Original Source Here
Why Don’t We See AI In Politics?
For the past year, I have been a technology and sponsorship director at a student-led organization called Girls For Politics and although I’m not involved immensely in the articles and interviews published from the organization, I have learned much more about how complex the field of politics actually is. One important observation is that AI seems to be absent in the field in general, at least at first glance. So, to answer the initial question about why we don’t see AI in politics, well…..we do. But, not in the way you’d think. Compared to other areas of study, AI is very much lacking in politics, mainly because of the risks that come with incorporating AI into politics. While AI is the star of the show in other fields like medicine and security, in politics, it is nothing more than a backstage tech worker adjusting the lights and opening the curtains. This difference in the role of AI reflects the inherently social nature of politics—there are opinions, bias, debates, and reputations at stake when AI is incorporated into a social science like politics, making its involvement a risky strategy for political campaigners.
However, let’s look at how AI is used in politics, because AI is not at all absent from the political sphere.
Well let’s first examine the primary, obvious usage: analyzing data. When a political candidate is campaigning, they will try to attract voters by fighting for causes their voters are interested in and understanding what issues they can work to solve. In the past, trying to figure out what causes voters feel strongly about has been more of a guessing game for politicians, but with the rise of AI and the availability of public data regarding political alignment and social interests, candidates can use machine learning to have a clearer understanding of how to bring in larger crowds and their donations and what causes to focus on while campaigning.
For instance, researchers from the Near East University developed a model that is believed to “associate opinions of American congress members on defined national issues with their political party affiliation as republican or democrat, thus providing an artificially intelligent anticipation system of the congressional voting outcome.”
In addition to the algorithm created by the scientists of the Near East University, the Washington Post also revealed that machine learning played a role in Hillary Clinton’s 2016 presidential campaign. Clinton’s team used an algorithm called Ada which, the Washington Post says would “run 400,000 simulations a day of what the race against Trump might look like.” If there were signs that Michigan or Pennsylvania was an important state for Clinton to win in, her team would put more of an emphasis in being active in those states. If a candidate running for office utilized a similar technique with the general population, they would have an upper hand with relating to their audience.
However, as we now know, Clinton lost against Trump in 2016, so what went wrong? Well, it would be precarious to assume that Trump didn’t utilize the power of AI in the election for his cause as well, although you could assume he chose a different strategy. According to the New York Times, Trump’s team released hordes of automated chatbots to overpower Clinton’s messages and pro-Clinton discussions and groups. Apparently, Trump’s chatbots had no other purpose besides ranting, joking about Clinton’s campaign, confuse users, and suppress discussions about the election that may have been pro-Clinton. Regardless of what the bots were made to complete, their impact on the election was drastic, assisting Trump in his victory even though Clinton was above him in the polls most of the time.
Even in Obama’s campaign, “Obama’s campaign team implemented Big Data Analytics and poked his voters. And it maximized the effectiveness of his email campaigns raising a huge US$1B in campaign donations.” The ability of politicians to utilize AI to have some more transparency into what their voters want to see has relevant effects.
In addition to understanding what voting audiences feels strongly about, machine learning is also used for targeted ads to reach a wider variety of people, although this has been a topic of controversy recently. In fact, the NYU Online Political Transparency Project collected data from different web users about the types of ads they received on Facebook related to politics, and ProPublica used the data to create a simulation for the types of ads one would see depending on their age, gender, background, location, political alignment, and more. Recently, however, large social media platforms like Instagram and Twitter have set limitations on advertisements that can be displayed, if not restricting all political ads. Facebook continues to allow almost all advertising with little supervision, which makes it a promising place for candidates to target certain groups of people, based on traits similar to the ones used in the NYUOPTP database.
AI has had massive effects in the field of politics, although it doesn’t usually get enough of the credit. In reality, the usage of a machine learning algorithm on the behalf of a campaign team can be a difference of billions of dollars or an entire election victory, and its effectiveness has made it a worthwhile investment for candidates. Each day, AI is proving to be a more constructive tool for politics, and it will continue progressing the field as a whole, likely in ways that will make it more powerful of a weapon for candidates.
As Microsoft’s Danah Boyd said:
“There will be abuses of power that involve AI, just as there will be advances in science and humanitarian efforts that also involve AI”
Trending AI/ML Article Identified & Digested via Granola by Ramsey Elbasheer; a Machine-Driven RSS Bot