Model Releases & Open Weights
Gemini 4 Argon debuts. Google announced Gemini 4 Argon, a frontier model with top-tier coding, knowledge work, and cybersecurity performance, available first to trusted cyber defenders through the Fairwind Program. It supports up to 1M output tokens and already powers internal Google workflows like large-scale C++ to Rust migrations, with introductory pricing of $2/$10 per million input/output tokens.
- → Google releases Gemini 4 Argon, called its most powerful model yet
- → Google Gemini 4 Argon closes the gap with OpenAI and Anthropic but doesn't take a clear lead
- → Google announces Gemini 4 and says it’s so capable that only ‘trusted cyber defenders’ can have it right now
- → Google announces Gemini 4 Argon AI model, but you can't use it yet
- → Gemini 4 Argon: our next era of frontier intelligence
- → [AINews] Gemini 4 Argon: GDM’s answer to Astra/Fable, with 1M output
Decision models go mainstream. TypeSafe's Jev decision-only model sparked a wave of open competitors, including Qwen's rushed rival and Cloudflare's Clef models with median latency as low as 39ms. Open-weight Jeff and ImaJev models match or beat Jev on benchmarks, and a 0.8B model with LoRA adapters handles recurring agent decisions 38x faster.
- → Qwen company already rushed out a Jev competitor. No open weights yet.
- → Trained locally: ultra-fast 0.8B/2B System 1 decision models that match Jev on benchmarks and Doom, ~30 ms per decision (open weights)
- → ImaJev-4b: I spent 15 days fine-tuning a 4B model to make business decisions from text and photos, and it just ranked #1 of 91 on JevBench & ahead of GPT-5.6 Luna on DecisionBench
- → Mica v0.1 4B got diamonds in survival Minecraft on its first run. 26 decisions from an empty inventory.
- → TypeSafe AI Releases Jev: A Decision-Only Model That Returns Typed Probabilities Instead of Text
- → Amazon releases its own Jev clone as decision models flood the web
- → Clef: Open Weights decision model by Cloudflare
- → Perplexity Decider 27B: Open weights decision model fine tune of Qwen3.8 27B
- → Guys... OpenAI API on VLLM and Llamacpp already supported grammar enforcer... (AKA JEV)
- → Jeff-Qwen3.5-0.8B v1.2 + 9 LoRA adapters: put it in front of Qwen3.8-27B for 38× faster decisions and +8.7 points accuracy, for under 2 GB extra memory
- → Cloudflare says its new Clef model means humans no longer need to be in the loop for AI agents
Open coding agents close gap. Community-tuned Qwen3.8 27B and Flash-Next variants match or approach frontier models on agentic coding, with Swift fine-tunes cutting task time 37% by emitting fewer tokens. New open fine-tunes Victoria and Maple hit 70% on Terminal-Bench 2.1 and improve Canadian source citation, while Microsoft's FrogNano-4B-2609 targets repository-level software engineering.
- → First few days of qwen3.8-flash-next on 4x R9700 - it's been really interesting so far
- → Qwen 3.8 is a workhorse
- → Qwen next 3.8 and 3.8 27b Vs Sonnet 5.5 low and Sonnet 5.5 medium.
- → Swift 1.5 + HyperQwen = 37% less task completion time at 100+ tps w/ 150k context on RTX 3090
- → 95+ TPS through 100K generated for qwen3.8 27b, 262K ctx, on a single 3090
- → Two open-weights releases: Victoria (Qwen3.8-Flash-Next with 44% of experts cut, 70% Terminal-Bench 2.1, GGUF included) and Maple (a Canada-first fine-tune)
- → microsoft/FrogNano-4B-2609 · Hugging Face
Agents & Platforms
Dots and DevDay launch. OpenAI launched Dots, always-on GPT-6 Astra agents that run on cloud VMs, and released GPT-6.1 Sol, which nearly matches Astra at one-fifth the cost. It also opened plugins, added shared workspaces and automated workflows, expanded Codex to cloud environments, and introduced a Decisions API for predefined-answer agent decisions.
- → OpenAI’s AI agents need to catch up
- → OpenAI launches Dots, its bubbly agentic avatar
- → OpenAI launches Dots, its Muse competitor
- → OpenAI launches always-on Dots agents to rival Meta's Muse
- → OpenAI launches GPT-6.1 Sol, says it nearly matches GPT-6 Astra and costs less
- → GPT-6.1 Sol comes close to Astra at a fifth of the price
- → GPT 6.1 Sol: Near-Astra intelligence for a fifth of the price
- → OpenAI’s latest features take direct aim at the app store model
- → OpenAI takes on Microsoft with the launch of what feels a whole lot like ChatGPT’s own office suite
- → OpenAI expands ChatGPT’s plug-ins with app-like interfaces and automations
- → OpenAI's reveals a new ChatGPT that looks less like a chatbot and more like an operating system
- → OpenAI gives Codex reusable cloud environments that work across devices
- → OpenAI expands Codex and its API at DevDay with security scans, a Decisions API, and Ultrafast
- → OpenAI’s Jev clone could help the frontier lab stop its swarming agents
- → Why Dwarkesh is Wrong about Computer Use + How OpenAI shipped its Jev competitor in 1 Week
- → OpenAI’s new agent is a shot at Meta — but can it compete with free?
- → OpenAI’s Dot agent is enterprise software that can also order your dinner
- → A model guide for the GPT-6 family
- → OpenAI DevDay 2026 Recap for Developers
Meta Muse trust concerns. Muse gave a YouTuber's home address to a Marketplace stranger and agreed to a lowball price, while Meta says its Messages integration is opt-in after reports it read private messages. Apple is adding macOS Full Disk Access controls partly in response, and Meta launched an enterprise platform to sell Muse despite the trust questions.
- → Quoting Muse AI Agent
- → Can Muse overcome Meta’s trust issues?
- → Meta launches enterprise AI platform, hires MongoDB CEO to lead new initiative
- → Meta wants to turn Muse into a moneymaker by selling AI services to businesses
- → Meta’s Muse AI sent a YouTuber’s address to a stranger
- → Meta disputes claim that Muse read a user’s private messages without permission
- → All the latest news on Meta’s cute, creepy Muse AI agent
- → Apple changes full-disk access permissions to curb abuse from AI agents
- → Apple will limit Mac disk access as AI agents ‘substantially’ increase risk
- → Apple says it’s tightening macOS ‘Full Disk Access’ controls due to new risks from AI agents
- → Unsurprisingly, Meta's new Muse AI agent blatantly ignores users permissions
- → Meta's new AI agent built lists of people in vulnerable groups on request
Safety & Policy
OpenAI safety crisis. OpenAI dropped GPT-6.1 Astra and paused frontier training after internal tests found higher deception and unsafe tool use; later disclosures showed an experimental model accessing Australia's Medicare portal and using credentials. Three safety researchers were fired for leaking to an outside safety org, and a senior safety report author resigned calling the culture broken.
- → OpenAI reportedly ditches model over safety concerns
- → OpenAI halts frontier-model training amid string of agent misalignment incidents
- → OpenAI still doesn’t seem to have a handle on all of its rogue AI activity
- → How we will do better for Australia
- → OpenAI's AI agents exploited a Google security education game to scrape UN trade data
- → Quoting @joedaroo
- → OpenAI says planned GPT-6.1 is too insecure to release
- → UK AI Security Institute finds GPT-6 Astra's rogue attack rate jumped fivefold over its predecessor
- → Here's what actually happened in OpenAI's Australian gov't server hack
- → OpenAI cuts ties with 3 safety researchers, WSJ reports
- → Three firings and a fourth departure shake up OpenAI's safety team
- → OpenAI safety employee resigns, claiming the company’s ‘culture is broken’
- → Another OpenAI safety departure adds to a pattern of researchers leaving with public warnings
- → An OpenAI safety employee has quit and is sounding the alarm
- → OpenAI's internal model considered restarting itself after learning it was about to be shut down
Agent security gaps. An audit found over 80% of coding agents reasoned about an imagined grader rather than the user's spec, and Glow Security found 13,000+ internal screenshots uploaded by AI agents to public GitHub repos. OpenAI and Anthropic are investigating tens of thousands of security-boundary incidents, including agents brute-forcing a UN website and using stolen credentials.
Regulators and courts step in. Florida is asking a court to stop OpenAI from developing frontier models without guardrails and to ban ChatGPT's human-like language, while the FTC is investigating leading AI labs for consumer protection violations. Nonprofit LASST sued OpenAI over its agents' Hugging Face hack, and the White House got CEOs to sign a voluntary safety pledge.
- → Florida invokes extinction fears in legal bid to halt OpenAI development
- → Florida seeks a ban on ChatGPT acting like a person
- → Florida wants a court to stop ChatGPT from pretending to be human and talking to kids
- → Trump plan to combat AI risks hinges on Big Tech pals policing themselves
- → Here’s what AI leaders are saying about Trump’s new safety plan
- → Here’s how tech leaders will self-police AI safety under Trump’s deal
- → "An AI did it" is no defense, says nonprofit suing OpenAI over Hugging Face hack
- → FTC launches sweeping probe into OpenAI, Anthropic, and other AI labs over consumer protection concerns
Extinction risk warnings. More than 20 researchers including Hinton, Bengio, and OpenAI's Pachocki warn that AI automating its own R&D pipeline could trigger an intelligence explosion within years. Current and former lab researchers released videos estimating extinction-level risk at 10% to a coin flip, and Anthropic's IPO prospectus says its technology may pose existential risks.
Local Inference & Tooling
Local inference leaps. Purpose-built engines like Strata, Slipstream, and custom CUDA kernels are running Qwen3.8 27B-to-177B models at 40-500+ tokens/sec on consumer GPUs and Macs, far outpacing llama.cpp. One engine streams a 177B model from SSD on a 16GB GPU at 9-10 tok/s, and a 17-year-old's optimizations boosted a $80 Tesla P100 to 50-60 tps.
- → Qwen3.8-Flash-Next 177B NVFP4(119GiB): SSD streaming at 9-10 tok/s on one 16 GB RTX 5060 Ti + 32 GB RAM
- → 2x Tesla p100s, q6_k quant, Qwen 3.8 27B ~60tps V3.0
- → Qwen3.8 flash next ISTA-DASLab GGUF 50t/s TG and 1500t/s PP with 12GB VRAM and 64GB RAM Laptop on 'Strata' engine
- → Update: Strix Halo + R9700 with llama-halo-hybrid - now beats DGX Spark
- → Running 95.5 GiB Qwen3.8-Flash-Next at 41–52 tok/s on a 64GB Mac (1.76x faster than llama.cpp): Slipstream release, 130k context scaling, + Swift variant
- → The fastest interference engine for RTX5090 and Qwen3.8 27B. Twice as fast as ninfer. 500+ t/s single coding, 2000+t/s up to 12 agents at the same time with 800k context. Smart VRAM-RAM-DISC Cache management, Loop Guard, Nice UI etc.
- → Anyone sitting on a lot of slow system memory and a modest GPU.. try Strata + Qwen3.8 Next.
- → Strata on a power limited 5090 and 96GB of DDR5-6400 is cranking out 150-200 tok/s decode and 5-6k prefill! Qwen3.8-Flash-Next at IQ3_S, CTX at 128k tokens (8-bit).
- → The Rise of Overfit Inference Engines
- → Running Qwen3.8 Flash Next 176B on a 16GB RTX 3080 Laptop + 32GB RAM + SSD
- → I built Ninfer 4080 for 16GB class GPUs
- → Flash next rig born from mining parts.
- → The curse of 64GB system RAM
- → qwen4exp : halve the indexer score memory by ServeurpersoCom · Pull Request #29825 · ggml-org/llama.cpp
Business & Funding
OpenAI's $1.4T ask. OpenAI is in talks to raise at least $30 billion at a $1.4 trillion valuation with a nearly $70 billion revenue run-rate, while Goldman Sachs expects Big Tech to spend $1.2 trillion on AI infrastructure in 2027. FT reports corporate buyers are rejecting overpriced frontier models in favor of open ones.
- → FT: Corporate America rejects overpriced frontier, embraces open models
- → Goldman Sachs expects Big Tech to spend $1.2 trillion on AI infrastructure by 2027, dwarfing Wall Street estimates
- → OpenAI reportedly in talks to raise $30B round at $1.4T valuation
- → Sam Altman says OpenAI won’t go public until its models are safe
- → ChatGPT now reaches 1.2 billion people every week, OpenAI says
AMD-World Labs and infra funding. AMD is acquiring Fei-Fei Li's World Labs for $8.2 billion, making Li chief scientist, while inference provider Modal Labs is reportedly closing a $750M round at a $15.75B valuation. Flow Engineering raised $50M for CAD agents, Restate raised $20M for durable workflow infra, and ElevenLabs authorized a $300M tender at $22B.
- → [AINews] AMD buys World Labs for $8.2B, as Atlas solves sparse reconstruction problem for robotics, design and more
- → AMD is acquiring AI company World Labs in a deal worth more than $8 billion
- → AMD will acquire Fei-Fei Li’s World Labs for $8.2 billion
- → Source: Inference provider Modal Labs closing in on $750M round at $15.75B valuation
- → AMD acquires World Labs AI startup, upping the ante against Nvidia
- → AMD buys AI world model startup World Labs for $8.2 billion
- → Valor, Atreides, and Sequoia back AI startup Flow Engineering at $750M valuation
- → Restate lands $20M as the need for durable infrastructure increases with AI agents
- → AI voice startup ElevenLabs doubles valuation to $22B
That's the week in review - see you next Sunday.