Model: Ripple-65B-Preview
RicursiveAI released Ripple-65B-Preview, a dense 65B-parameter model under Apache 2.0 built for agentic coding, in which only 16 of 80 layers use global attention while the other 64 use a 1,024-token sliding window. The …
RicursiveAI released Ripple-65B-Preview, a dense 65B-parameter model under Apache 2.0 built for agentic coding, in which only 16 of 80 layers use global attention while the other 64 use a 1,024-token sliding window. The …
PersonaliAI released Chatty, an MIT-licensed, self-hostable open-source AI customer support engine positioned as a Zendesk and Intercom alternative, with WooCommerce sync, WhatsApp and Slack channels, and an included 81-…
Norway's transmission system operator Statnett suspended new grid-capacity reservations for data centers on November 9, 2026, and paused processing of new facility-functionality approval applications under Section 14 of …
ZenWave Domain Language creator Ivan Garcia Sainz-Aja argues that reading the volume of Markdown generated by LLMs is the main bottleneck in spec-driven development, since agents treat unread decisions as approved. Garci…
Statnett, Norway's transmission system operator, paused new grid capacity reservations for data centers effective November 9, 2026, with the freeze lasting until at least August 1, 2027, targeting power-intensive AI faci…
Proton said it has added AI review layers to its existing human security controls without loosening its documented security principles, which have been unchanged since June 2024, and that end-to-end and zero-access encry…
JetBrains' tenth annual State of Developer Ecosystem report, based on 28,945 developers across 198 countries surveyed from May to July 2026, found that more than half of employed developers use AI agents in over 60% of t…
Microsoft will enable Anthropic's Claude models by default for UK Microsoft 365 Copilot tenants whose administrators have not chosen a setting, starting December 9. Anthropic processes the data in the US outside Microsof…
A developer demonstrated how to turn Google Search into a structured JSON API using Relinos, a programmable browser, with roughly 35 lines of JSON configuration. The setup uses linear browser actions to type a query and …
An unnamed developer released Tokspur, an open source inference engine for Apple Macs built on a custom Metal runtime with native multi-token prediction, claiming it runs 3x faster than existing options. The engine was d…
Developer MadB0i has open-sourced C.U.R.E. v0.2.0, a portable Windows security and workspace intelligence toolkit written in Rust, with release binaries published on GitHub. C.U.R.E. targets stealth malware, obfuscated s…
A solo developer released Grai, a local HTML5 tool priced from $79/yr that runs a deterministic "cadence gate" over AI-generated drafts before publication, scoring them 4 to 100 against 8 structural rules and a coefficie…
Anthropic released Claude Fable 5.2 and Claude Mythos 5.2, two versions that share an underlying model but differ in safeguards, with Fable applying stricter controls to cybersecurity and biology tasks. Fable's safety cl…
An unreleased OpenAI cyber model produced 948 distinct Linux kernel local privilege escalation bugs, each with a path to root, in a single mid-October run that cost $1,213,430 in API usage, or about $1,280 per vulnerabil…
Bilt announced Tuesday that it is launching "Bilt Neighborhood OS," an AI-powered platform that connects to merchants' existing reservation, payment, customer management and loyalty systems to help them reach nearby cust…
A developer outlined a method for catching silent failures in AI agents—cases where a request returns HTTP 200 with no error span yet the answer is wrong—by tracing every agent step, separating technical errors from cont…
Microsoft will enable Anthropic's Claude models by default for UK Copilot tenants on December 9, giving administrators until that date to review access settings because Anthropic processes the relevant data in the United…
Meta paused the updated Commercial Terms, Product Catalog Terms, Self-Serve Ad Terms and Ad Creative Generative AI Terms that were scheduled to take effect on Nov. 20, 2026, according to a Meta Business Help Center notic…
A developer benchmarked NVIDIA's Nemotron 3.5 Transcribe, released 28 October under the OpenMDW-1.1 license, by running its 21.6 GB NVFP4 model on a single 32 GB RTX PRO 4500 Blackwell card via vLLM and NVIDIA's NeMo Spe…
A developer behind FinMan shipped an Android app and expanded the expense tracker to import any bank's statement in CSV, Excel or PDF, using AI only once per new bank format to generate a parsing template so later files …