The latest AI news, in focus.

The sleep framework: A new way for LLMs to retain knowledge in continual learning

Education Innovation

Why dbt Labs thinks AI agents will finally make the semantic layer stick

Why AI agents can make semantic layers stick, if teams can solve the trust problem.

b11935

rpc : catch std::length_error in recv_msg (#31204) **Website:** - **Attestations:** - **macOS/iOS:** - [macOS Apple Silicon (arm64)](https://github.com/ggml-org/llama.cpp/releases/download/b11935/llama-b11935-bin-macos-arm64.tar.gz) - macOS Apple Silicon (arm64, KleidiAI enabled) [DISABLED](https://github.com/ggml-org/llama.cpp/pull/23780) - [macOS Intel (x64)](https://github.com/ggml-org/llama.cpp/releases/download/b11935/llama-b11935-bin-macos-x64.tar.gz) - [iOS XCFramework](https://github.com/ggml-org/llama.cpp/releases/download/b11935/llama-b11935-xcframework.zip) **Linux:** - [Ubuntu x64 (CPU)](https://github.com/ggml-org/llama.cpp/releases/download/b11935/llama-b11935-bin-ubuntu-x64.tar.gz) - [Ubuntu arm64 (CPU)](https://github.com/ggml-org/llama.cpp/releases/download/b11935/llama-b1

Frontier AI labs say their models refuted the Exponential Time Hypothesis. What does that mean?

TechCrunch spoke to the MIT computer scientist charged with explaining OpenAI's latest mind-bending discovery

How We Built Our Design Agent

How LangChain built Content Studio, an open-source design agent in Slack that makes on-brand videos, graphics and decks, and the harness that powers it.

Announcing SOCI-indexed images for faster SageMaker AI jobs

Seekable OCI (SOCI) indexing for Amazon SageMaker AI training and processing jobs takes full container image download off the critical path. SageMaker AI lazy-loads your image to start containers in seconds instead of minutes, with no changes to your job definition. See the benchmarks and how to adopt it.

NVIDIA RTX Spark Powers Neural Rendering and AI Characters in Games

NVIDIA RTX Spark is a new Windows PC platform for AI development, creation, and gaming in slim laptops and compact desktops. Powered by an NVIDIA Blackwell RTX...

Fix evaluation bias to unlock better models on Amazon Bedrock

Property Finder found that 96.8% of their vision-AI evaluation labels were inherited from the incumbent model, hiding its true accuracy. See how they built unbiased ground truth, applied evolutionary prompt optimization, and selected a more accurate, cost-effective model on Amazon Bedrock, cutting cost by 86% across 11 image-quality attributes.

Exowatt beats frontier models on accuracy with SageMaker serverless model customization

In this post, we show how Exowatt used Amazon SageMaker AI serverless model customization to build a repeatable AI pipeline that produces task-specific fine-tuned models for two deployment targets, edge and cloud, achieving higher accuracy than frontier models at up to 49x lower cost.

Cloudera bets portability, not agents, will set it apart in enterprise AI

Enterprises racing toward AI automation are finding that where their models and data run matters as much as what the models can do. Regulatory pressure, rising token costs and fear of lock-in are pushing companies to demand more control over their AI workloads. That demand is shaping how data platform providers pitch their next act. […] The post Cloudera bets portability, not agents, will set it apart in enterprise AI appeared first on SiliconANGLE.

Google’s Genie 4 world model adds audio, more realistic worlds, and real-time steering

Google DeepMind released Genie 4, a world model with real-time audio, more realistic simulations, and mid-simulation text prompts, in research preview for trusted testers.

How KOHO built a Claude-powered self-serve analytics tool on years of dbt governance

KOHO explains why AI self-serve analytics needs a strong data culture and a governed dbt foundation first.

Firefox Smart Window adds an AI form-filling feature that can automatically extract information based on the current tab.

On October 27, 2026, Mozilla updated Firefox’s AI browsing feature, Smart Window, introducing a new capability to intelligently fill out forms based on the user’s currently open tabs. This feature supports extracting information from websites such as LinkedIn and personal portfolios to automatically populate multi-field forms like job application forms.

Why it mattersThis update represents a significant advancement in Firefox’s recent AI feature iterations. As reported by The Verge and clearly dated October 27, 2026, the announcement focuses on specific interaction logic and use cases, with complete information elements and clear sources. However, it does not disclose technical implementation details, data processing methods, or privacy policies, nor does it specify whether this feature has been rolled out to all users or remains limited to certain testing channels.

Fine-Tuning NVIDIA Nemotron to Know Local Rules, Not Just Local Languages

As AI adoption expands globally, people across languages and cultures are using it in more aspects of their daily lives. General-purpose language models serve...

NVIDIA Nemotron 3.5 Transcribe and Granary v2: Intelligent Multilingual Transcription You Can Prompt and Adapt

NVIDIA Nemotron 3.5 Transcribe and Granary v2: Intelligent Multilingual Transcription You Can Prompt and Adapt

b11934

tests : speed up tensor data generation and model saving (#31145) * tests : speed up tensor data generation and model saving Refactor save_models to create only the model, not a llama_context, which is not needed for saving. Replace the slow std::normal_distribution draw with a splitmix64 + box-muller normal generator. Assisted-by: dsh:llama.cpp/ggml-org/DeepSeek-V4-Flash-Vision-Exp-GGUF (Mac Studio, M2 Ultra, 192GB) * tests : use a shared pool of normal draws for tensor data Generate a pool of normal draws once with std::mt19937 and std::normal_distribution, then have every tensor copy a cyclic slice of it instead of running a per-tensor box-muller. The name-keyed slice selection keeps the draw deterministic. Assisted-by: dsh:llama.cpp/ggml-org/DeepSeek-V4-Flash-Vision-Exp-GGUF (Mac Studi

b11933

vulkan: fix CI llvmpipe by allowing dim 3 broadcast in MUL_MAT (#31239) * vulkan: allow dim 3 broadcast in MUL_MAT supports_op rejected every MUL_MAT whose src0 and src1 differ in ne3, although mul_mm, mul_mmq and mul_mat_vec already broadcast src0 over dim 3 through broadcast3. Such a matmul then ran on the CPU with a copy of the weight, and the split layout changed with the number of streams in the ubatch. Keep the strict check for MUL_MAT_ID only, and route the broadcast cases away from mul_mat_vec_nc, the one path that requires equal ne3. * tests : harden MUL_MAT with permuted dim 3 broadcast cases Harden the MUL_MAT coverage with a weight broadcast over dim 3 against a permuted per-stream activation, the shape of an MLA output projection with a non unified KV cache, for a single token

UK Home Office May Face Investigation Over Use of OpenAI Tools

A government watchdog is considering a probe into the Home Office’s use of ChatGPT technology to process asylum claims amid concerns over the tools’ accuracy and secrecy surrounding their use.

How AI Is Changing the Pace of Discovery for Millions Living With Dementia

There are 57 million people worldwide living with dementia today. For the researchers working to help them, the obstacle isn’t a lack of effort — it’s too much data, too few experts to synthesize it and a pace of discovery that can’t keep up with the disease. The Consortium for Biomedical Research and AI in […]

3 Ways to Boost MoE Inference on NVIDIA DGX Spark

Generating tokens faster starts with identifying what limits inference performance. In the batch-size-1 evaluation of Ling-3.0-flash from Inclusion AI on NVIDIA...