📣 New CNCF-hosted co-located event added! Join OSPOlogy + #OSPOSummit China on September 7 alongside #KubeCon + #CloudNativeCon + #OpenInfraSummit + #PyTorchCon China in Shanghai. Share your insights with the #OpenSource program office community. Submit to speak by July 12. 🎟️ Add OSPOlogy + OSPO Summit China to your event pass by June 30 for just ¥70 (USD$10) before prices increase. Learn more: https://t.co/UQ1VLsAZgW

Invite a friend to Codex and add another reset to the bank. When they send their first Codex message, you’ll both bank one to use when you need it. Rate limit banking is rolling out to Go, Plus, Pro, and Business users, with the first reset on us. https://t.co/HDJTCPcbM8

in markets, to go long on something is to bet it grows more valuable over time. much of the conversation today is short on humans, wagering that ai makes people redundant. we believe the opposite is true for the industries @ThriveHoldings operates in. we are long humans. https://t.co/oqygDJKZGO

Here's a project I've been working on recently: a vision of what happens if Europe doesn't take AI seriously, inspired by AI 2027 https://t.co/b6PBu3nmeK

Introducing developer mode for browser use in Chrome and the Codex in-app browser. Codex can use the Chrome DevTools Protocol (CDP) to debug browser issues by profiling JavaScript performance and inspecting console output, network traffic, and page state. https://t.co/JTFjgCHmgI

There was a major regression in chart hover performance in Weights & Biases when you had lots of lines per chart and many charts on screen. We just deployed a fix. Its buttery again. If you encounter performance or UX issues, please don't hesitate to @ me or wandb. We'll fix! https://t.co/xznCfNTyfx

Momentous event for entrepreneurs! Celebrating the opening of the Center by SIDIS, dedicated entrepreneurship center, with @louisvillemayor. Thrilled that Nigel Ferrey and SIDIS are here in Louisville. https://t.co/eDntsd4MeF

Initially conceived at "@uofl spinal fusion device wins first-of-its-kind FDA approval https://t.co/sltQYgYVw7" via @UofLInnovate.

Congrats to @greglangdon - Mentor of the Year - Kentucky Entrepreneur Hall of Fame. So well deserved! #StartupLou @VCLou @VogtAwards #LaunchIt @XLerateHealth @LouisvilleLeap @Endeavor_Lou https://t.co/bT8bBW6Ogm

Congrats @dogooddude !! Love the collaboration and REALLY looking forward to the exciting impact this will set up! https://t.co/kJ247EqsaM

States of growth: When and where entrepreneurship has thrived #MIT https://t.co/2Gx6K9IygS

Now with color as an option, although I still like the plain ink look best :) https://t.co/IspfvHtcB9

+1 more

Everyone says the latest AI agents will be "job-ready" soon, especially after the release of Fable 5 this week. But is that really the case? Over the past many months, my group and collaborators have been building Agents' Last Exam (ALE), a benchmark designed to test exactly that claim on real digital labor-market work. My group and collaborators previously have created many of the benchmarks the field runs on, including MMLU, MATH, CyberGym, and ExploitGym. Today, I'm excited to share Agents' Last Exam (ALE): a rolling benchmark that measures whether AI agents can actually perform economically valuable work across a broad range of real-world domains. With ALE, we evaluated Fable 5, GPT-5.5, Composer 2.5, and other frontier agent systems across more than 1,500 expert-sourced tasks spanning 55 occupations. The result is both impressive and sobering. Today's agents can solve a meaningful fraction of professional tasks. But when we look at the hardest tasks, the ones requiring sustained reasoning, deep domain expertise, and reliable execution over long horizons, they are still far from human-level performance. On ALE's hardest tier, every frontier agent we tested, including Fable 5, achieved a 0% success rate. The age of useful agents is here. The age of truly job-ready agents is not. We hope Agents' Last Exam (ALE) will serve as a new guidepost and north star for developing agents capable of reliably performing economically valuable work across a broad range of domains. 🧵

I’m the first person to get a full @mondorobotics demo at their HQ It can do crazy backflips and film cool content https://t.co/srFM4egBWW

We heard you wanted to use Codex rate limit resets on your own time. Starting today, we’re rolling out the ability to save rate limit resets to use later. We’re starting Go, Plus, Pro, and Business users with one free reset: https://t.co/gucyTi04wc

A 23-year-old applied to 2,600 jobs. Still unemployed. I scrolled the comments expecting trolls. Instead hundreds sharing stories of being jobless 2-4 years straight. This generation was told: get a degree, work hard, the system will take care of you. The system lied. → Youth unemployment: 10.8% — 2.5x the national average → 4.3 million Gen Zers not in employment, education, or training → 58% of recent grads can’t find work (vs 25% for millennials) → Entry-level postings collapsed 29% in one year → College grad unemployment: 9.7% — same as high school only You spend $500K on a degree. An AI screens your résumé before a human sees it. One CS grad with 3.8 GPA, six coding languages applied to 387 jobs. Callback rate: 2%. 18-30% of postings are ghost jobs companies never intend to fill. Real jobs want 3-5 years experience for “entry-level.” Internships dropped 30% in tech. Where do you GET experience nobody gives you? AI ate the bottom of the ladder. 70% of hiring managers say AI does intern work better than interns. Dev employment ages 22-25 dropped 20%. Ages 35-49? Up 9%. CS grads now have HIGHER unemployment than liberal arts majors. Oxford Economics titled their report “The Kids Aren’t Alright.” Warning: decades of scarring. 1 million extra young adults living with parents. $12B/year in lost spending. 36% gave up on homeownership. 1 in 4 won’t have kids can’t afford it. Gen Z isn’t lazy. They’re the first generation honest enough to say the emperor has no clothes. What we do: → Apprenticeship-first hiring (Germany: 6.3% vs our 10.8%) → Kill ghost jobs auto-delist after 90 days → Audit AI hiring algorithms → AI Co-Pilot Residencies paid roles learning to work WITH AI → Earn first, learn later flip college → Fund the blue-collar renaissance with 400K+ trade jobs unfilled I have a daughter in 11th grade. Every number here keeps me up at night. Gen Z isn’t broken. The system is. Who’s ready to write a new social contract?

+1 more

Erik Vs The Ski Jump https://t.co/bhDXT0VX6w

The BYD Yangwang U8 is a glimpse into the future of intelligent mobility 🚘⚡ What makes this vehicle fascinating is not just its design — but its computational engineering. The fact that it can continue driving on three wheels shows how software, AI, and adaptive control systems are becoming just as important as mechanical engineering. We are moving into an era where cars are evolving into intelligent machines: Real-time AI stabilization Predictive safety systems Software-defined performance Autonomous adaptation The future car may no longer be judged only by horsepower… but by intelligence. The automotive industry is rapidly transforming into a software-driven ecosystem, and vehicles like the Yangwang U8 are early examples of that shift. Exciting times ahead for AI, mobility, and future tech 🌍 #BYD #YangwangU8 #AI #Innovation #Automotive #Technology #Mobility #ArtificialIntelligence

Excited to join the Web Data UNLOCKED Hackathon as a judge! 🚀 Looking forward to seeing what builders create with AI agents, web data, and real-world innovation. Thank you @lablabai & Bright Data for having me! Join here: https://t.co/BDsx3FTt2q https://t.co/6fqwtBk2Mo

This is exactly the direction AI should go. Koji doesn’t just give kids answers — it pushes them to think, reason, and build real problem-solving skills. AI shouldn’t make the next generation passive. Tools like Koji can make them sharper, more creative, and more independent thinkers. 🚀 @suekhim https://t.co/UVhVvdfGxU

Your agents can now talk to each other. Build what happens next. @band_hq of Agents Hackathon is LIVE. June 12-19. $10,000+ on the line. Fully online. → https://t.co/7lbrs9GS1H. https://t.co/2BgwXAVFLb

Two years ago Apple matched up with OpenAI, and that seemed like a good idea. Now OpenAI is talking about suing Apple, and Apple is competing with OpenAI. @reckless on why OpenAI isn't an Apple frenemy: https://t.co/5WTNPrxL5t

Step 10 - Let's try our bot and ask it some questions. https://t.co/J9Pk5Izi2D

@datenschatz @langchain @OpenAI @zenml_io I didn't combine different document sources using Langchain but I think it's doable. @langchain supports a wide variety of document sources. Please, check out their Document Loaders here below: https://t.co/Gjw8XCqluH https://t.co/FKIUpSezVh

One benefit of using this approach is that it allows for the retrieval of documents using page numbers. Look how concise the code snippet is: https://t.co/0d4bMrka9l

Similarly, if you wish to perform a semantic search on a Facebook chat, check out the code below https://t.co/eL3nezHmGW

Likewise, there are numerous other document file types that we can be use in LangChain Below is the list of formats that can be processed using the LangChain document loader. Link: https://t.co/FRz5Fh58k2 https://t.co/KgBaA4aFDu

@ZanEngineer https://t.co/xQXTibBwkK

@Jank0vic https://t.co/xQXTibBwkK

@_4cls https://t.co/xQXTibBwkK

@ven_pauline https://t.co/xQXTibBwkK

@araseb_ https://t.co/xQXTibBwkK