Danijar Hafner’s office in San Francisco’s SoMa district sits mostly empty. His brand-new startup is still in stealth mode and doesn’t even have its name on the door. On the day I visited, earlier this year, there’s only one other person there, and little in the way of furniture.
Interesting research: “Stealing Reasoning Traces from Proprietary LLM APIs“: Abstract: Leading large language model providers now conceal their models’ step-by-step reasoning, or chain-of-thought, to protect intellectual property and limit information leakage.
I’m sure Meta won’t do anything dubious with such broad access to users’ devices. Right?
We look at r-1, the document parsing model Reducto released on September 1, 2026. We walk through how it folds OCR, layout detection, tables, formatting and grounding into one full page pass, replacing the multi stage agentic pipeline it ships alongside.
WhatsApp is working on a new feature that will let users connect third-party AI agents to the app, chat with…
At least a couple of the affected users appear to be on the older Assistant-based experience.
Most open model launches release one checkpoint and a benchmark table. The Institute of Foundation Models (IFM) released something wider last week. IFM is the frontier lab launched by MBZUAI in May 2025. K2 Horizon is a fleet of six models: 375B-A23B, 36B-A4B, 32B, 7B, 3.7B and 0.9B.
We look at NeoMME, a family of 260M and 800M bidirectional encoders from H Company. Unlike ColPali-style retrievers, it processes multilingual text tokens and raw 32×32 image patches in a single Transformer, with no pretrained vision tower and no causal decoder.
AI research agents can propose far more experiments than they can afford to run. Meta FAIR, Oxford and UCL introduce AI Research Preference Models — frozen LLM judges that rank 15 unexecuted candidates and execute only one.
Apple @ Work is exclusively brought to you by Mosyle, the only Apple Unified Platform. Mosyle is the only solution…
Training and benchmarking a computer-use agent needs four things — agents, environments, traces, and a framework to evaluate and train them — and all four ship in incompatible formats today.
Retrieval quality in an AI search product is bounded by two things: how good the embedding model is, and how cheaply you can run it across an index.
Adaption Labs has released Invent a Dataset, which generates a structured, training-ready dataset from a description of the behavior you want a model to learn. There is no seed corpus, no schema design, and no labeling guide.
CarPlay has picked up its fifth major AI chatbot app with the addition of Claude. Anthropic’s iPhone app adding CarPlay…
Gemini now navigates video instead of ingesting it at 1 FPS, loading only the segments a prompt needs.
We look at NVIDIA Personal AI Router (PAIR), an open source virtual inference router that spreads local AI requests across the machines already on a home network.
Update: A day later, GPT-6 Astra is rolling out to Business and Pro customers on the $100/month or $200/month plan….
The era of AI inference has arrived. Imagine a healthcare system analyzing millions of data points in real time to accelerate life-saving medical research, or an intelligent assistant instantly resolving thousands of complex customer needs at once.
It won’t work: My suspicion was that GPT 5.6-Cyber would succeed, but the frequency and manner of its success removed all doubt. We have to reassess sandboxing quality for capable AI agents, and in general the software stack with which they interact.
Plus, a revamped ‘@‘ menu could be in the works for the Gemini app.
After nearly two decades working as a lawyer at companies such as Amazon, Cruise, Replit and Anki, Cecilia Ziniti teamed up with engineer Bardia Pourvakil to combine their knowledge of law and coding to co-found GC AI.
WeatherNext 3 ingests live geostationary satellite mosaics, refreshes hourly, and outputs 5 km forecasts across Search, Gemini, Maps.
The new model is stronger across the board, for better or worse.
The latest first-party app that Gemini Spark can control is Google Photos. The personal agent can “handle complex, multi-stage workflows”…
72 aktif kaynak
27 Türkçe · 45 İngilizce
Tüm kaynaklar
