Bots of conviction
Another type of bots is discussed by Mark Sample, who enunciates bots of conviction. Inspired by Jurgen Habemas' “journalism of conviction”, which was polemical, political and constantly debating the needs of society and the role of the state, “bots of conviction” are topical, data-based, cumulative, oppositional, and often uncanny.
He gives the example of NRA Tally, a bot that generates and tweets school shootings as a response to the murders of Elliot Rodger who on May 23, 2014 killed six people and injured fourteen more near the campus of UC-Santa Barbara.
The parameters of the bot's tweets are not accidental: all of them are drawn from real data of mass shootings that have happened in the U.S. For example, it selects a number between 4 (the threshold for what the FBI defines as mass murder) and 35 (just above the Virginia Tech massacre, the worst mass shooting in American history). The location, the victims, the response from the NRA are all variables which have at some point defined such an event.
Clearly NRA_Tally refers to a frustration that is shared by many Americans. However, one might wonder what is the real reach of the bot. Sample calls it a tactical bot because it challenges the NRA without drawing attention away from the victims of the mass murders. One might wonder why there was a need to generate mass murders, when there is already quite a substantial existing list.
Parliament Edits Bot
Other cases where bots are used as social vigilantes, one example being the Parliament Edits bot. If you edit on Wikipedia without making an account, your IP address will be made public. Journalist and developer Tom Scott exploited this paradoxical situation to make a bot that will look at recent edits made by anonymous users and retweet the ones that were made in the IP address range belonging to the British Parliament.
The bot was resurrected by Ed Summers in 2014, after the IP addresses had been intentionally modified following a rejected Freedom of Information Request. Summers made the code available on Github and instructed how users could modify the IP addresses and make their own bot. This inspired multiple other iterations, among which the most well-known one is congress-edits.
It turned out that most edits made by these IP addresses were fixing punctuation, spelling or grammar. Nonetheless, they sparked the imagination of many others which installed their own instance of the bot to patrol their national governments.
Arguetron is another example of informed coding by Sarah Nyberg. It started after identifying the emotional labour that goes into having discussions and arguments with inflammatory commentators on the internet. So she decided to create a bot called Liz that posts comments that appear extremely left leaning. Soon enough, human users who are looking for specific keywords will engage with the bot in an argument. She calls them honeybots.
Botivist: project to help activists find volunteers on Twitter
In a similar vein, Botivist is another project that was initiated by Saiph Savage, Tobias Hollerer and Andres Hernandez, the last of which is associated with the Microsoft company. The project was looking to alleviate the labour of engaging volunteers by applying face-to-face strategies of communication to possible interested parties to the Twitter platform. The group attempted to address activists in Mexico that were tweeting about corruption and impunity. The findings were not particularly surprising in that the strategy that was most succesful was the direct approach. The more transparent, and the less empathetic the bot, the more succesful.
As opposed to propaganda bots, Sample's bots of conviction can also be considered rhetoric bots, where the same process of persuasion takes shape but with different variables:
- revealed intent versus hidden intent
- dialogic relationship versus unilateral communication
- technique used to make the truth effective versus effective technique to hide the truth
Other examples include:
Bots of conviction tend well towards journalistic practices, many of them take on the task of making data legible for non-technically minded people.