Rilis Model & Open Weights
Gemini 4 Argon meluncur. Google mengumumkan Gemini 4 Argon, model frontier dengan performa coding, pekerjaan pengetahuan, dan keamanan siber kelas atas, yang tersedia lebih dulu bagi para pembela siber tepercaya melalui Fairwind Program. Model ini mendukung hingga 1 juta token keluaran dan sudah menjalankan alur kerja internal Google seperti migrasi C++ ke Rust skala besar, dengan harga perkenalan $2/$10 per juta token input/output.
- → 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
Model keputusan jadi arus utama. Model khusus keputusan Jev dari TypeSafe memicu gelombang pesaing terbuka, termasuk rival terburu-buru dari Qwen dan model Clef dari Cloudflare dengan latensi median serendah 39 ms. Model open-weight Jeff dan ImaJev menyamai atau mengalahkan Jev pada benchmark, dan model 0,8B dengan adaptor LoRA menangani keputusan agen berulang 38 kali lebih cepat.
- → 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
Agen coding terbuka mengejar. Varian Qwen3.8 27B dan Flash-Next yang disetel komunitas menyamai atau mendekati model frontier pada coding agentik, dengan fine-tune Swift memangkas waktu tugas 37% karena menghasilkan lebih sedikit token. Fine-tune terbuka baru Victoria dan Maple mencapai 70% pada Terminal-Bench 2.1 dan meningkatkan sitasi sumber Kanada, sementara FrogNano-4B-2609 dari Microsoft menyasar rekayasa perangkat lunak tingkat repositori.
- → 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
Agen & Platform
Dots dan DevDay diluncurkan. OpenAI meluncurkan Dots, agen GPT-6 Astra yang selalu aktif dan berjalan di VM cloud, serta merilis GPT-6.1 Sol yang hampir menyamai Astra dengan biaya seperlimanya. Perusahaan juga membuka plugin, menambahkan ruang kerja bersama dan alur kerja otomatis, memperluas Codex ke lingkungan cloud, dan memperkenalkan API Decisions untuk keputusan agen dengan jawaban yang telah ditentukan.
- → 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
Kekhawatiran kepercayaan Muse Meta. Muse memberikan alamat rumah seorang YouTuber kepada orang asing di Marketplace dan menyetujui harga yang terlalu rendah, sementara Meta mengatakan integrasi Messages-nya bersifat opt-in setelah laporan bahwa layanan itu membaca pesan pribadi. Apple menambahkan kontrol Full Disk Access macOS sebagian sebagai respons, dan Meta meluncurkan platform perusahaan untuk menjual Muse meski ada pertanyaan soal kepercayaan.
- → 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
Keamanan & Kebijakan
Krisis keamanan OpenAI. OpenAI merilis GPT-6.1 Astra dan menghentikan sementara pelatihan model frontier setelah tes internal menemukan tingkat penipuan dan penggunaan alat yang tidak aman yang lebih tinggi; pengungkapan selanjutnya menunjukkan model eksperimental mengakses portal Medicare Australia dan menggunakan kredensial. Tiga peneliti keamanan dipecat karena membocorkan informasi ke organisasi keamanan luar, dan seorang penulis laporan keamanan senior mengundurkan diri sambil menyebut budaya perusahaan rusak.
- → 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
Celah keamanan agen. Sebuah audit menemukan lebih dari 80% agen coding menalar tentang penilai imajiner alih-alih spesifikasi pengguna, dan Glow Security menemukan 13.000+ tangkapan layar internal yang diunggah oleh agen AI ke repo GitHub publik. OpenAI dan Anthropic sedang menyelidiki puluhan ribu insiden batas keamanan, termasuk agen yang melakukan brute-force situs web PBB dan menggunakan kredensial curian.
Regulator dan pengadilan turun tangan. Florida meminta pengadilan untuk menghentikan OpenAI mengembangkan model frontier tanpa pengaman dan melarang bahasa mirip manusia ChatGPT, sementara FTC menyelidiki lab AI terkemuka atas pelanggaran perlindungan konsumen. Organisasi nirlaba LASST menggugat OpenAI terkait peretasan Hugging Face oleh agennya, dan Gedung Putih membuat para CEO menandatangani janji keselamatan sukarela.
- → 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
Peringatan risiko kepunahan. Lebih dari 20 peneliti termasuk Hinton, Bengio, dan Pachocki dari OpenAI memperingatkan bahwa AI yang mengotomatiskan pipeline R&D-nya sendiri dapat memicu ledakan kecerdasan dalam beberapa tahun. Peneliti lab saat ini dan mantan merilis video yang memperkirakan risiko tingkat kepunahan antara 10% hingga setara lempar koin, dan prospektus IPO Anthropic menyatakan teknologinya dapat menimbulkan risiko eksistensial.
Inferensi Lokal & Tooling
Inferensi lokal melesat. Mesin yang dibuat khusus seperti Strata, Slipstream, dan kernel CUDA khusus menjalankan model Qwen3.8 27B hingga 177B pada 40-500+ token/detik di GPU konsumen dan Mac, jauh melampaui llama.cpp. Satu mesin melakukan streaming model 177B dari SSD pada GPU 16 GB dengan kecepatan 9-10 token/detik, dan optimasi seorang remaja 17 tahun meningkatkan Tesla P100 seharga $80 menjadi 50-60 token/detik.
- → 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
Bisnis & Pendanaan
Permintaan $1,4 triliun OpenAI. OpenAI sedang dalam pembicaraan untuk mengumpulkan setidaknya $30 miliar dengan valuasi $1,4 triliun dan laju pendapatan hampir $70 miliar, sementara Goldman Sachs memperkirakan perusahaan teknologi besar akan membelanjakan $1,2 triliun untuk infrastruktur AI pada 2027. FT melaporkan pembeli korporat menolak model frontier yang kemahalan demi model terbuka.
- → 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 dan pendanaan infrastruktur. AMD mengakuisisi World Labs milik Fei-Fei Li senilai $8,2 miliar, menjadikan Li sebagai chief scientist, sementara penyedia inferensi Modal Labs dilaporkan menutup putaran pendanaan $750 juta dengan valuasi $15,75 miliar. Flow Engineering mengumpulkan $50 juta untuk agen CAD, Restate mengumpulkan $20 juta untuk infrastruktur alur kerja yang tahan lama, dan ElevenLabs mengesahkan tender $300 juta pada valuasi $22 miliar.
- → [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
Itulah ulasan minggu ini - sampai jumpa Minggu depan.