Home Technology AI The Myth of AIs’ Predictive Power

The Myth of AIs’ Predictive Power


One of the most promising advantages of AI (artificial intelligence) seems to be its ability to predict the future preferences of consumers. Through means of data collection on past transaction or search histories, the ability of AI to develop a detailed digital profile for every individual so that individual’s future preferences can be estimated under specific circumstances seem to be a nightmare among many users. The fact that leading companies such as Amazon already started to use dynamic pricing models further increase concerns on personal data collection by machine learning algorithms for the purpose of price predicting the willingness of consumers to make specific purchases. Needless to say, such an ability of companies would mean to be able to persuade the individuals at the right time to make purchases. Given the fact that due to dynamic price modeling, users would be exposed to different prices on their company profiles, the FTC (Federal Trade Commission) in the U.S is rightly concerned that the use of big datasets in order to manipulate users to increase their own profits could do damage to consumer welfare. 

Assuming that user preferences can be predicted by means of technology would mean forgetting how information differs from knowledge. While information is searchable and collectible, not all information exists objectively. Although data, as a subset of information- includes particular observations of specific actions or choices that are visible to individuals, often times, these choices are colored by subjective data due to an individual interpretation of the information available to them. Although machine learning algorithms might excel at making predictions regarding objective data such subjective knowledge cannot be reflected upon by algorithms regardless of how advanced they might be. This is why knowledge is contextual ad it is developed on a subjective interpretation of information within a particular setting and time.

Given this contextual nature of knowledge, predictions provided by AI would also be restricted. In fact, this was foreseen a few decades ago by the famous economist Hayek when he made a distinction between pattern and point predictions. While the former one refers to generic trends within the system the latter one refers to the next action of a specific individual or component within the system. Although this underlying difference might seem to be a technological one, it is, in fact, an epistemic one. According to Lavoie- as he mentioned in National Economic Planning (1986); given the dispersed form of knowledge, it cannot be externalized by a single individual within the society; yet such an extraction would be necessary in order to make use of knowledge for economic decision-making.

Each individual interaction with the market necessitates a self-discovery about one’s own preferred price level or personal choices. Each moment throughout this interaction process provides a variety of both external and internal factors which lead to the organic emergence of the final decision. Ideally, the decision-maker should bear in mind that pattern predictions could not provide much information on this organic evolution process of individual preferences and therefore they should not be seen as a guarantee for future acts. As such, AI should only be conceptualized as a technology which could help reduce the cost of pattern predictions through means of data collection and data interpretation rather than enabling human-beings or machines to make specific point predictions.   

Making big claims about the potential of AI regarding its influence on shaping individual decisions would mean to underestimate the value of individual action when it comes to using these services. Regardless of the amount of data collected and the sophistication of the AI technology used, individual preferences are not something that can be predicted ahead of time as the individual choice patterns develop throughout time as a result of interactions among other individuals or factors in the environment.

Individual ability to make a choice does not only make one empowered, but also feel authentic. Human-beings are not automatons who carry a static price tag within their heads around. Human-beings update their choices on an ongoing basis based on contextual conditions. This fact has also been asserted by the famous economist Buchanan who mentioned that unless individuals get involved in a  mutual process of exchange they also could not know for sure what their choices would be. In other words, individual choices are constantly being shaped and can be considered as a moving target which would make it impossible for AI to be forecasted in advance.

The use of platforms which collect individual data does not necessarily mean the erosion of consumer welfare. Both overestimating the prediction capabilities of AI and underestimating the entrepreneurial capabilities of the human actors within the market might result in reducing the benefits of AI for the individuals. What should be done is to provide space for both individuals and enterprises so that they can collaboratively determine the tradeoff between personalized services and privacy issues.

Previous articleLightning torch initiative – spreading the word about scalable bitcoin micro payments
Next articleWho heals the healer when depression strikes?
Ayse Kok
Ayse completed her masters and doctorate degrees at both University of Oxford (UK) and University of Cambridge (UK). She participated in various projects in partnership with international organizations such as UN, NATO, and the EU. She also served as an adjunct faculty member at Bosphorus University in her home town Turkey. Furthermore, she is the editor of several international journals, including those for Springer, Wiley and Elsevier Science. She attended various international conferences as a speaker and published over 100 articles in both peer-reviewed journals and academic books. Having published 3 books in the field of technology & policy, Ayse is a member of the IEEE Communications Society, member of the IEEE Technical Committee on Security & Privacy, member of the IEEE IoT Community and member of the IEEE Cybersecurity Community. She also acts as a policy analyst for Global Foundation for Cyber Studies and Research. Currently, she lives with her family in Silicon Valley where she worked as a researcher for companies like Facebook and Google.


Please enter your comment!
Please enter your name here