Key Takeaways
- Google Search is testing a new feature called “What People Suggest” for health-related queries.
- It pulls information from online discussions on sites like Reddit, Quora, and X.
- The goal is to provide insights from people with lived experience alongside authoritative sources.
- Google uses AI (Gemini) to filter suggestions for safety and relevance, excluding harmful or misleading content.
- This feature is currently only available to mobile users in the United States.
- Concerns exist about the potential for spreading misinformation despite Google’s filtering efforts.
- The launch comes as AI chatbots like ChatGPT increasingly compete with traditional search engines.
Google Search, often called “Dr. Google,” is experimenting with a new way to answer health questions by including advice shared by the public.
The feature, named “What People Suggest,” gathers information from online conversations found on forums such as Reddit, Quora, and X (formerly Twitter). This content is then analyzed and filtered before being presented to users.
Currently, this new section is only visible to mobile users within the U.S. Google says it introduced this feature because users asked for real-world perspectives alongside expert medical information.
A major concern is whether sourcing information from online opinions could inadvertently spread incorrect or even dangerous health advice. Imagine someone seeking alternative cancer treatments or asking if vitamins can prevent diseases like measles.
In an interview highlighted by Forbes, Google’s chief clinical officer, Dr. Michael Howell, addressed these concerns. He explained the feature’s purpose and Google’s commitment to ensuring the information is helpful and accurate.
Howell emphasized that people seek both authoritative medical facts and the “lived experience” of others dealing with similar conditions. “We want to help them find that information as easily as possible,” he said.
Google isn’t showing these suggestions for every health search. If a query suggests an emergency, like symptoms of a heart attack, the feature won’t appear. Instead, users will see prominent links to established sources like the Mayo Clinic.
Similarly, searches related to suicide prioritize resources like the 988 Suicide and Crisis Lifeline. Suggestions about prescription drugs or specific medical procedures are also excluded.
To filter the content, Google uses its Gemini AI model. This AI scans online forums, identifies suggestions related to specific conditions, and groups them. A specialized AI application then checks if the advice aligns with medical consensus and avoids harmful content.
Before launching, Google also had panels of physicians review sample responses generated by the AI to assess their accuracy and helpfulness.
The suggestions that pass these checks are presented as short, clickable descriptions within a boxed area further down the search results page, not right at the top. There’s also a feedback button for users.
Dr. Howell believes this approach taps into the valuable insights people gain from managing their own health conditions. He showed how a suggestion for using “colloidal oatmeal” for eczema itching, sourced from patient discussions, is actually supported by scientific research.
However, some experts remain wary. Dr. Danny Sands, an advocate for patient involvement in healthcare, expressed concern to Forbes about whether widely shared online “common wisdom” is always truly wise.
This move by Google happens at a time when AI chatbots like ChatGPT are gaining popularity and presenting a challenge to traditional search engines for information seeking.
Health queries represent a significant portion of Google searches and contribute substantially to its revenue. Keeping users engaged is vital, especially as the company faces competitive and regulatory pressures.
While Google maintains that search and chatbots offer different experiences, evidence suggests chatbots are attracting users, partly due to their conversational style and concise answers. Some early research indicates users might even trust chatbot responses more than traditional search results.
This new feature fits into Google Health’s broader ambitions to leverage technology across its platforms, including Search, YouTube, and Fitbit, potentially using “lived experience” data to generate insights and personalized health information in the future.
