East Asian Technology Intelligence
Japan & China tech news — translated, contextualized, and delivered for Western readers.
Free. Unsubscribe anytime.
This story ran in Issue #98, alongside three other stories.
Google rolled out Gemini 3.8 Live. The “Extended Thinking” version is a big step to make conversational AI act like an agent. It goes beyond a simple chatbot to automate business processes. The Japanese tech press, including *ITmedia AI+*, praised its ability to run background tasks during a conversation. This feature changes AI from a passive responder to an active tool. It can handle many steps or hard questions in real time.
For business users, the AI can draft emails, get data, or schedule meetings during a chat. This mimics human executive function. A human might nod and say “uh-huh” while starting a follow-up action. Gemini also supports 97 languages and can switch between them instantly. Global firms in Asia can use this right away for support, sales, and internal work. It cuts the friction of language barriers without slowing down the conversation.
Western analyses often focus only on benchmarks and model size. They miss how AI fits into daily work. This fit is vital in markets like Japan, where “omotenashi” or hospitality shapes service design. Gemini handles complex, multi-turn talks while managing backend tasks. This could change how Japanese firms automate customer support, where the flow of talk is key. It is the Japanese way to make AI polite and efficient for business.
However, these “Extended Thinking” features might stay trapped in demos. They will fail to boost output if IT departments cannot link them to legacy systems. A smart model is only as good as its connection to existing business data. Firms that treat this as a chatbot will miss the point. They must see it as a new way to automate their work.
To see the true impact, look for integration news from Japanese banks or factories. Track how fast firms in Asia adopt Gemini Enterprise or Workspace for core work. Do not just look at pilots. Watch for case studies with clear data on efficiency gains, not just praise. This will prove the move from conversational AI to automated workflows.
This story appeared in AsiaAI.FYI Issue #98.
Get this in your inbox each week — subscribe free.