Insights from a Former Bing Search Director: Transforming Search in the AI Era
In the ever-evolving world of search engines, it’s easy to forget just how far we’ve come. Stefan Weitz, the former director of Microsoft’s Bing search engine, recently shared his thoughts on the state of search in the AI era. Despite significant advancements, Weitz believes that search is still broken and has room for improvement.
Weitz points out that while search engines are great for many things, they still struggle to effectively help users accomplish their goals. He argues that search feels like too much work for complex tasks, with users needing to analyze results and perform multiple queries to dig deeper or take action. The idea of connecting queries and intent through action, which was pursued during Weitz’s time at Bing, is still not fully realized.
One interesting finding is that people are now looking through more search results compared to five years ago. This suggests that users are becoming more discerning and demanding when it comes to finding the information they need. Weitz believes that voice agents, which were once seen as revolutionary, are not the ultimate solution either. While language models (LLMs) have made great strides, they still lack true knowledge and struggle to understand what users are asking for.
Looking to the future, Weitz emphasizes the importance of AI overviews and answer engines. Google and its competitors are continuously developing new AI search experiences, but the key to success lies in ensuring that these experiences truly add value for users. Weitz warns that if people start using a technology and it fails, they may lose trust in it and abandon it altogether.
Changing user behavior is a challenge in search innovation. Weitz acknowledges that his own search behavior has changed in the past two years, and he expects this trend to continue for more users. The rise of multimodal search, combining text, images, and videos, is becoming a natural way for users to find answers to complex questions. Traditional search, which often returns generic results for specific queries, is seen as a pain.
If given the opportunity to improve Google Search, Weitz would focus on several key areas. Firstly, he believes there is a need for better understanding of queries and providing accurate answers. Generative AI tools have made progress in understanding complex sentences, but there is still work to be done. Additionally, Weitz emphasizes the importance of task orientation, where search engines provide follow-on actions for users after identifying an issue. Lastly, search should be less device-specific and aim to become a true assistant that helps users remember information they have encountered in the past.
Overall, Weitz’s insights highlight the ongoing challenges and opportunities in transforming search in the AI era. While progress has been made, there is still work to be done to create a seamless and efficient search experience that truly meets users’ needs. As technology continues to advance, it will be interesting to see how search engines evolve and adapt to the ever-changing demands of users.