How can we design good goals for arbitrarily intelligent agents? Reinforcement learning (RL) is a natural approach. Unfortunately, RL does not work well for generally intelligent agents, as RL agents are incentivised to shortcut the reward sensor for maximum reward – the so-called wireheading problem. In this paper we suggest an alternative to RL called value reinforcement learning (VRL). In VR...

When talking about unfair competition, business ethics has to be involved, which means to judge whether the competition is fair or not is very much ethical debate. And its connotation is also evolving along with the development of social economy, so in this sense, defining “fair competition” or “unfair competition” is very difficult. Therefore, whether it is related to international conventions...

The notion of ‘wireheading’, or direct reward centre stimulation of the brain, is a wellknown concept in neuroscience. In this paper, we examine the corresponding issue of reward (utility) function integrity in artificially intelligent machines. We survey the relevant literature and propose a number of potential solutions to ensure the integrity of our artificial assistants. Overall, we conclud...

When thinking about the future, people tend to envision dramatically advanced technology, architecture that resembles something from another planet, and social norms that have shifted in radical ways—all of it vivid, dynamic, and exciting. What people don’t realize is that the story they tell themselves about the future actually has rules. The mind applies these rules to ensure that people arri...

In my thesis, I examine the discursive space of the detective fiction genre following Kasai Kiyoshi’s periodization in his two-volume seminal work Tantei shōsetsuron (The Theory of Detective Fiction, 1998). I investigate how Japanese detective fiction has developed in relation to Japan’s modernization, industrialization, nationalism, and globalization, specifically in the 1920s-30s, the 1950s-6...

The study presented in this paper deals with the modelling of extreme emotions occurring in abnormal situations. The aimed application is civil safety and surveillance in the public places in particular. A corpus of fiction (SAFE Corpus) is selected illustrating rich and varied contexts with the presence of extreme emotions, mainly fear. An annotation strategy adapted to the application is then...