China
Africa
Explore chapters and articles related to this topic
A Look at Top 35 Problems in the Computer Science Field for the Next Decade
Published in Durgesh Kumar Mishra, Nilanjan Dey, Bharat Singh Deora, Amit Joshi, ICT for Competitive Strategies, 2020
Deepti Goyal, Amit Kumar Tyagi
The collaboration and cooperation of skilled and learned agents from the different fields including economics and scientific perspective leads to the emergence of Algorithmic Game Theory. It basically provides an overview of the agents based on the choices they make supported by various incentives to keep them motivated and encouraged. This system includes human members who are self-motivated and keenly interested to work along with other agents with limited resources. Note that a Multi-Agent System (MAS or “self-organized system:) is a loosely held network of software agents/computerized system indulged in by various interacting intelligent agents who link with each other to tackle problems which are way ahead of the individual capacities or knowledge of each problem solver (http://igiglobal.com).
Using Decision Theory and Value Alignment to Integrate Analogue and Digital AI
Published in Maurizio Tinnirello, The Global Politics of Artificial Intelligence, 2022
Condorcet understood different sets of background conditions could suggest different decision procedures to aggregate votes to lead to better decisions. For example, in the case where a group of low knowledge and low independent thinking 3-year-old preschoolers and a high knowledge adult preschool teacher are deciding when the class should cross the road, it is easy to model the situation and show that if we want to maximise the probability of the class crossing the road safely, we should effectively allow the adult teacher to be a dictator when it comes to deciding when to cross the road.12 Unfortunately, for Condorcet, he did not have access to computers that would allow him to approximate good enough decision methods when background conditions were not simple corner case examples, like the ones he could easily calculate. Fortunately for us, we have access to computational resources and MASs, which allow us to get approximate answers to two questions: (1) given a particular set of background conditions, what are best or good enough voting methods by which to aggregate individual judgements into an aggregate group judgement? And (2) given the goal to use a particular group decision-making process, how do we need to change background conditions so that the particular group decision-making process produces good judgements? For example, with respect to the preschool class case, teacher dictatorship is the best decision rule given the background conditions, but let us suppose our goal is for each student to eventually learn how to cross the road using their own judgement. Given such a goal, it is clear that we must increase the knowledge and independence levels of the preschoolers, perhaps through institutions that teach kids how to cross roads. Again, this is just a corner case, but more complicated cases can be dealt with using MASs and algorithmic game theory.13
Artificial Intelligence in Purchasing: Facilitating Mechanism Design-based Negotiations
Published in Applied Artificial Intelligence, 2020
Ines Schulze-Horn, Sabrina Hueren, Paul Scheffler, Holger Schiele
Given the complexity of mechanism design-based negotiations, the idea emerged that AI might also facilitate this type of negotiation method. Previous research suggests that strong connections exist between game theory and AI (Tennenholtz 2002). Both research fields are concerned with decision theory and assume players to be rational (Elkind and Leyton-Brown 2010; Russell and Norvig 2010). Computer scientists began to study game theory in the 1990 s due to two major reasons: first, the computational properties of economic systems became too complex for practical use so that economics needed to collaborate with computer scientists, and second, the advances of the Internet required computer scientists to deal increasingly with settings in which players interact and thereby influence each other (Elkind and Leyton-Brown 2010; Parkes et al. 2010). Algorithmic game theory constitutes the research area resulting from the combination of game theory and AI (Nisan 2007). Three key research areas in algorithmic mechanism design concern game playing, social choice, and mechanism design (Elkind and Leyton-Brown 2010). With the help of AI methods, it is believed that also mechanism design-based negotiations could be supported by reducing the dependence on experts in this field of study, developing more sophisticated negotiation designs, shortening the preparation time for negotiations, and facilitating a large-scale application of the negotiation method. The ultimate result is higher cost savings for buying organizations.