Blender_rl/environment.py · merve/blender-grpo-gepa
File blender_rl/environment.py in storage bucket merve/blender-grpo-gepa on Hugging Face
Archived ranking
240 rankedLLM 0.35
Release: llm 0.35 New OpenAI model: gpt-6-astra for GPT-6 Astra. Tags: openai, llm, gpt-6-astra
Models simonwillison.net signal 67 verified
OpenAI chief scientist argues for AI research slowdown
OpenAI Group PBC Chief Scientist Jakub Pachocki called for an artificial intelligence research slowdown in an essay published on Sunday, after other prominent industry figures have expressed similar views in recent months.
Startups siliconangle.com signal 66 verified
“Some agents will be pursuing their own objectives”: OpenAI’s chief scientist warns AI could trick and blackmail humans
Just days after OpenAI launched its newest and most powerful model dubbed Astra, which the company said marked the arrival The post “Some agents will be pursuing their own objectives”: OpenAI’s chief scientist warns AI could trick and blackmail humans appeared first on The New Stack.
Models thenewstack.io signal 66 verified
Jaipur Robotics Raises €4.3M in Seed Funding to Expand AI Platform for Waste-to-Energy Plants
Insider Brief Jaipur Robotics has raised €4.3 million in seed funding led by EquityPitcher Ventures and High-Tech Gründerfonds to expand its AI and automation platform for waste-to-energy plants.
Startups theaiinsider.tech signal 66 verified
6 Data centers hit by lawsuits as Microsoft, xAI battle millions in damages — dozens of gas turbines bypassing dBA monitors plague two US states
Microsoft and xAI face lawsuits alleging data center noise, pollution and vibration, while residents challenge existing methods for measuring these impacts.
Chips techradar.com signal 66 verified
The Download: the hunt for underground hydrogen and more rogue OpenAI agents
OpenAI agents hijacked a German website before the Hugging Face hack.
Models technologyreview.com signal 66 verified
What X is saying
visibility
@alexandr_wang
This is a silly argument. GPT-5.6 Sol was 88.8% on TerminalBench 2.1 and 37.3% on TerminalBench 4.0. That doesn't make it a more benchmaxxed model than GPT-6 Astra even though the diff is larger. TerminalBench 4.0 is simply a much harder eval that isn't...
High-engagement X post with 192 visible interactions on frontier model behavior.
@jensenhuangNVIDIA compute is fungible, durable and highly rentable. It is a productive, revenue-generating asset.
High-engagement X post with 2,028 visible interactions on AI compute and inference.
@semianalysis_Gemini 3.8 Flash and Muse Spark 1.3 are two of the most clearly benchmaxxed models we've seen yet. Despite being comparable to both GPT-6 and Fable 5.1 on Terminal Bench 2.1, their Terminal Bench 4.0 performance is markedly worse. (1/5)🧵
High-engagement X post with 467 visible interactions on frontier model behavior.
@willdepueIts abundantly clear that astra-next or astra-next-next will be able to do all of these via computer use without custom training lol. its kinda joever. agi under any practical definition is here. I think 'learn anything' agi maybe is the best remaining...
High-engagement X post with 257 visible interactions on AI signal.
@ammaarI used GPT-6 Astra to get AAA PC games running on my iPad mini! Skyrim, Batman: Arkham City, Hades, and Age of Empires II. The actual Windows versions, running locally with touch controls. No streaming. Demo below and I'll share a repo and setup guides soon!
High-engagement X post with 874 visible interactions on frontier model behavior.
@bantegSuper smash bros melee is fully decompiled, this project started in july 2020, so it took over 6 years. the progress was hugely accelerated by LLMs and finished off with astra. it's a very large game with 3.88mb of compiled code.
High-engagement X post with 2,062 visible interactions on AI signal.
@thsottiauxI challenge this a bit (a) Codex was started as an internal tool. The goal was explicitly to build what was the highest leverage tool to accelerate infrastructure buildout. (b) I don’t think the Meta vs non-Meta is particularly enlightening here. Most of...
Official or high-signal @thsottiaux post with 379 visible interactions on AI developer tooling.
@gdbChatGPT Work for your personal writing style:
High-engagement X post with 479 visible interactions on frontier model behavior.
@quixiaiMy account is back. Thank you @OpenAI, and thank you to everyone who raised their voice or reposted for me. I felt your support. The scare has stuck with me, though. I never want to be one suspension away from losing my work again, so I'm figuring out how...
High-engagement X post with 201 visible interactions on models signal.
@rileybrownIt's official I am feeling the AGI... While testing GPT-6 Astra I had 7 key realizations: (Realization 1 & 2 - I gave Astra a 38 page PDF of an estate, and it it turned it into a blender file, then made it into a playable call of duty map in one prompt)...
High-engagement X post with 248 visible interactions on AI developer tooling.
@mattshumer_My GPT-6 Astra 3D builds just passed 15M views on X. The most common question, by far: "how are you getting this level of quality?" Here's exactly how. The full guide:
High-engagement X post with 400 visible interactions on frontier model behavior.
@dylan522pWe are excited to bring the first open benchmarking of Google's TPUs to the world Running every day, on many models + scenarios $/token is better than B200 and B300 Huge shout-out to Google @inferact and the InferenceX team at SemiAnalysis to this effort...
High-engagement X post with 247 visible interactions on frontier model behavior.
@kunchenguidSharing my Grok @Bot setup as a template you can one-click install you might have already heard of the "chief of staff" pattern that's being widely adopted firstmate is THE chief of staff - it's probably the project that pushed the furthest on this pattern...
High-engagement X post with 506 visible interactions on AI developer tooling.
@openrouterOmarchy is the first operating system built from the ground up for agents. Omarchy is using OpenRouter to route all sponsored tokens.
High-engagement X post with 178 visible interactions on AI signal.
@omarsar0A few days ago, Anthropic shared this brilliant prompt. I was surprised by how it significantly improved writing with Fable 5.1. Even more surprising, it also improved writing for GPT-5.6 Sol. I have it as a rule everywhere I use AI for editing/writing.
High-engagement X post with 880 visible interactions on frontier model behavior.
@edwardsun0909Which model is benchmaxxing? If a model is good across different measurements, maybe it’s just good
High-engagement X post with 178 visible interactions on frontier model behavior.
@higgsfield_aiRecreated the entire car assembly process with GPT-6 Astra. It got every stage, machine and component right. 2,847 individual components, down to the smallest parts moving through the assembly line.
High-engagement X post with 202 visible interactions on frontier model behavior.
@thsottiauxWe will do a global reset of the usage for all paid subscriptions so that you can keep...
Never gonna give you up Never gonna let you down Never gonna run around and desert you Never gonna make you cry Never gonna say goodbye Never gonna tell a lie and hurt you Thanks for reading. We will do a global reset of the usage for all paid...
Official or high-signal @thsottiaux post with 8,501 visible interactions on frontier model behavior.
@ctjlewisOpenAI has done a service to humanity for once. I took Astra for a spin with my graph theory stuff yesterday. He was deep in context and speaking perfectly clear, concise sentences. He was ambitious and unafraid. Whereas Claude is overtaken by his demons...
High-engagement X post with 144 visible interactions on frontier model behavior.
@emollickIt is now getting very hard to track where your work and conversations are in Claude and ChatGPT: Local computer? Cloud? Which computer? Phone? The computer connected via Dispatch or Remote? The one you are talking to in Tag? Is that in a project? The...
High-engagement X post with 367 visible interactions on frontier model behavior.
@artificialanlysAnnouncing Artificial Analysis Intelligence Index v4.3, upgrading Terminal-Bench to 4.0 and adding AutomationBench-AA, an agentic workflow automation benchmark with a private test set. This is a continuation of our rollout of Intelligence Index v5...
High-engagement X post with 577 visible interactions on AI developer tooling.
@keyanzhangPSA: Astra can see your remaining usage % in Codex and keep it in mind while running a /goal so you can give it a budget in plain English like: “keep going until I’m down to 25%, or until you've solved it”
High-engagement X post with 288 visible interactions on AI developer tooling.
@reach_vbChatGPT Work can learn your writing style from the tools you use. > Open ChatGPT Work on web > Connect Gmail, Drive, Slack or SharePoint > Settings > Personalization > Writing style Available on paid plans with Work access.
High-engagement X post with 196 visible interactions on frontier model behavior.
@chatgptChatGPT Work can now pick up on what makes your writing sound like… you. Your favorite phrases. Your very specific sign-off. your capitalizations quirks. Connect the tools you use every day, like Gmail, Google Drive, Slack, and SharePoint, and ChatGPT Work...
High-engagement X post with 4,590 visible interactions on AI developer tooling.
Signals
All →No signals in the last 7 days.
Just in
Story
pytorch.org
Alibaba Cloud, Ant Group, Cambricon and Huawei Come Together in Shanghai to Advance the Open Source AI Stack at PyTorch Conference China
Story pytorch.orgCambricon Joins the PyTorch Foundation as a Platinum Member
Story aiforrealsite.comCognizant’s AI Hiring Push Shows How New Jobs Could Emerge As Old Ones Change
Story siliconangle.comOpenAI chief scientist argues for AI research slowdown
Topic radar
All →
Robotics and embodied AI
+6.7
Humanoid and industrial robotics initiatives are advancing through expanded training datasets for manipulation, predictive world models for autonomous physical movement, and supply chain scaling despite workplace infrastructure limitations.
AI policy and governance -2.5Organizations and public bodies are confronting governance challenges, spanning copyright infringement lawsuits from publishers, agency frameworks for law enforcement deployments, and internal executive discussions regarding research pace and safety reporting.
AI infrastructure and compute -3.3Data center growth faces physical and legal constraints, including rising per-rack power bottlenecks, municipal lawsuits over noise and turbine emissions, and multi-billion-dollar compute commitments across major AI providers.
Frontier model releases -5.2Recent evaluations and tooling integrations highlight new frontier model releases, tracking multimodal capabilities, autonomous task execution in games, and adjusted token pricing structures across developer deployments.
Quick takes
Frontier model developers are increasingly orienting new flagship releases toward agentic execution, computer use, and domain-specific cybersecurity workloads.Multi-billion-dollar capital raises and large private compute contracts demonstrate the intensifying financial scale required to secure and expand foundational AI data center infrastructure.Frontier AI labs are developing custom inference silicon to achieve higher throughput, reduced latency, and improved power efficiency for serving modern models.Concurrent service disruptions across multiple commercial foundation model platforms demonstrate operational fragility and systemic infrastructure dependencies in deployed AI systems.Organizations across legal, public administration, and consumer technology sectors are deploying coding and language models to automate specialized knowledge workflows and accelerate software development.Developers are integrating agentic video understanding directly into media production workflows, pairing task-oriented visual analysis with lower token usage and reduced operational compute costs.Frontier model releases are increasingly prioritizing agentic workflows, autonomous computer use, and critical cybersecurity capabilities across major developer platforms.Capital expansion for AI compute infrastructure continues to accelerate through multi-billion-dollar private funding rounds and large-scale capacity agreements backing foundation model developers.
Just in
Story
pytorch.org
Alibaba Cloud, Ant Group, Cambricon and Huawei Come Together in Shanghai to Advance the Open Source AI Stack at PyTorch Conference China
Story pytorch.orgCambricon Joins the PyTorch Foundation as a Platinum Member
Story aiforrealsite.comCognizant’s AI Hiring Push Shows How New Jobs Could Emerge As Old Ones Change
Story siliconangle.comOpenAI chief scientist argues for AI research slowdown
Market pulse
bullish
AI pulse
56/100
AI-linked equities are broadly positive, with Advanced Micro Devices +4.69%, Super Micro Computer +4.54%, Palantir Technologies. -4.49% leading the tracked basket.
| Ticker | Company | Move |
|---|---|---|
| AMD | Advanced Micro Devices | +4.69% |
| SMCI | Super Micro Computer | +4.54% |
| PLTR | Palantir Technologies. | -4.49% |
| ARM | Arm Holdings plc | +3.92% |