In 2012 Barack Obama and Mitt Romney were opponents in the election for becoming the president of the United States of America (Bergdahl, 2019). Bergdahl (2019) mentions that the results generally depend on a few swing states. Therefore, Barack Obama performed the biggest Artifical Intelligence (AI) operation ever observed by that time in politics. He hired a machine learning professional, called Rayid Ghani, to gather voter data from sources like social media. Using this data, Rayidwas able to estimate how likely voters were to vote for Obama, how easy voters would change their vote decision, and how likely those people were to actually be at the polling station at the election day (Domingos 2017). Based on simulations results, they were able to find specific and convince voters to vote for Barack Obama. Romney knew Obama was up to something by going targeting specific neighborhoods, but he did not know the reason behind it.  Consequently, Obama won the elections (Bergdahl, 2019).

Currently, you’re probably well aware of companies making product and service  recommendations for their customers using their data. As mentioned in the previous example, Politicans do exactly the same. Four years after Barack Obama won the election, Donald Trump would execute an even bigger AI operation to defeat Hilary Clinton. The organization Cambridge Analytica created 220 million voter profiles for adults in the USA, using 5000 data points (Anderson, 2017). These profiles enabled Cambridge Analytica to present specific adds to voters. For instance, if a voter is actively supporting your party, positive news according this party can be used to encourage to go vote. Conversely, voters that actively support the opponent can be encouraged not to vote by emphazing mistakes made by the opponent (Bergdahl, 2019). The most incredible thing about targeting voters using AI, is you’re in complete control of your own adds. Opponents will never know what sort of adds are shown to their voters and therefore cannot influence your strategy.

AI is the biggest threat to democracy today, and the strongest tool for anyone aiming to win an election. How do you think about collecting data on citizens and manipulation opinions through microtargeting to win an election? Is this ethical?

References:

Anderson, B. (2017). The Rise of the Weaponized AI Propaganda Machine. [online] Medium. Available at: https://medium.com/join-scout/the-rise-of-the-weaponized-ai-propaganda-machine-86dac61668b [Accessed 11 Oct. 2019].

Bergdahl, J. (2019). How AI Can Make You The President. [online] Medium. Available at: https://towardsdatascience.com/how-ai-can-make-you-the-president-4756f6b1c0c0 [Accessed 11 Oct. 2019].

Domingos, P. (2017). The Master Algorithm (2017). The Master Algorithm: How the Quest for the Ultimate Learning Machine Will Remake Our World. New York, America :Penguin Groups.

Moneyball is a movie based on a true story in which Billy Beane, the director of Oakland athletics, tries to find undervalued players using a new statistical approach for baseball. This is because Oakland lost a few important players after the play-offs in the season of 2001. By chance Billy meets Peter Brand, a young economist who focuses on statistical idea’s about how to evaluate players. Using this statistical approach the Oakland athletics win 20 consecutive games even though they have a small budget (wikipedia, 2018). This is can be seen as a initial example of data-driven performance optimization. This presents how data can atone for a lack of money and/or resources and add to performance by more adequate decision-making. After this occurence the sport world has emerged tremendously. Different sorts of technology, such as AI are now integrated in sport analytics, since so many things in sports are quantifiable. A few examples of AI in sports are: scouting and recruitment, training and performance, maintaining player health and fitness, and broadcasting and advertising (Forbes, 2019).

Even though, it is almost impossible to evaluate humans, it is possible to quantify their performance with AI. By using individual and team performance data, players’ capacity and potential can be measured. Here, we are not talking about common stats, such as home runs or amount of hits, but about using complex metrics. This allows to include for numerous factors, which humans can only partly take into consideration. This way performance improvement can become more easy and reliable, since it enables coaches to determine players strengths and weaknesses more accurate. Also, AI enables the use of historical data, to forecast players market value and optimal performance. Lastly, AI can help in the preparation for matches by identifying patterns in for instance antagonist’s tactics. All of this makes it possible to predict sports more reliable.

However, humans behavior and therefore performance still has some unpredictability. Will at some point human behavior /sport performance become completly predicatable? And if sport performance becomes completly predictable will it still be exciting to watch?

references:

Forbes. (2019). Here’s How AI Will Change The World Of Sports!. [online] Available at: https://www.forbes.com/sites/cognitiveworld/2019/03/15/heres-how-ai-will-change-the-world-of-sports/#3861c18a556b [Accessed 1 Oct. 2019].

Wikipedia. (2019). Moneyball. [online] Available at: https://nl.wikipedia.org/wiki/Moneyball [Accessed 1 Oct. 2019].