Around the world this week

Google's AI can read phone screens now
Google has started introducing new artificial intelligence features to its Gemini Live platform, enabling real-time interpretation of smartphone screens and camera feeds. These enhancements, part of the broader “Project Astra” initiative, were confirmed by company spokesperson Alex Joseph. Among the upgrades are screen-reading and live video analysis tools. These allow users to ask questions about what’s currently visible on their screen or through the camera lens, receiving context-aware responses. The features are being made available exclusively to Gemini Advanced users subscribed to the Google One AI Premium plan, with a gradual rollout expected over the month. Joseph explained that the screen-reading function helps users interact with any visible on-screen content, while the live video tool can identify objects, offer design advice, or assist with tasks such as selecting paint colours.

Chinese firms are slashing AI costs
Alibaba-affiliate Ant Group is using both Chinese and US-made semiconductors for building more efficient artificial intelligence models, according to media reports. The combination of chips not only reduces the time and cost of training AI models, but also limits reliance on a single supplier such as Nvidia, the reports said, noting the industry trend of tapping multiple networks, known as mixture of experts — a technique that allows models to be trained with much less compute.