Language is often considered a key aspect of human thinking, providing us with exceptional abilities to generalize, explore, plan, replan, and adapt to new situations. However, Reinforcement Learning (RL) agents are far from human-level performance in any of these abilities. We hypothesize one reason for such cognitive deficiencies is that they lack the benefits of thinking in language and that we can improve AI agents by training them to $\textit{think like humans do}$. We introduce a novel Imitation Learning framework, Thought Cloning, where the idea is to not just clone the behaviors of human demonstrators, $\textit{but also the thoughts humans have as they perform these behaviors}$. While we expect Thought Cloning to truly shine at scale on internet-sized datasets (e.g.

Luke says many fans "gave up hope" that a sequel was on the way, given the six-year gap between the two titles. But when the new game was revealed, he says, there was "this incredible reception and everyone went absolutely crazy". It also reignited an online discourse that had erupted around the release of the original KCD. Daniel Vávra, the co-founder and creative director of Warhorse, is a regular poster on social media and is quick to answer critics. He defended the first KCD, when it was criticised for its lack of diversity, as being historically accurate to the time and location of its setting, although there is not universal agreement about this.

Language is often considered a key aspect of human thinking, providing us with exceptional abilities to generalize, explore, plan, replan, and adapt to new situations. However, Reinforcement Learning (RL) agents are far from human-level performance in any of these abilities. We hypothesize one reason for such cognitive deficiencies is that they lack the benefits of thinking in language and that we can improve AI agents by training them to \textit{think like humans do} . We introduce a novel Imitation Learning framework, Thought Cloning, where the idea is to not just clone the behaviors of human demonstrators, \textit{but also the thoughts humans have as they perform these behaviors} . While we expect Thought Cloning to truly shine at scale on internet-sized datasets (e.g.

Model based reinforcement learning (MBRL) uses an imperfect model of the world to imagine trajectories of future states and plan the best actions to maximize a reward function. These trajectories are imperfect and MBRL attempts to overcome this by relying on model predictive control (MPC) to continuously re-imagine trajectories from scratch. Such re-generation of imagined trajectories carries the major computational cost and increasing complexity in tasks with longer receding horizon. This paper aims to investigate how far in the future the imagined trajectories can be relied upon while still maintaining acceptable reward. Firstly, an error analysis is presented for systematic skipping recalculations for varying number of consecutive steps.% in several challenging benchmark control tasks. Secondly, we propose two methods offering when to trust and act upon imagined trajectories, looking at recent errors with respect to expectations, or comparing the confidence in an action imagined against its execution. Thirdly, we evaluate the effects of acting upon imagination while training the model of the world. Results show that acting upon imagination can reduce calculations by at least 20% and up to 80%, depending on the environment, while retaining acceptable reward.

Sony Corp. on Wednesday showed off the latest tricks of its signature robotic dog, Aibo, including a new security capability of patrolling inside a house. Teaming up with security company Secom Co. Ltd., the Japanese electronics giant will roll out the new roaming security feature for free in a system software update scheduled for mid-February. Specifically, up to 10 names and faces can be registered on the My Aibo app, and the user's Aibo can then be programmed to walk around predetermined areas in a home or other structure at designated times to check whether any of the registered individuals are present. The user will subsequently receive a report from the Aibo. The patrolling task is the first feature that showcases the Aibo's ability to remember and recognize spaces, said Izumi Kawanishi, Sony's senior vice president in charge of the artificial intelligence robotics business.

Studies have shown we like happy faces because positive emotions in others immediately boosts our own mental state. But what are the emotional consequences of trying to seem happy in order to please others? Researchers have found that we can adjust facial expressions and body gestures without actually changing our emotional state – for example putting on a smile without being happy. People performing surface acting'put on a mask', which creates an unhealthy inner conflict between expressed and felt emotions, researchers found. In an article for The Conversation, Milda Perminiene, a senior lecturer in occupational psychology at the University of East London, shows that this can cause emotional exhaustion, strain and reduced job satisfaction.

Quentin Tarantino indulges in some of his worst impulses in this widescreen western, loading it up with violence and vulgarity to an almost nihilistic degree. Yet as tone-deaf and ugly as the film often is, it's also beautifully shot (by Robert Richardson) and masterfully acted (by an all-star cast that includes Kurt Russell, Samuel L. Jackson, Walton Goggins and an Oscar-nominated Jennifer Jason Leigh), with a stirring Oscar-winning score from Ennio Morricone. And it's always a pleasure to listen to Tarantino's dialogue, with its winding speeches and stories within stories. He bites off more than he can chew with this claustrophobic tale of post-Civil War animus, boiling over at a snowed-in Wyoming trading post, but while the movie is uneven, it's often thrilling. Buyer beware, though: While the film itself is certainly worthy, this first DVD and Blu-ray release contains the shorter, nonroadshow cut, with just a couple of featurettes.