Web scale reasoning refers to systems that attempt to use the knowledge available on the web, or corpera of similar scale, as part of formal reasining mechanisms. This is the goal of the semantic web and linked data, cxreating machine readable web data that can then be used as a vast knowledge base for AI, that can be accessed on demand. However, large language models and other large sacledeep learning approaches can be seen as a form of web-scale reasining where the 'reasononh' is enabled by fully processing vast amounts of data rather than treating it as a formal knowledge base.
Used on Chap. 17: page 417