The unintended consequences of implementing AI for personal use

Published on in ,

I truly believe that machine learning and artificial intelligence will impact every realm of society. Generally, I am quite excited about seeing more ML/AI creep into my every day life, whether that’s Siri, Netflix recommendations, or chat bots in Messenger. I am convinced that the ubiquitous use of AI can enhance the user experience in consumer products. However, as with the emergence of any new technology, there are often unintended consequences. For that reason, I co-authored a report that addressed the research question:

What are the unintended consequences of implementing artificial intelligence for personal use?

The report was published by the UK’s Digital Catapult and you can get a free copy here (Creative Commons BY-NC-4.0).

The report makes three significant contributions to the field of AI research

  1. It provides a definition for Consumer AI as the commercial development of intelligent agents for personal use.

  2. The application of agency theory to AI.

  3. A novel regulatory approach based on Agent Decoupling.

To borrow from the report’s Executive Summary…

The field of artificial intelligence (AI) has been attracting unprecedented interest from business, governments, and academia (Lunden, 2016; Simonite, 2017; Smith, 2016; Agrawal et al., 2016; The Economist, 2016d). This trend has been fuelled primarily by: the emergence of Big Data, extensive research in Information Technology, and ever-growing compute power (Kelly, 2014). It is no surprise that, as a result, AI applications are no longer restricted to computer science labs but find their way into marketplaces. Despite that, there is no consistent source that businesses and policymakers can refer to in order to build their understanding of the nature and effects of these consumer-facing AI applications. This report fills this gap by focusing on the commercial developments of intelligent agents for personal use, defined by the authors as Consumer AI.

Based on comprehensive primary research, this original analysis is structured around the notion of unintended consequences. Thus, it focuses specifically on the benefits and drawbacks beyond the intrinsic motivations of Consumer AI implementers. Key findings were derived from a combination of semi-structured interviews with 30 experts across business, government, and academia regarding the topic of AI, and a survey of 300 consumers based in the United Kingdom. Secondary sources and case studies identified in the literature provided context and further evidence. As a result, this report identifies ten interest areas where unintended consequences of Consumer AI are evident. As such, it contributes to the literature by identifying the interest areas in which the consequences are most applicable to Consumer AI, and maps the effect on consumers as direct or indirect.

Therefore, by identifying Consumer AI as a newly-emerging focus area in the field, and conducting a systematic analysis of primary research in ten interest areas, this report aims to stimulate conversation among implementers and regulators of AI-based consumer products.