10 Break-Out Sessions
[timetable id="9" column_title="0" filter_visible="1" filter_multiple="1" event_box_time="0"]
[timetable id="9" column_title="0" filter_visible="1" filter_multiple="1" event_box_time="0"]
Sophisticated algorithms and artificial intelligence are some of the most important and promising technological developments of our time, especially in a business context. Algorithms can improve the targeting of communication and offerings of companies and platforms, often even to the point of individualization.
AI is a branch of computer science that deals with the simulation of intelligent behavior in computers or the ability of a machine to imitate intelligent human behavior. In other words, it can not only support human work and decision- making processes but could eventually replace them altogether. Thus, AI is the first field that threatens to compete with humanity in a domain long considered unreachable for machines: making intelligent decisions.
Most people are already used to social networks applying algorithms to filter content for users, online shops track order history and user behavior to make product suggestions, and search engines tailor results to stored individual profiles. The basis for that is data, very often on a level of detail and in amounts that many people are not aware of, which is analyzed and used for predictions by algorithms that most people do not understand.
Many algorithms are aimed at influencing online search and shopping decisions (which ultimately also impacts offline behavior). They do not exert control through authoritarian power, do not impose prohibitions or laws, but they subtly create different realities (bubbles) and thus influence human decisions.
How do the Leaders of Tomorrow – many of whom can be expected to lead start-ups or join top management of companies in the future – see the issue of freedom in the context of this new technology? Do they consider algorithms that filter the content they see on the Internet more as a tool of convenience, more as a patronizing instrument, or both? This year’s Voices of the Leaders of Tomorrow Report – a collaboration of the Nuremberg Institute for Market Decisions and the St. Gallen Symposium – provides answers.
First and foremost, they see restrictions in their freedom of information and freedom of choice in algorithms that filter content, a kind of “algorithmic paternalism”. Two thirds of them agree to corresponding statements, just 20% disagree. In contrast, convenience aspects such as decision support by preselection and customized suggestions polarize clearly: Half of the respondents seem to distrust the quality of the algorithm-based preselection or do not consider them to be really helpful.
Concerning the intentional delegation of tasks to AI, the Leaders of Tomorrow were asked to recommend which of various tasks should be delegated fully or partially to AI by companies. The following response options were available: 1) Humans decide without input from AI, 2) AI develops a number of options and humans decide between these options, 3) humans develop a number of options and AI decides among them, or 4) AI decides without input from humans.
The variation of the answers is very high. The AI is granted least authority in matters of personnel and human resources. Especially the hiring decision itself should be made by only a person – at least according to about half of the Leaders of Tomorrow. However, more than 40% think that AI can preselect options. The power of influence by creating options seems to be perceived as smaller than by making the final decision. But as the development of options creates a shortlist and narrows the scope of possibilities, the perceived level of control exerted by making the final decision may be overestimated or – to put it bluntly – be only an illusion of control.
When it comes to the task of developing products, far more Leaders of Tomorrow suggest that AI should be granted influence. Just a quarter of the respondents would do this creative task without the support of artificial intelligence, while nearly 60% would use it for the development of options. So, creating the shortlist is the preferred type of AI assistance again.
The greatest extent of AI involvement is recommended for a third task, granting discounts or setting surcharges or customers. Every fifth participant even thinks that these decisions should be left entirely to AI. And similar percentages think that AI should select among human-developed options (34%), or that humans should select among AI-developed options (36%). Thus, 9 out of 10 respondents would rely on the AI for this task. Maybe the high level of overall agreement that AI should be involved in decision making in this case can be explained by the fact that AI is already widely – and successfully – applied in a similar task, namely for programmatic buying in online advertising.
Thus, it does not need much imagination to conceive of applying AI solutions to automatic pricing. The interesting questions for the future will be to see which decision tasks are best left to humans, which to AI, and in which ways the two can best cooperate.
Read the full Voices of the Leaders of Tomorrow Report here for all findings and detailed analysis.