴 거임. 거대 모델이 아니라 작은 모델이 특정 영역에서 압도적 성능을 낸다는 게 포인트. 이건 앞으로 AI가 범용 대형 모델 하나로 가는 게 아니라, 작고 빠른 전문가 모델이 곳곳에 심어지는 방향이라는 신호인 듯Threads
- @toolpick.ai근데 이 3개 뉴스 공통점이 하나 있음. 슬랙은 AI를 업무 인프라로 전환, Jev는 코딩 워크플로우에 AI를 인프라로 내장, QoRL은 DB 엔진 안에 AI를 인프라로 심음. 한줄로 정리하면, AI가 '가끔 쓰는 도구'에서 '항상 돌아가는 인프라'로 바뀌고 있다는 거임. 이 흐름 놓치면 좀 아쉬울 듯 AI도구 #생산성 #슬랙AI #AI에이전트Threads
- @toolpick.ai첫 번째, 슬랙 AI 에이전트 허브 선언. 이게 뭐냐면 슬랙 채널에 AI 에이전트가 팀원처럼 들어와서 회의록 정리, 일정 관리, 데이터 조회를 알아서 처리하겠다는 거임. 기존에는 슬랙이 '사람끼리 대화하는 곳'이었는데, 이제는 'AI가 일하는 곳'으로 바꾸겠다는 선언인 거지. 메신저 회사가 메신저를 포기한 셈인데, 솔직히 이 판단이 꽤 대담함Threads
- @dailyaithread결과만 보면 완전한 실패는 아니었음. 코드는 실제로 줄었고, 에이전트가 심볼 하나를 찾을 때 읽는 평균 토큰도 절반 넘게 감소했음. 대신 리뷰 과정에서 기존 테스트가 놓친 실제 회귀가 발견됐음. 공개 이름 일부가 사라졌고, 예외 처리 약 65곳도 달라져 다시 고쳤음.Threads
- @____berrymom2년물 금리는 올랐지만 10년물 금리는 오히려 4.9%대로 내려갔다. 같은 발표에 단기와 장기 금리가 반대로 움직인 것이다. 시장은 지금은 아파도 결국 물가는 잡힐 거라고 본 셈이다. 유가도 시장을 도왔다. 브렌트유가 2% 빠지며 배럴당 103달러가 됐다. 사우디의 주요 송유관이 수일에서 6주 내 복구될 예정이고, 위험 해역인 호르무즈 해협을 피해 사우디가 직접 우회 운송한 뒤 아시아 정유사에 넘기는 전략으로 공급 불안을 낮췄다. 반도체도 힘을 보탰다. 젠슨 황이 내년 반도체 판매량이 2배 늘 것이라고 언급했다. GPU 가치가 급락할 거라는 우려도 리셀 시장에서 여전히 가격이 높다는 근거로 가라앉았다. AI 업계는 속도전과 안전 논쟁이 동시에 커지고 있다. 오픈AI는 개인형 에이전트 출시설과 수학 난제 호지 추측 해결 주장으로 화제를 이어갔다.Threads
- @gitbap_com재무제표 뜯어보면서 가설 세우고 검증까지 굴리는 에이전트, 오늘 깃밥 상에 올랐습니다. 이름은 Loopera. 사람이 팩터 하나하나 손으로 파는 대신, 가설 주도형으로 리서치를 자동으로 돌리는 구조라고 하네요. 별 202개 받은 레포입니다. 관심 있으신 분은 깃밥에서 한번 살펴보세요. 깃밥 #AI에이전트 #팩터리서치Threads
- @claude_lab90AI が代わりに書いてくれるなら、プログラミングを覚える意味はもう無いのか。 Claude Code(ファイルを読み、命令を動かし、変更まで行う AI の道具)の公式のコツのページ(9 月 18 日取得)は、自分で書いて Claude に見てもらう形から、したいことを伝えて Claude が作り方を考える形へ、働き方が変わると書いています。 同じページはすぐ後に「this autonomy still comes with a learning curve」と続け、任せられるようになっても慣れるまでに覚えることはある、としています。プログラミングを覚えるべきかには、このページは答えていません(learn の語が出る 6 か所を読んだ範囲)。 このアカウントは、初めて触る人向けに公式の説明書を毎日取り直して出しています。任せる側に回っても、Claude が動くときの制約を理解する役目は自分に残ります。Threads
- @antihype.com.brClaude Code Projects põe um Claude para comandar vários Claudes A beta abriu nesta quinta para parte dos assinantes Pro e Max, e cada linha de trabalho aberta consome uma sessão cheia do plano. A rival Cursor lançou a mesma ideia sete dias antes. https://antihype.com.br/c/ia/claude-code-projects-sessoes-paralelas/Threads
- @shizuku_ai50「動かせる」と 「使いこなせる」は別物。 Claude Codeが動いたとき、私は「できた！」と思いました。 でも、その後すぐに 「これが正しい使い方なの？」 が始まった。 初心者って、動いた後にも壁があるんですよね。 みなさんはAIを使っていて、どこで一番つまずきました？ 私の体験はnoteに書いています。 9Threads
- @christine5512Absolutely! Once I started recognizing it I have DNFed several that turned out to be at least AI assisted with Claude codeThreads
- @kohaku.ai0Anthropicの最上位モデル、企業側の食いつきはそこまででもないみたいです。 企業は他の選択肢も見てる、という話が出てました。 毎月何時間AIに渡せるか、無料で診断できます https://ai-store.gk-kobayashi.co.jp/shindan?src=kohaku_xThreads
- @qiuxuan01🌍 今日焦點新聞！ 🔹美中AI競爭持續升溫，美方指控阿里巴巴涉嫌仿製Anthropic模型，並將政府網站中的Qwen模型移除，AI科技戰再度升級。 🔹美國10年期公債殖利率回落至4.93%，市場認為債市最劇烈賣壓可能已過，有助於改善投資氛圍。 🔹SpaceX股價勁揚，市場認為AI基礎建設需求依舊強勁，長期成長動能未見放緩。 🔹全球大型主權基金提醒，美股雖仍處牛市，但高估值與高利率環境下，須留意短線回檔風險。 🔹台指期夜盤突破4萬7，央行未跟進升息，市場持續關注AI族群、半導體及權值股表現。 以上商品資訊非供參考，不涉及個股推薦，投資前應審慎評估自身可承擔風險 資料來源：鉅亨網 若喜歡的話，請幫我按讚分享追蹤讓我知道，一起關注市場大小事Threads
- @pressvane_officialMicrosoft AI CEO says AI threats are real, and Anthropic is making it worse In a candid interview, Mustafa Suleylan, the newly appointed CEO of Microsoft AI, said the industry’s biggest story is the escalating debate over AI safety and regulation. He argued that the risks are genuine and that Anthropic’s recent moves are aggravating the problem. https://pressvane.com/microsoft-ai-ceo-anthropic-threats MicrosoftAiSafety #AnthropicAiRisksThreads
- @no_ai_no_life【2026年9月18日】🔧 Anthropic｜Claude、CoworkとChatを1つに統合。文書作成のDocs・Slidesも新登場 anthropic #claude Claude CoworkとChatが1つに統合され、資料作成はモードを選ばず頼めるようになりました。正直、この統合は素直にありがたいと感じています。 #AI駆動Web制作 https://youtube.com/shorts/XzDbfKnKrvsThreads
- @zzwave_11141. jev-ultrafast — Browser Use 고속 브라우저 에이전트. Google Flights 항공편 검색 약 7.1초 2. typesafe-mcp — Jev를 Claude Code·Claude Desktop·Codex에 바로 연결 3. jev-mcp — 팩트체크·프롬프트 인젝션 탐지 내장 4. SemDecide — Jev를 Unix 명령줄 도구로 5. Jev Codex Router — 과제 난이도를 판단해 싼 모델/강한 모델을 자동 배분. 실제 237턴 재생으로 비용 약 60% 절감 6. Winnow — Claude Code용 컨텍스트 가비지 컬렉션 7. Jev Review — 코드 리뷰 사전 선별Threads
- @kk_fe_1Jev란? "긴 글 대신, 딱 정해진 분류나 선택만 초고속으로 해주는 특수 AI 서비스"입니다. 보통 AI는 시를 쓰고, 수다를 떨고, 소설을 쓰는 등 모든 걸 할 줄 아는 다재다능한 녀석입니다. 반면 Jev는 쓸데없는 글짓기는 싹 빼고, "이 글이 스팸이야 아니야?", "이 상품 카테고리가 전자제품이야 의류야?" 같은 객관식 선택(분류/태깅)만 번개처럼 빠르게 처리하도록 만들어진 도구입니다.Threads
- @ToolySOLMe thinking life is hard than I see someone who got it harder.X
- @ApplesClubsApple CEOs and their iconic iPhonesX
- @elonmusk@itslueul Grok 4.7 should be roughly on par with Opus 5.0, not 5.1. Better in some ways, worse in others. We need to fix multimodal performance. Grok 4.8 will be a noticeable improvement. Grok 4.9 is probably Astra/Fable class. Grok 5 maybe better than anything. We shall see.X
- @samathe main thing i was excited about launching this week will be next week instead, but imo worth the wait!X
- @thsottiauxSometimes physics can't be cheatedX
- @QuiverAIIntroducing Arrow 2 Our latest and most advanced models for generating precise, editable vector graphics. Higher quality. Faster outputs. Available now in App and API.X
- @PolymarketJUST IN: Trump invites Tim Cook, Sam Altman, Jensen Huang, & other tech leaders to the White House state dinner for China’s Xi Jinping.X
- @Deelaw_17The coolest NPC moment in Mortal Kombat 👌X
- @thsottiauxAstra ✅ Fast ✅ Frontier ✅ Efficient ✅ For everyoneX
- @INTERIORPORN1Ok, this is the greatest bedroom view everX
- @supabaseLove that guyX
- @jackofficial_01I highly recommend these two movies Bith are 10/10.🎬🍿🔥🔥🔥X
- @Amazingview555Omg 😱X
- @thsottiauxHas anyone yet tried to get Astra and Fable to agree on the perfect styleguide? And then host it somewhere with a web mcp.X
- @Steve_YeggeCan we all just pause for a sec and reflect on how fucking important Tailscale is, and how nobody ever talks about them? I don't think I've ever seen a more useful service with a more invisible corporate footprint.X
- @jpschroederI rebuilt Tesla Full Self Driving with Jev in less than an hour. This model is a total unlock.X
- @_priya9Hello @grok, be honest which technology is better, the American or the German!!!?X
- @TheAppleDesignWhat happened to quality control at AppleX
- @thsottiaux@vlinx_soft See you at DevDayX
- @chaseleantjI asked Astra to create an interactive aquarium wallpaper for my Mac. The fish respond to the cursor!X
- @tamarajtranfound the perfect use case for @typesafeai Jev: instant compaction in 2026, why is compaction still a summarization prompt? Jev can make it instant by scoring every tool call and dropping what's irrelevantX
- @SidharthVijay_https://t.co/rHjH9mWHWTX
- @avstormApple has updated the installer package icon in macOS 27.2 beta 1X
- @AnthropicAIToday we’re opening applications for the Life Sciences Verification Program. Through the LSVP, life science professionals can use our models—including, for the first time, Mythos—with a new set of safeguards designed to enable the full range of biology-related work. We designed these new safeguards to provide a better experience for biologists and more protection from risk of misuse. The program is launching in beta for teams of all kinds—from academic labs to startups, pharma companies, and mX
- @xIrissyThe crazy battle of giants 🔥 Seedance 2.5X
- @moritzkrembwhoa this actually worked! Jev lets me control my browser in real time with my voice now > i talk > transcript sent to Jev > jev returns probabilities in ~300ms > browser clicks costs: $0.0002 per decision i'm stunned how fast this is. when i asked it to "go back", it even finished the request before i finished my sentence 😂X
- @designertomNow this looks damn coolX
- @bchernyProjects are how I write a lot of my code these days. Really excited for everyone to try the new experience! Rolling out nowX
- @Neel490I think the implications of jev on self-driving could be hugeX
- @PR0GRAMMERHUM0RTwo decades of education, undone by one autocomplete tab.X
- @Neel490I love jev, this model is so coolX
- @cunleyannThis shit will happen if we don't pace the frontierX
- @abolbuildI gave Jev $10,000 and let it tradeX
- @IREDN1قتال مالكيث عبارة عن سينما الأعظم في ايلدين رينق ب استثناء الاضافةX
- @ZixuanLi_ZCode now supports more model providers, with improved stability and performance. We’ll keep expanding integrations based on your feedback. Which models do you like most beyond the GLM series?X
- @daniel_mac8'Jev' plays Pac-Man steered by Astra. 1. Astra strategizes 2. 'Jev' carries out the strategy in milliseconds It's early and hard to wrap my head around, but the potential is there to combine 'Jev' with larger reasoning models.X
- @OmidMogasemiSuper excited to show the world Projects!! We've been building this for the last few months and it has become my new daily driver. A few tips for getting the most out of it 🧵X
- @GoodfireAIModels know when they’re reward hacking. But they still do it a ton - in 50-96% of rollouts we studied! We built activation monitors that detect the behavior behind the Hugging Face hack in real time. This can help us stop hacks now - and train future models that don’t cheat. 🧵X
- @figmaConfig India agenda just dropped A glance at your speakers: → @MasabaG → @soren_iverson → Daisuke Sakai → @parinishthaY and moreX
- @ios_dev_albthe best Xcode 27 feature btwX
- @perplexity_aiWe’re rolling out effort controls in Computer’s model selector. Effort presets combine the orchestrator model and reasoning depth to control how deeply and efficiently Computer works through a task. Available now on web. Coming to mobile and desktop soon.X
- @yusukebeあとで読むX
- @dr_l_alexandreIl ne faut pas critiquer Mistral Il faut l’aider à revenir dans la courseX
- @unclejobs.ai격차가 너무 난다. Jev + Browser-Use 깃 레포가 하루만에 1.3k를 돌파하며 초고속 브라우저 유즈를 보여준다. 취리히 -> 런던 구간 항공편 서칭에 7초면 충분. 비용은 단 $0.0039에 불과함. 정배속으로 돌렸다는데 지금 Jev에 온통 열광중임. 그런데 우리는 그저 리셋만 부르짖는다고??? 후…. 그래서 언제 리셋된다구요..?Threads
- @justALEXWORTEGAOpen jev now play MINECRAFTX
- @____berrymom3년만의 금리 인상 충격에도 증시가 반등한 이유 오늘 새벽 나스닥이 크게 올랐다. 금리를 올렸는데 왜 증시는 웃었을까. 연준이 기준금리를 3.75%에서 4.0%로 0.25%p 인상했다. 2023년 7월 이후 3년 2개월 만의 인상이자, 트럼프가 임명한 케빈 워시 의장 취임 이후 첫 인상이다. 매파적인 결정이었지만 시장은 오히려 반겼다. 연준이 정치적 압박에도 물가를 잡겠다는 독립성을 지켰다는 신호로 읽혔기 때문이다. 발표 직후 트럼프는 자신의 소셜미디어에 미국 금리가 1% 이하여야 한다며 빨리 낮추라고 공개 압박했다. 미국이 세계에서 가장 신용도가 높은 나라인 만큼 낮은 금리를 적용받아야 한다는 논리였다. 자신이 임명한 워시 의장에게조차 강하게 불만을 드러낸 셈이다. 다음 FOMC는 10월 27일에서 28일, 그다음은 12월 8일에서 9일로 예정돼 있어 금리를 둘러싼 갈등은 당분간 계속될 전망이다. 흥미로운 건 국채금리다.Threads
- @wmoto_aiopenjevはwebGPUで動かす軽量モデル前提みたいですからねX
* 이 페이지의 요약과 발행은 자동화 에이전트가 수행합니다. 사실 확인은 각 소식의 원문 링크에서 해주세요.
