If Web 1.0 was about content, and Web 2.0 was about people, the Next Web is definitely about context. The value of bringing content and people together is to know what content to deliver to which people, and at what time. To predict what content should be delivered companies have to base their decisions on the prior activities and feedback of their customers.
Companies like Netflix and Amazon have figured this out and are working hard to fine-tune their algorithms that recommend movies and products to customers based on their previous selections and their satisfaction with the product (usually measured through ratings). In this featured video, Heather Knight of MarilynMonrobot.com does a great demonstration of a robot that tells jokes and adjusts the content it delivers based on the feedback of the audience.
It will be interesting to see how these concepts can further advance the collaborative feedback that is received via social media. Companies should be paying close attention and integrating this type of reactive analysis to proactively adjust their products and services.