THE NEGATIVE CoNSEQUENCES
DATA COLLECTION
While video games and Artificial Intelligence have undeniably helped each other greatly in the past decades, these advancements came at a great cost: Data collection. Indeed, as discussed previously the unfortunate consequence of developing Artificial Intelligence has always been the need for using extensive data, this consequence has not been avoided by the gaming industry. Furthermore, the latter has the advantage of being able of collecting data directly from their players, which usually are counted in the millions, for improving their games and algorithms for their next game. As with any instance of data collection, it is highly unappreciated, especially in the gaming industry which is often discrete about the fact that it collects their users’ data.
In fact, computer scientist at the IT University of Copenhagen and his team, Georgios Yannakakis conducted a study regarding player behavior in Tomb Raider: Underworld, a single-person game in which a weapon-wielding female archaeologist steals artifacts from ruins. They analyzed data from 10,000 players on the Xbox Live network (an internet-based network on which player can connect to compare their results, discuss etc.).
This is known as game-mining as the researchers searched for insights on human behavior in the data logs generated by video games, which reach a size quantified in terabytes. In fact, a single game session might involve thousands of button clicks and controller movements, each expressing a specific conscious or unconscious decision. Most of these decisions are minor, such as whether to go out of the way to pick up a coin or not. However, other inputs have major implications, like involving whether to help a game character in need, or what kind of taxation policy to employ for the simulated country you lead. (Bohannon, 2010)
Credits: KarmelSoft
In any case, Yannakakis’ analysis covered 35 different variables such as the use of weapons, the rate of progress, and whether it was tigers, traps, or other hazards that killed them. The goal of the study was to develop a computer which could predict the level at which any given player would eventually quit the game out of frustration. Once again, the underlying aim of the researchers was to create "personalized" games which can one day adapt to each player's abilities and interests. Their newly developed program was able to predict with 77% accuracy when any given player would give up. According to Yannakakis, the more data they can get their hands on, the more the program will be able at making predictions. These thousands of people which data are used by unknown entities for research reasons are usually unaware that they are giving these kinds of data. As the communication between people and the gaming company is almost inexistent regarding this data collection, many find basing the development of AI algorithms this way as unethical.
This explains why, according to a study by IBM’s Institute for Business Value, 60 percent of companies willing to use Artificial Intelligence fear liability issues even though 82 percent of all companies are willing to use AI. Most companies recognize the impressive benefits of implementing AI in their daily work, such as drive revenues, improve customer service, lower costs, and manage risk. Nevertheless, they undoubtedly fear issues regarding the ethical aspects of developing and using AI, including where the data comes to be able to develop it. (Rossi, 2018)
Researcher and professor at Les Gobelins, Nicolas Esposito, stated that the actual issue on data collection by video game companies regards the nature of the data rather than the quantity. Indeed, he highlighted the potential risks that could arise in the event that gaming companies face hacking (“banking data”) and if such data gets manipulated for political means (just like Cambridge Analytica). “The latter is clearly linked to the personal characteristics which can be mined in certain games” he explained.
Credits: CleanPNG
The main difference between video game companies collecting data from us and social media companies doing the same thing is that people are mostly aware of the kind of personal information we give out. Indeed, social medias want to know who we are through our given information we are mostly aware of and we know that through countless controversies and mediatized legal cases. On the other hand, players are not aware of this goal from video game companies, which makes it a key privacy difference. It must be stressed that social medias are far from being saints in this matter. (Togelius, 2019) Nevertheless, awareness remains an important factor to take into account when looking at privacy issues, which video game companies have not succeeded when collecting data from players to develop new Artificial Intelligence algorithms.