
This article on News.com details how Google is offering a web accelerator service to speed up internet speed. Since Google caches other people’s web site pages frequently, whenever your browser makes a request for, say, “www.cnn.com”, it will request the page from Google’s cache instead of hitting cnn.com directly. I’m not sure how it will handle dynamically-driven sites, or when sites whose pages change frequently, like CNN’s because the news is always changing, come from the Google cache or cause the browser to go fetch it from CNN.com directly.
There is one potential hidden benefit to this all for Google: each user is a crawler for Google, so if a page is not in Google’s cache index, the browser could pull the page down from the author’s site directly (like a browser normaly would), then upload the content to Google directly, doing Google’s crawling work for them.
Also, Google gets data about who is going to what sites, very valuable data to any advertiser or marketer.
Google Labs Web Accelerator page