Search results

11/04/2019 - Recent advancements in Artificial Intelligence (AI) research result from the combination of deep neuronal networks and reinforcement learning. In the latter, agents are able to learn rewarding behaviours in unknown environments by an iterative trial-and-error behaviour update process. But this process is not yet fully understood. Reinforcement learning agents are a specific area of AI. As AI can have a big impact on society, a better understanding AI systems is crucial to assess potential challenges and risks. Already today, AI is employed to steer cars, manage production lines, or even draft texts. A team of scientists from the Potsdam Institute for Climate Impact Research has developed a new method to investigate those algorithms using insights from statistical physics. Published in the journal Physical Reviews E, their insights can help to improve the design of large-scale AI reinforcement learning systems.

02/26/2018 - Even countries that tend to act in an egoistic way in the long run have an incentive to participate in international climate stabilization pathways and couple CO2 pricing systems, a new game-theoretical study shows. Yet they might only do this if pioneer coalitions for pricing greenhouse gas emissions make the first steps. If this is the case, the egoistic countries temporarily enjoy the benefits of avoided climate change without paying for it, but in the longer term can join the pioneers and link to their already established models of CO2 pricing. Forming larger and larger coalitions always reaps additional benefits of avoided damages from climate change. These benefits, even though unequally distributed across the coalition members, can be distributed via financial transfers. This makes it attractive to join, even for egoistic countries.