CEO Sundar Pichai's cheeseburger flex sums up Google's amazing comeback (www.businessinsider.com)

Google this week unveiled Gemini 3 and an upgraded image generator called Nano Banana Pro, marking a visible leap in its generative-AI capabilities. The new models demonstrate much stronger spatial reasoning and 3D understanding, producing realistic photos, diagrams and charts that fix longstanding image-generation quirks — memorably illustrated when CEO Sundar Pichai shared an AI-rendered cheeseburger with the cheese correctly placed above the patty, a tongue-in-cheek capstone to a 2017 emoji controversy. The launches helped restore market confidence (Google’s stock hit a record and briefly topped Microsoft by market value) and refocus attention on Google’s years-long “AI-first” engineering push. Technically, Gemini 3 and Nano Banana Pro show progress in modeling object relationships, spatial orientation and physics priors — areas where prior generative systems often erred. That improvement matters beyond novelty images: better implicit 3D/world models can translate into more reliable decision-making in design, engineering and safety-critical placement tasks (for example, recommending optimal siting for barriers or components down to fine tolerances). The releases also reinforce the idea that scale plus long-term infrastructure and research investment can yield practical advances in multimodal reasoning, reasserting Google’s competitive position in generative AI.