0
0
During Google I/O, as I live-blogged the event, a memory resurfaced—my spouse’s successful bet that ChatGPT could mimic my writing style better than I thought possible. It reminded me of the complexities of asking AI the right questions. Google’s emphasis on AI’s potential impact on our lives during the keynote further intensified this realization. While Google showcased innovative features like Sidekick and Search Generative Experience (SGE), which provide proactive prompts and tailored search results, I couldn’t help but feel unsettled as a professional question-asker. The queries and prompts presented by Google didn’t align with my own search habits and questions, raising concerns about the evolving role of AI in our lives.
The Challenge of Asking the Right Questions:
Google’s demonstration of Duet AI highlighted how prompts can guide users on how to interact with AI tools effectively. However, as someone who regularly relies on asking questions, I found myself feeling out of sync. The queries and prompts displayed on stage were vastly different from the toddler-like questions I often type into the search bar, supplemented with “Reddit” to access non-SEO content. My approach involves unconventional phrases like “Bald Dennis BlackBerry movie actor name” or “Site:theverge.com Peloton McCarthy ship metaphors” when seeking specific information. In contrast, Google presented more generic questions like “What should I do in Paris for a weekend?”—a style that rarely crosses my mind when using search engines.
The Dilemma of Generative AI:
Generative AI, such as ChatGPT, presents another challenge. Even after witnessing countless demos, I find myself unsure of how to leverage its capabilities. It feels reminiscent of being called on by a stern teacher in second grade, unable to answer the question. When I do ask something, the results often fall short, requiring more effort to refine than if I had done the task manually.
Unveiling Google’s Vision:
Google’s vision extends beyond ChatGPT-like models. Duet AI aims to extract contextual information from emails and documents to intuit users’ needs—an ambitious endeavor considering how frequently I struggle to define my own requirements. SGE, on the other hand, strives to answer questions, including those without clear-cut answers, and anticipate follow-up inquiries. Achieving this level of intuition requires programmers to design AI systems that anticipate users’ questions before they even ask them. The complexity of this task is enough to give anyone a headache.
The Quest for Meaningful AI:
In a more philosophical sense, life is about discovering the right questions to ask. However, in the AI era, defining our expectations from AI remains challenging. Google, OpenAI, Microsoft, and other industry players have their own visions of AI’s purpose. Yet, when engaging in conversations about AI with everyday individuals, the common question that arises is simple: “How will AI transform and impact our lives?” The answer to this question eludes us, leaving us uncertain about what we truly desire from AI.
Conclusion:
As Google showcased its advancements at I/O, the challenge of asking the right questions in the AI era became evident. AI tools like Duet AI and SGE offer impressive features but can leave question-askers feeling out of sync. The ability to extract contextual information, predict users’ needs, and generate intuitive responses raises the bar for programmers and users alike. Furthermore, in our quest for meaningful AI, we must grapple with the uncertainty of defining our expectations and understanding how AI will shape our lives. As we navigate this AI-driven landscape, the quest for the right questions continues, challenging us to define our desires and expectations from AI in a rapidly evolving world.
