GoKawiil
Tech News
Meta has moved its agentic AI assistant, Muse, into a limited closed testing phase, according to Testing Catalog, gating access through a waitlist and invite codes. An App Store listing reveals the app can connect to Gmail, Google Calendar, health apps, and even users' bank accounts and cards to track spending and suggest canceling unused subscriptions.
Thoughtworks' Europe CTO Giles described an internal experiment in Barcelona where 10 engineers used heavily agentic AI coding practices to build a simulated airline irregular-operations (IROps) system in just four days. Working in a single shared code repository, the team discovered that requiring agents to constantly commit and rebase against the main branch produced unexpected coordination effects, effectively turning the codebase into a shared communication point between agents.
At IFA 2026, Nvidia introduced the Personal AI Router (PAIR), software that distributes sub-tasks from a local AI agent across idle GPUs on other PCs in the same home network. Instead of overloading one machine with multiple sub-agents competing for resources, PAIR spreads the workload and returns results to the main system, adapting in real time as household GPUs become busy or free again.
Zeke Hausfather, a climate scientist who relies on AI for research, spent eight weeks tracking 1,138 prompts he sent to Claude Code to estimate the electricity consumed at the data centers processing them. His findings suggest that AI agents, which perform multi-step autonomous tasks, draw significantly more power than the simple chatbot queries tech companies typically cite as low-energy examples.
An MIT Technology Review Insights webcast produced with NiCE examines how companies are moving AI agents from small pilots to enterprise-wide use. Chandra argues that agents need consistent access to data, context and back-end systems, and warns that scaling without integration can create new silos across teams.
The developer has released Muse Spark 1.3, an updated AI model available now through Muse Code and the Meta Model API. The update focuses on sustaining longer, multi-step tasks, following complex instructions more reliably, and collaborating more actively with users by asking clarifying questions and checking in before consequential actions. A more advanced reasoning mode is expected soon pending additional safety testing.
Google has shipped Gemini 3.8 Flash, its third Flash-tier model in six weeks, built to handle complex agentic workflows more effectively than earlier versions. Alongside it, the company introduced Gemini 3.8 Flash Cyber, a specialized version aimed at vulnerability detection and automated patching, though that variant is currently restricted to participants in Google's Fairwind Program.
Google launched Gemini 3.8 Flash, its third Flash-tier model in six weeks, improving on coding, agentic tasks and multi-step reasoning while keeping the same pricing as 3.7 Flash ($0.75/$3.75 per million input/output tokens). Alongside it, Google introduced Gemini 3.8 Flash Cyber, a specialized version tuned for vulnerability detection and automated patching, accessible only to vetted security teams through a new invite-only Fairwind Program.
Google has upgraded Gemini's video analysis with a new 'agentic' approach that lets the AI selectively choose which frames, audio, or transcript segments to review instead of processing every frame. The company says this cuts token usage by up to 88% while improving accuracy by up to 7%, and enables tracking of object counts, movement, and sub-second events across multi-hour videos. The feature currently works through the Gemini API in AI Studio and the Enterprise Agent Platform, with plans to reach the Gemini app and YouTube's 'Ask YouTube' tool soon.
Shopify COO Jess Hertz says the company lets its engineering teams spend without limits on AI tools, a bet that's paying off as revenue climbs 34% even as headcount holds steady. Hertz frames this as part of Shopify's push to build the infrastructure that lets AI agents discover and purchase products on behalf of consumers, a shift away from the old model where big-budget brands dominated search rankings.
Anthropic has released Claude Fable 5.1 and Mythos 5.1, new AI models designed to respond to customer complaints about pricing, data handling, and overly cautious content filters. Fable 5.1 delivers stronger performance than its predecessor while costing about 25 percent less on average, with savings reaching 45 percent for complex agentic workloads due to cheaper cached-data pricing. Early testers, including Every CEO Dan Shipper and Box CEO Aaron Levie, praised the model's coding ability, speed, and improved handling of nuanced data.
Perplexity has extended its agentic AI feature, previously exclusive to Mac, to Windows 10 and 11 users on paid Pro, Max, or Enterprise plans. The tool can access local files, control native Windows applications, and connect to cloud services like OneDrive, Google Drive, Box, and Dropbox to complete multi-step tasks with minimal user input. A ZDNET reviewer tested it on five complex tasks to gauge its real-world usefulness.