The document discusses how aggregators on the internet can lead to "SpamBack" if they are not properly programmed. It describes how a feedback loop could be created by posting content from one service to another and back again, without recognizing that the content has already been seen. To prevent this, the document suggests that aggregators check for a unique content ID and drop any content that has already been associated with that ID. As more services aggregate content, addressing this issue of SpamBack will become more important to avoid overloading systems on the internet.