AI news roundup: Meta's Muse Code, Google's research exit, and Anthropic's $10B cloud bet

Meta launched an AI agent built for large code bases, top Google researchers are leaving to start a science company, and Anthropic is spending billions on cloud capacity and custom chips. Shopify's numbers also suggest AI search is becoming a real sales channel. Each story points the same way: AI is moving from chat into business infrastructure.

Meta launches Muse Code for large code bases

Meta has expanded its AI coding lineup with Muse Code, an agent for complex tasks in large code bases. Most coding assistants handle small, well-scoped jobs. Muse Code targets big software projects, where one change can ripple through many files, as TechCrunch reported on Wednesday.

OpenAI, Google and GitHub already ship coding agents of their own. Meta's bet is that huge repositories need a different tool from the one that writes a quick function. The launch puts Meta in the enterprise developer market, one of the most visible AI battlegrounds.

Google's top researchers leave to start a science company

Jeff Dean, the longtime Google engineer who helped build Google Brain, is leaving the company along with other senior AI researchers. Their plan, reported by TechCrunch, is to launch a startup that applies AI to scientific discovery.

Dean helped shape modern machine learning, so the departure signals where senior talent thinks the next big opportunity sits. Scientific discovery is a growing industry focus, from protein-folding models to AI systems that propose new materials. A startup founded by people who shaped deep learning will not struggle for attention or funding.

Anthropic bets on its own chips and a $10 billion cloud deal

Anthropic has signed a reported US$10 billion deal with AI cloud startup Volta, the latest in a string of cloud partnerships. It is also hiring an AI chip design team to co-design hardware and models so Claude runs faster and more efficiently.

The two moves point in the same direction: Anthropic wants more control over its computing future. Frontier models are expensive to train and run, and dependence on one cloud provider is risky. Custom silicon is a long game, but the chip team shows Anthropic is attacking cost at the hardware level, not just the model.

Shopify says AI search is driving sales, not stealing them

Shopify says AI-driven traffic and orders to its stores tripled year over year in Q2. The company argues AI search is not cannibalising traffic the way it has for publishers: shoppers find products through AI assistants and still buy through Shopify stores.

This is the week's most practical AI story for small businesses. As shoppers ask AI assistants where to buy, stores need product pages and listings that AI systems can read and summarise clearly. The merchants found in AI answers tend to have clean, complete product data. That is a job most businesses can start today without buying new tools.

The thread running through today

Four stories, one theme: AI is moving from chat into the plumbing of business. Meta wants to own the developer workflow, Google's researchers see science as the next frontier, Anthropic is building the hardware layer it needs, and Shopify's numbers suggest AI is becoming a real sales channel. For Australian businesses the practical signal is the same as it has been all year. Make your content and product data easy for AI systems to understand, because that is how more customers will find you.

This article was written by Triweb AI's editorial team based on analysis of today's leading AI news sources.