Google is updating its search functionality to enhance its AI experience by incorporating more context to links, such as excerpts from forums and blogs, alongside a feature that emphasizes links from a user’s news subscriptions.
While referencing forums and discussion boards can assist users in finding answers to niche questions, this approach may introduce chaos.
Two years ago, Google revamped its search to prioritize AI, often generating an “AI Overview” for queries, which received mixed feedback. Users noted its susceptibility to sarcasm and questionable sources, citing examples like using The Onion for dietary advice or Reddit suggesting glue on pizza for cheese adherence.
Despite improvements in Google’s AI Overviews, they’re still prone to hallucinations. A recent analysis by the New York Times showed the AI was accurate nine times out of 10. Given Google’s vast query processing, this rate could result in numerous inaccuracies every minute.
Not all queries have clear-cut answers, leading Google to draw insights from forums where such topics are discussed. As Google explains, more people seek advice from others, thus AI responses will now preview perspectives from online discussions, social media, and firsthand sources. Additional context, like a creator’s name or community, aids in evaluating which discussions to engage with.
This new approach complicates the role of AI Overviews. Are they meant to provide answers or a collection of potential sources for information seekers? This resembles traditional Google searches.
Google plans to add context to AI Overview sources, helping users assess source trustworthiness, similar to how ChatGPT or Claude present supporting links. Nonetheless, it’s advisable to verify these citations’ validity against potential hallucinations.