AI Scale, Security Crises, and Robot Ambitions Define This Week in Tech

This week in technology brought a mix of significant advancements and concerning developments. AI continued its rapid expansion, with Google's Gemini achieving 1 billion monthly users and Apple negotiating multiyear deals for real-time news to enhance Siri. OpenAI prepared to release a portable AI smart speaker, while ByteDance trained a model with 10 trillion parameters.

Microsoft unveiled its Maia 300 chip in a bid to challenge Nvidia's dominance, and Amazon backed a natural-gas plant to power an off-grid AI data center in Texas, raising environmental concerns. China's Fengwu AI model forecasted Typhoon Dolphin's landfall with five days' notice, though its accuracy in predicting intensity was limited.

Biotech researchers at Stanford and Arc Institute created complete viral genomes using Evo genomic models, potentially opening new therapeutic avenues but also raising biosecurity concerns. The farmer who lost 25 acres of sesame to AI-recommended chemicals highlighted the need for independent verification in automated agricultural guidance.

Anthropic added machine-readable watermarks to Claude models for better transparency, and Agility Robotics' Digit v5 humanoid robot was set to work for extended periods. On Android, Aptoide Games became the first third-party storefront in the US Google Play Store, while automated publishing services like RuntimeWire published news faster than traditional outlets.

However, the rapid deployment of AI infrastructure came with significant energy demands, as seen with Amazon's proposed 7.65-gigawatt natural-gas plant. Financial institutions faced a coordinated social-engineering campaign targeting over 200 companies, and ShieldBreak zero-day exploits compromised Microsoft Defender on fully patched Windows systems.