Open your Facebook feed, a newspaper or turn on the news and you'll likely see something about the dangers of machine learning, the increasing amount of fake news or even the dangers of AI on our privacy. Yet, these technologies are continuing to develop and thanks to new developments in automation and machine deception – they will continue to shape the use of AI over the coming year. While full automation might still be a way off, there are many workflows and tasks that lend themselves to partial automation. In fact, McKinsey estimates that "fewer than 5 per cent of occupations can be entirely automated using current technology. However, about 60 per cent of occupations could have 30 per cent or more of their constituent activities automated."

A model of music needs to have the ability to recall past details and have a clear, coherent understanding of musical structure. Detailed in the paper is a deep reinforcement learning architecture that predicts and generates polyphonic music aligned with musical rules. The probabilistic model presented is a Bi-axial LSTM trained with a pseudo-kernel reminiscent of a convolutional kernel. To encourage exploration and impose greater global coherence on the generated music, a deep reinforcement learning approach DQN is adopted. When analyzed quantitatively and qualitatively, this approach performs well in composing polyphonic music.

I'm looking for an implementation of a Regression Tree from scratch and have only been able to find classification trees.

If AI is to become a meaningful facet of society, identifiable and understandable by consumers, its value must be articulated. And for that to happen, designers of AI-driven experiences must make the invisible visible; they have to give AI a good, old-fashioned brand identity. A skeptic might wonder why AI needs branding in the first place. If it's meant to silently toil away in the background of our lives, why does it need to announce itself? Why give consumers yet another thing to think about?

'We' would have to be the countries doing AI development. There have been successful treaties that affect what the counties do. Montreal Protocol is the most famous, but the Outer Space Treaty is probably a better model. Probably most important is that treaty get produced and ratified before someone fields a systems, because they will not want to give it up. While countries do not know if they are ahead or behind in fielding such a system, they have a motivation to do the treaty, so they don't lose.

With the aim of creating virtual cloth deformations more similar to real world clothing, we propose a new computational framework that recasts three dimensional cloth deformation as an RGB image in a two dimensional pattern space. Then a three dimensional animation of cloth is equivalent to a sequence of two dimensional RGB images, which in turn are driven/choreographed via animation parameters such as joint angles. This allows us to leverage popular CNNs to learn cloth deformations in image space. The two dimensional cloth pixels are extended into the real world via standard body skinning techniques, after which the RGB values are interpreted as texture offsets and displacement maps. Notably, we illustrate that our approach does not require accurate unclothed body shapes or robust skinning techniques. Additionally, we discuss how standard image based techniques such as image partitioning for higher resolution, GANs for merging partitioned image regions back together, etc., can readily be incorporated into our framework.

The increasing role of recommender systems in many aspects of society makes it essential to consider how such systems may impact social good. Various modifications to recommendation algorithms have been proposed to improve their performance for specific socially relevant measures. However, previous proposals are often not easily adapted to different measures, and they generally require the ability to modify either existing system inputs, the system's algorithm, or the system's outputs. As an alternative, in this paper we introduce the idea of improving the social desirability of recommender system outputs by adding more data to the input, an approach we view as as providing `antidote' data to the system. We formalize the antidote data problem, and develop optimization-based solutions. We take as our model system the matrix factorization approach to recommendation, and we propose a set of measures to capture the polarization or fairness of recommendations. We then show how to generate antidote data for each measure, pointing out a number of computational efficiencies, and discuss the impact on overall system accuracy. Our experiments show that a modest budget for antidote data can lead to significant improvements in the polarization or fairness of recommendations.

The amount of dialogue history to include in a conversational agent is often underestimated and/or set in an empirical and thus possibly naive way. This suggests that principled investigations into optimal context windows are urgently needed given that the amount of dialogue history and corresponding representations can play an important role in the overall performance of a conversational system. This paper studies the amount of history required by conversational agents for reliably predicting dialogue rewards. The task of dialogue reward prediction is chosen for investigating the effects of varying amounts of dialogue history and their impact on system performance. Experimental results using a dataset of 18K human-human dialogues report that lengthy dialogue histories of at least 10 sentences are preferred (25 sentences being the best in our experiments) over short ones, and that lengthy histories are useful for training dialogue reward predictors with strong positive correlations between target dialogue rewards and predicted ones.

Stanley Kubrick's 2001: A Space Odyssey will help launch the world's first super-high definition 8K television channel on Saturday. Japanese broadcaster NHK said it had asked Warner Bros to scan the original film negatives in 8K for its new channel. Super-high definition 8K pictures offer 16 times the resolution of HD TV. However, few people currently have the necessary television or equipment to receive the broadcasts. NHK says it has been developing 8K, which it calls super-hi vision, since 1995.

Amazon is pretty platform-agnostic when it comes to music. Its Echo speakers obviously work with Amazon Music as well as Spotify, Pandora, IHeartRadio and a number of other options. Starting on December 17th, Echo owners will have another option: Apple Music. The second-biggest streaming service in the US will work with Alexa, with users able to request songs, albums, artists, radio stations and playlists from Apple Music. While Amazon says that Apple Music will work with Echo speakers starting the week of December 17th, it doesn't say how fast the rollout will go, so Echo owners may have to be a little patient.