The new, ChatGPT cop-piloted Bing Search has got a favorable market response, claimed Microsoft, one week after it was introduced for testing in 169 countries, along with the Edge Web browser.
Microsoft said in a blog post that as many as 71% of the initial testers had given AI-powered answers a “thumbs up”. There was increased engagement across traditional search results & with the new features like summarized answers, the new chat experience, & the Content creation tools.
Also, Microsoft had received feedback on how to improve. Some suggestions were:
Better Search & Answers: Users were happy with the citations & references that underly the answers in Bing. But the challenge was throwing up answers based on timely data like “live” sports scores. For queries where users were looking for direct & factual answers such as numbers from financial reports, Microsoft said it was now planning to 4x increase the grounding data in its ML model. It was also weighing the option of adding a toggle that will give users more control on the precision vs creativity of the answer to tailor to your query.
Chat. The ease of use and approachability of chat was an early success, said the IT major. There was also a lot of engagement which was delivering value for improving search & answers. A new use-case for chat was how people were using it as a tool for more general discovery of the world, & for social entertainment. This was a great example of where new technology was finding product-market-fit for something Microsoft had not fully envisioned.
What Microsoft also realized as part of its feedback was that in long, extended chat sessions of 15 or more questions, Bing could become repetitive or be prompted/provoked to give responses that were not necessarily helpful or in line with the designed tone.
