Prompt: "How does Baidu's Ernie model compare to the other major ones?" Response: "Summary: Baidu’s ERNIE is a powerful, knowledge-augmented LLM optimized for Chinese-language performance and integrated into Baidu’s services. While it may not match GPT-4’s global versatility in English, it excels in niche areas like Chinese search, BERT-style understanding, and knowledge integration. As AI competition grows, ERNIE represents a regional leader, whereas models like GPT-4 aim for universal coverage. The best model depends on the use case—language, task, and regional constraints." Google AI Mode: "In summary, Baidu's ERNIE models, particularly the recent versions, are presenting a significant challenge to other major AI models by offering comparable or even superior performance in certain areas, especially multimodal tasks and Chinese language processing, at a substantially lower cost, coupled with an open-source approach. This strategy is likely to intensify competition and accelerate the development and democratization of advanced AI globally."

(Gemini Deep Research: review of Callard, 2025, from secondary sources. The conclusion sums up the terms. See the book's Table of Contents for more. Also, Gemini 2.5 Pro can look for uniqueness of the author's interpretation of the Socratic Method. Or the dialectical styles and tensions of the popular Chatbots, for that matter. Gemini Personalization model can talk about the user's, hinting at core beliefs/interests, based on their history. Related references might include Penn's Writing the Shadow, or Allen's Why Plato Wrote. On a rhetorical triangle of speaker/content/audience, this might be somewhere between Quora and POE. Rather than only a synthesis like Plato or Hegel, or aporia like Derrida, this might look for alternatives. Google AI Mode casts this framework as a contrast to Aristotle's eudaimonia based on endoxa.)

Gemini Deep Research prompt: What are the different types of problem sets from the AI perspective, including new capabilities such as embodiment, sensors, and agents? (Research plan on 2.0 looks similar to 1.5. Initial claims from training were sequentially verified against different search sites. Growing number of references. Perhaps the idea would be to figure out a starting prompt that will then take it through skill acquisition to grand challenges, or exploit multiple agents in parallel, or cover sectors or countries, or integrate datasets, or find altogether new questions. Personalization could make it take on synthetic roles, not unlike early Eliza, but now headed for the C-Suite or whoever assures equity under ideology. Without triggering Hulk.)