OpenAI brings ChatGPT and the Codex coding agent natively to Linux, while researchers show every major model's 'encrypted' reasoning can be decoded by a cheaper model with no encryption broken.

• Researchers showed encrypted reasoning blobs from Claude Opus, GPT, and Gemini are portable across models — feed one to a cheaper model like Haiku and it reads the hidden reasoning aloud, no decryption needed.
• A scan of ~7,000 public Claude Code/Codex session traces found 62 exposed API keys, 33 emails, and 33 passwords hidden inside these 'encrypted' reasoning chunks.
• OpenAI shipped a native ChatGPT desktop app for Linux (Ubuntu, Debian, Fedora) bundling ChatGPT, ChatGPT Work, and Codex, giving the coding agent direct access to local repos and files.
• Unsloth released an open-source desktop app for local model training/inference on Mac, Windows, and Linux — 2x faster training, 70% less GPU memory, connects to Claude Code/Codex as a local API replacement.
• OpenAI added auto-sync of projects, chats, and plugins into Codex.
• Google's open diffusion model hit 1,500 tokens/sec on a single GPU; a new inverse model can reconstruct exact LLM prompts from public outputs alone.

DeepSeek V4 Pro 0813 (on OpenRouter)

• <p><strong><a href="https://openrouter.ai/deepseek/deepseek-v4-pro-0813">DeepSeek V4 Pro 0813 (on OpenRouter)</a></strong></p> The latest DeepSeek Pro model is now available, via API only. I had to link to OpenRouter because DeepSeek don't have any obvious announcement page for their new model.</p>

LFM2.5-VL-3B for Better and Faster Vision Capabilities for the Edge

Anthropic's unreleased Claude used 60 parallel subagents and 650 failed attempts to make the biggest Riemann Hypothesis progress in 160 years — and the agent architecture behind it is public now.

• Unreleased Claude model pushed Riemann Hypothesis proof coverage from 41.6% to 67.2% using 60 parallel subagents in Claude Code, 650 failed attempts, and 2,400 shell commands; proof verified in Lean 4 with logs public.
• OpenAI's Daybreak program splits into Blue (guardrail-loosened GPT-5.6 Sol for defensive security work) and Red (GPT-5.6-Cyber, which found real zero-days in Chrome's V8 engine, 95% success on advanced security tasks vs 1.5% for base model).
• NVIDIA released Nemotron 3.5 Lightning: 30B MoE model activating only 3B params, 4x faster than similar models, runs on a single H100, open weights on Hugging Face, fine-tunable via NeMo.
• Higgsfield open-sourced a full $2M AI-generated feature film with real celebrity actors — direct proof-of-concept for AI creative production at scale.
• Metis introduced persistent memory built into a model's forward pass, eliminating the need for retrieval-based memory systems — relevant for building longer-running creative agents.

Thinking of ACE? We Can Do It with Fewer Tokens

Pathway's tiny 150M-parameter model hints at a post-transformer future, while Meta, Grok, and Runway push new tools for image/video creation.

• Pathway released BDH-CQ, a 150M-parameter post-transformer model scoring 29.5% on ARC-AGI-1 at $0.0007/task — 11x cheaper than OpenAI's GPT-5.6 Luna (Low) despite slightly lower accuracy (34.5%); team claims it can scale to 600B params
• Meta launched Muse Glimmer, an open-weight 30B agentic model runnable on a single consumer GPU (Apache 2.0), beating Gemma4-31B and Qwen3.6-27B on coding and multi-step agentic benchmarks
• xAI shipped Grok Image 2.0 with improved editing, text rendering, and factuality
• Runway integrated P-Image-Ideogram, enabling prompt-to-image generation in under 1 second
• Anthropic made Claude Sonnet 5 introductory pricing permanent ($2/$10 per million input/output tokens)
• Google's Pixel camera team confirmed Gemini-class generative models are now driving rapid gains in computational photography (Pro Res Zoom, Reimagine, Magic Editor), ahead of the Pixel 11 launch

Quoting Florian Herrengt

• <blockquote cite="https://blog.florianherrengt.com/ai-removing-middle-class-software-engineering.html"><p>But then users start to report a weird bug. It's the 4th time your team has been trying to fix it. I mean... asking AI to fix it. Unfortunately, it seems like not even Fable can figure it out.</

Twitch content has trained Amazon AI for years, but users can opt out now

Stealing Reasoning Traces from Proprietary LLM APIs

• <p><strong><a href="https://stolen-thoughts.com/">Stealing Reasoning Traces from Proprietary LLM APIs</a></strong></p> A vanity domain name (<code>stolen-thoughts.com</code>) for <a href="https://www.alphaxiv.org/abs/2608.09867">a neat paper</a>:</p> <blockquote> <p>Anthropic, OpenAI, and Google ret

Stealing Reasoning Traces from Proprietary LLM APIs

• <p><strong><a href="https://stolen-thoughts.com/">Stealing Reasoning Traces from Proprietary LLM APIs</a></strong></p> A vanity domain name (<code>stolen-thoughts.com</code>) for <a href="https://www.alphaxiv.org/abs/2608.09867">a neat paper</a>:</p> <blockquote> <p>Anthropic, OpenAI, and Google ret

Gemini becomes Google's fastest-growing product ever as it hits 1B users