Google announced in a blog post that it will no longer answer silly questions from users. While it might sound a bit insulting, it might actually be a pretty good idea.
Since the beginning, Google aims to direct users to the information they need and lists the results it thinks are most relevant to the searched topic in the search results. However, in recent years, it gives the answer to the question it has asked users directly, without having to go to any website, just above the search results.
For example, Google the “How far is the Moon from Earth?” When we ask, “384,400 kilometers” appears on the screen in response. Although it is very successful in answering questions about well-known facts, when users ask questions that can be considered ridiculous, the algorithm fails and can display extremely ridiculous answers. For example, “Am I pregnant?” It is not possible for Google to answer such a question, but since it still has to give an answer, it can come up with a meaningless answer to users.
Google is aware of this situation and is working to fix it. “Using our latest artificial intelligence model, the Multitask Unified Model (MUM), our systems can now understand the concept of consensus where multiple high-quality sources on the web agree on the same fact and provide answers around it,” the dev team wrote in a blog post. He added: “We trained our system to detect false results. We were able to reduce incorrect responses by around 40 percent in this system.”
Google will also add more context to the “about this result” section, providing information about companies’ ownership, reviews of resources, and whether Google has found much information about a resource, helping users think a little more critically about the information.