A common theme in our 21st century is the amplification of enabling technologies. In the optical fiber world, for example, it is increasing the distances between amplifying repeaters through the use of erbium-doped fiber. In the cloud computing world, it is increasing the total amount of computation available to users to accomplish goals more quickly. A good example is the use of massive numbers of processors to index the World Wide Web, enabling users to find information in fractions of a second that would otherwise be buried forever in exabytes of data. In the machine learning world, amplification arises through training of deep neural networks with massive numbers of examples allowing the trained network to discriminate among and classify an endless array of samples. The power of social networking is evidenced by the amplifying effect of these services to spur collaboration and coordinated action. Some neural networks amplify our individual capacity to understand languages we don’t speak through automatic translation.
The Internet itself is an amplifier of voices that might otherwise remain unheard. Its very existence is evidence of the amplifying effect of network collaboration in the form of network interconnection, often accomplished through Internet exchange points (IXPs). The “network effect” noted by Robert Metcalfe claims that the amplification grows proportionally to the square of the number of users.
These examples of amplification explain the power and popularity of so many online services. It must be observed, however, that not all amplification is beneficial. Network denial of service attacks are amplified by the hijacking of millions of Internet-enabled devices, such as laptops, webcams and other programmable devices. Malware is propagated by means of compromised network servers. Spam email is generated and propagated by so-called bots. Misinformation and disinformation is propagated and repeated through the use of the same tools that enable beneficial collaboration.