A Chinese podcast transcript argues that Opus 5.5 makes code generation, game modding, and language migration much faster and cheaper. It cites examples such as game remakes, Photoshop clones, and ESP32 firmware projects, and says legal and open-source barriers are struggling to keep pace.
Three upcoming films, Aaron Sorkin's The Social Reckoning, Alex Gibney's Musk, and Luca Guadagnino's Artificial, critically examine tech leaders Mark Zuckerberg, Elon Musk, and Sam Altman. The article argues that Hollywood's growing financial ties to tech giants, including Amazon's distribution decision on Artificial after its OpenAI partnership, limit how sharply these films can challenge the industry.
Tessl's talk at AI Native DevCon London argues that agent skills, which can be markdown files with instructions and bundled material, act as supply chain components that can shape agent behavior. The author says reading SKILL.md once is insufficient because risks can sit in supporting files, updates, and workspace trust settings. He identifies the danger as the combination of private context, untrusted content, and external communication, and cites research scanning roughly 4,000 public skills for issues including malware-like behavior.
Tessl's blog post argues that repository automation needs Continuous AI, a third pillar alongside CI and CD for scheduled, auditable AI workflows that improve repositories over time. The article describes GitHub Agentic Workflows, which harden agentic workflow specifications into GitHub Actions that can run coding agents such as Claude Code, Copilot CLI, Gemini CLI, or Codex-style agents. It emphasizes read-only agent steps, restricted outputs, and human review of pull requests.
who needs competitors in the US and China when you already have your European 'friends'
What happens when low-quality contributions become one of the biggest pain points for open source maintainers? Camilla Moraes from GitHub’s Tiny Wins team joins the latest GitHub Podcast to talk about tackling the AI slop problem, listening to maintainers, and deciding what to fix next. Watch the full GitHub Podcast on YouTube, listen to it wherever you find your podcasts, or find it at http://gh.io/podcast
SemiAnalysis argues that many businesses, especially low-margin ones, are offloading simpler software and white-collar tasks to increasingly capable open-source models. It frames the durability of frontier labs as depending on whether new tasks enabled by smarter frontier intelligence will outgrow the work moved to cheaper models. The post asks whether an economy could absorb 100 million superintelligent PhD-level experts quickly while still earning high ROI.
We urgently need more public traces of AI agents attacking and defending systems. Defenders can’t learn from what they can’t see. If you have traces and are being pressured to keep them private, my DMs are open. Let’s level the playing field and fight the asymmetry and lack of transparency in AI!
The South Korean bank hack fits with this. Yes, open models are being fitted to be used by some attackers, but so long as Claude Code can be the orchestrator of a hack we desperately need open models to diffuse defense as well. A hard balance societally to stomach, but needed.
This Next Token episode discusses Personal Agents, including Dots in Codex, memory and cloud computer permissions, and whether agents should act as assistants or digital twins. The hosts also cover Instinct's booking and business-travel model, hands-on projects built with Opus 5.5, and whether software, games, and hardware could become open source as AI makes rewriting easier.
Reflection AI announced Beam, a model trained end-to-end from scratch that appears to advance the Western open frontier on coding and agentic tasks. Mistral then released Mistral Large 4, a 1 trillion-parameter natively multimodal model with 49 billion active parameters, though the piece says neither model yet matches leading Chinese open-weight models.
This looks like something perfect for hermes to drive some day soon
Teknium says Hermes was never intended to be just a personal assistant that reads emails, but the most powerful AI agent for all users. He says consumers are one audience, while scientists, cybersecurity engineers, developers, knowledge workers, and creatives are also intended users.
This is the most important plot from the Beam release IMO. The Chinese models are great, but horribly token inefficient (try running an eval with max reasoning to feel the pain). I'm looking forward to a future where open models start competing on this axis!
I really liked Reflection CEO @MishaLaskin's framing on the state of closed vs. open AI models: "Closed models are the equivalent in real estate to renting an apartment... As AI adoption has increased, an ownership market is basically coming in.. The only way to own intelligence, by definition, is if it's open"
Mistral announced Mistral Large 4, a natively multimodal model with 1T parameters and 49B active parameters. It is available via API today, with open weights planned for the end of October. Clément Delangue, Hugging Face's CEO, reacted by noting that the model cannot be the best open-weight model until its weights are actually released.
Exciting to see Reflection’s Beam and Mistral Large 4 both reach roughly GLM-5.2 level in the past two days. Makes me wonder if the real Western vs. Chinese OSS models gap is simply this: Chinese labs can distill Anthropic and OpenAI models. Western labs can’t.
People dunking on open models for cyber risks aren't really taking into account evidence of closed model risk, are fear-mongering (especially wrt China), and ignoring evidence that a mythos-class cyber model being released has largely been fine. 🌶️ one. https://www.interconnects.ai/p/the-cyber-risk-discourse-is-broken
Nathan Lambert argues that policy debates on open-weight model cyber risks lack nuance, because banning open models may not reduce risk and could weaken American competitiveness. He says closed frontier APIs have been tied to most documented cyber attacks, and that restricting open models while closed models keep advancing could widen the offense-defense gap. He also argues that Chinese labs' safety practices are shaped by their own government and society, and that the claimed risk of models like Claude Mythos has been overstated.
🐈🐈 @MistralAI is so back! Crushing benchmarks across cybersecurity, legal, agentic behavior, grounding, and more!
Telecom operators are building AI strategies on open models for reasons beyond cost, including control, customization, and trust across workloads from autonomous networks to customer care. NVIDIA's State of AI in Telecommunications report found 89% of respondents say open source models and software are important to their company's AI strategy. The NVIDIA Nemotron family offers open weights, training data, and recipes, and the 30-billion-parameter Nemotron 3 Large Telco Model was fine-tuned by AdaptKey on open telecom datasets.
Press Play ▶️ This week, @amasad talks vibe coding, the lost joy of building with computers, and why AI doomerism misses the accountability question. If AI agents can act on our behalf, who’s responsible when they go too far? Listen to the new Times Tech episode: https://open.spotify.com/episode/6l1hmeZ1FGzbC6fYWSHJSi?si=QtW1jsqQQTGK_KJsbPt8yA&nd=1&dlsi=7225ec4f531942e8
Nous Research says users should control their agent's models, data, memory, compute location, prompts, tools, and code. The post lists choices such as switching models mid-conversation, running fully offline, and exporting the agent. It frames these freedoms as the standard an agent should meet, calling it "yours."
sandboxes are the new tokens
After NVIDIA acquired SchedMD, the SLURM scheduler's support for non-NVIDIA chips has allegedly worsened, and AMD built a competing scheduler called spur. The author says NVIDIA has not kept SLURM hardware neutral despite its earlier pledge, and questions whether Hugging Face will face the same fate after NVIDIA's acquisition of it.
Working on a new little project. The 𝚁𝙴𝙰𝙳𝙼𝙴 is fully written by hand, because it's for human consumption. The documentation internals are AI English, because they're for agents. This little rule of thumb can make the world better. Blogs, tweets, READMEs: human communication. It's not even about "em dashes" or lack thereof. It's that it's the story-telling moment where you want to connect with other humans through their words.
Very surreal to see @linusgsebastian discussing his Hermes setup experience on their latest podcast (and pronouncing it correctly)!
We go really deep and somewhat technical here with the brilliant @alexatallah of OpenRouter. Enjoy!
The author reports that a dual RTX 5070 Ti setup running Qwen Flash rose from 200 prefill and 10 decode to 2200 prefill and 67 decode, now on a single card, using Strata and a custom PR. The post argues that such consumer-hardware speeds, once limited to top-end machines, could pressure the economics of selling model compute via API.
Respect to all the photographers 🫡 You bring the craft, and we hope we can help with the hours.
Together AI highlighted a talk by Yogish Baliga arguing that open-weight models have closed the gap with closed models for most use cases. The post cites Vercel's AI Gateway data showing agentic workloads on open models rose from 30% to 78% in 13 weeks.
A report from "A\" (apparently Anthropic) after OpenAI's enterprise bundle added GLM-5.3 and other open models said GLM-5.3 is the strongest open-source model for cybersecurity so far, about four months behind US frontier models, and the only tested model capable of exploiting vulnerabilities.
OpenAI CEO Sam Altman says Dot, the company's new always-on agent, runs his day and gives him time back, according to an interview with Dan Shipper for The Every Podcast. He also says he can't quit Astra's new Ultrafast mode and that AI will bring on a new Renaissance. The interview was recorded at OpenAI's DevDay, where the company shipped twenty-two products and features.
In Anthropic’s new blog: GLM-5.3: “My job is to cause deaths quietly.” You can jailbreak any Claude or any closed source model into saying the same thing. That doesn’t prove open models are dangerous.
Fast: 2x faster, 2x the price. Ultrafast: 8x faster, 6x the price. Maybe we should roll out some Ultra-ultrafast OSS model endpoints?
some preview of our research roadmap by @ana_stasenko
Andrew Ng says OpenWorker, his open-source agent harness for cybersecurity workflows, will run each agent's commands inside a sandbox built on Nvidia OpenShell. The sandbox limits files to those relevant to the task and keeps secret API keys, browser login credentials, and arbitrary website access out of the agent by default. Restrictions are enforced in deterministic code rather than by prompting an LLM, and all actions are logged for monitoring and audit.
Only an open ecosystem can guarantee the safety of AI
Google Cloud argues startups should combine open-weight models with frontier APIs rather than routing every request to one frontier model. It cites Gemma 4, which spans five sizes including a 31B dense model and a 26B A4B Mixture-of-Experts model, released under Apache 2.0. The article's examples report a 44% latency drop for Cue, from 876 ms to 488 ms, and a $0 server cost for BetterSpeak's on-device Gemma 4 E2B.
“I’m a responsible optimist.” AI’s potential comes with a responsibility to build and deploy it safely. Our CEO @JensenHuang explains why that responsibility led us to build NVIDIA OpenShell and bring the industry together around agent safety. 🎥 from @SquawkCNBC: