The best answer I got in talking to Google about recent launches is that people expect capabilities like deep research and canvas out of modern LLM-powered offerings. That does make sense, with the other new feature announced for AI Mode being the ability to upload a PDF with your prompt.

Deep Think is based on the same foundation as Gemini 2.5 Pro, but it increases the "thinking time" with greater parallel analysis. According to Google, Deep Think explores multiple approaches to a problem, even revisiting and remixing the various hypotheses it generates. This process helps it create a higher-quality output.