The folks over at Microsoft Bing put together a blog post explaining the differences between indexing for Search versus indexing for Grounding (AI responses) and the differences. Krishna Madhavan, Knut Risvik, Meenaz Merchant from Microsoft wrote, "Indexing for grounded AI answers is not a reinvention of search – it is a major evolution of it. Grounding commits to an answer."
That is a big statement that seems very different from Google's messaging around similar topics.
"The shift we described at the opening is worth restating plainly: search indexing was built to help humans decide what to read. Grounding indexing is being built to help AI systems decide what to say. The infrastructure required to do those two things well is not the same – even when it starts from the same foundation," Microsoft Bing's team wrote.
Traditional search asks: which pages should a user visit? Grounding asks: what information can an AI system responsibly use to construct a response? These questions sound similar. They are not. Microsoft posted this table with examples:
This means you need to rethink what "index quality" means - Microsoft wrote. Here is another grid showing this:
"A common misconception is that grounding replaces search. It does not. Grounding builds on the same foundational infrastructure – the same crawlers, the same quality signals, the same deep understanding of the web – but it adds a new optimization layer on top," they also wrote.
The team ended by saying, "Search optimizes for likelihood of relevance. Grounding must measure strength of evidence. Understanding that difference is not just an engineering concern. It is the starting point for building AI systems that people can actually trust."
Krishna Madhavan from Microsoft added on LinkedIn, "In our new blog post published today, we examine the transformation that occurs when the index evolves from merely assisting users in finding pages to aiding AI systems in providing answers. Co-authored with Knut Risvik and Meenaz Merchant, we discuss how the unit of value transitions from documents to evidence, and why this shift is crucial for establishing AI that people can genuinely trust."
Jordi Ribas from Microsoft added on X, "In the era of the agentic web, the role of the web index needs to evolve to support very different needs across agents and humans."
What do you all think?
