Meta Superintelligence Labs released Muse Glimmer, a 30-billion-parameter dense model under Apache 2.0 in August 2026, designed for local, single-GPU inference with a 262,144-token context at 24GB VRA…
Meta released its 30B-parameter Muse Glimmer model under the Apache 2.0 license, a shift from the Llama Community License that removes restrictions on monthly active users, competition with Meta produ…
Meta launched Meta Compute on 1 July 2026 to sell surplus AI training and inference capacity, backed by its custom MTIA chips and a planned 2026 capital expenditure of $125 billion to $145 billion, in…
Meta Platforms Inc. gives away its most powerful AI models, including Llama and Muse-class releases, under the Apache 2.0 license, while CEO Mark Zuckerberg's 6,500-word manifesto 'The Future is for E…
Meta's Llama 4 open-weight models, touted as beating GPT-4o and Gemini 2.0 on coding benchmarks, failed independent reproduction tests, with Rootly scoring Llama 4 last at 69.5%, raising trust and tot…
The browser market is fragmenting in 2026 as Chrome's global usage slips to roughly 67–70%, down about 1.9 points year over year, while Google still spends $26 billion annually on default-placement de…
Chrome's browser market share has fallen to 68.02% from a peak of 73.17%, according to Backlinko's analysis of StatCounter data, as generative AI absorbs search queries and the browser wars shift from…
Google's Manifest V3 update to Chromium has disabled the full uBlock Origin extension in Chrome and Edge, replacing the blocking webRequest API with the rule-limited declarativeNetRequest API, a chang…
Employees spend over 80% of the working day in the browser, yet enterprise security controls remain on the network and OS, leaving the authenticated browser session unobserved and creating gaps in vis…
A CTO's evaluation framework for choosing an enterprise browser in 2026 recommends a weighted rubric, POC design, and build-vs-buy maths, noting that Gartner projects a quarter of organizations will a…
Apple's Siri AI, rebuilt on a new architecture for iOS 27 and macOS 27, is a ground-up replacement for the assistant, featuring on-device models with screen and cross-app awareness. The company paid $…
Apple's OS 27 now routes larger Apple Intelligence workloads off-device onto Google Cloud and NVIDIA GPUs, extending its Private Cloud Compute (PCC) privacy architecture beyond Apple's own data centre…
Apple's Apple Intelligence is a hybrid system that runs privacy-sensitive tasks on-device but routes heavier workloads to Private Cloud Compute (PCC), with offloads including agentic tool use, complex…
Apple's macOS 27 'Golden Gate' introduces Siri AI alongside third-party models Claude (Anthropic) and Gemini (Google) as governed enterprise AI integrations, with Apple Business settings separating ex…
Apple is retiring legacy MDM mechanisms across all OS 27 releases with no grace period, forcing a cutover to Declarative Device Management (DDM), a framework where servers declare desired state and de…
Realtime voice agents, which combine streaming speech-to-text, a large language model, and text-to-speech in a cascaded pipeline, must achieve a time-to-first-audio (TTFA) of around 200 milliseconds t…
Function calling, the Model Context Protocol (MCP), and the Agent-to-Agent (A2A) protocol together decouple AI agents from vendor-specific interfaces, enabling enterprise voice agents to act on backen…
A vendor-neutral scoring method for evaluating whether to build or buy AI voice agents is outlined, with buying favored under about 50,000 minutes per month and building past about 150,000, where a cu…
A technical survey reports that a fully self-hosted voice AI pipeline can achieve around 755 milliseconds from end of speech to first audio output, while a provider comparison shows the same speech-to…
Gallup polling shows 71% of Americans opposed an AI data center near them by March 2026, up from 47% in late 2025, making data centers less popular than nuclear power plants, which 53% oppose. At leas…