Machine Learning technology as a tool for manipulation
Net Neutrality has been a hot topic since the regulations to check the censorship and monopoly of telecommunications has passed in 2015. However, with the recent shift in leadership, FCC has been trying to repeat the net neutrality act and return the power to the monopolizing companies. The article linked above exposes the extent which the telecom. Companies and other pro-repealers do to try to repeal the act. The author states that often, pro-repeal comments on net neutrality articles and threads are often fake and generated by spambots – bots that uses machine learning to generate coherent sentences for the purpose of spamming.
He detected spambot patterns by first finding similar sounding phrases in various difficult comments, then detecting repeated sentence structures. Through this pattern detecting, the author was able to detect more than a million spambot generated pro-repeat comments on net neutrality articles; these included only those which he were able to detect. Personally, reading the spambot generated comments, they seemed very normal and not machine-like. If the reader was not reading a lot of the comments generated by the same spambot, it would be difficult for them to tell whether or not these were written by a chatbot or a machine.
People’s opinions are influenced by what the they read and what they see majority thinks. People’s votes may change depending on the opinion which they were exposed to most of the time. The false comments, though seemingly harmless, it can have networks effect and convince people to change their minds based on myriads of pro-repeal comments which may effectively brainwash those people who jump on the bandwagon. The only solution to prevent the further spread of false majority is to increase awareness in topics such as net neutrality. Topics such as net neutrality may seem irrelevant to some people, but it will ultimately affect so much in a bad way since without net neutrality, the telecommunications companies will be less regulated and be presenting biased, expensive information.