We propose an approach to build a neural machine translation system with no supervised resources (i.e., no parallel corpora) using multimodal embedded representation over texts and images. Based on the assumption that text documents are often likely to be described with other multimedia information (e.g., images) somewhat related to the content, we try to indirectly estimate the relevance between two languages. Using multimedia as the "pivot", we project all modalities into one common hidden space where samples belonging to similar semantic concepts should come close to each other, whatever the observed space of each sample is. This modality-agnostic representation is the key to bridging the gap between different modalities. Putting a decoder on top of it, our network can flexibly draw the outputs from any input modality. Notably, in the testing phase, we need only source language texts as the input for translation. In experiments, we tested our method on two benchmarks to show that it can achieve reasonable translation performance. We compared and investigated several possible implementations and found that an end-to-end model that simultaneously optimized both rank loss in multimodal encoders and cross-entropy loss in decoders performed the best.

Given that in practice training data is scarce for all but a small set of problems, a core question is how to incorporate prior knowledge into a model. In this paper, we consider the case of prior procedural knowledge for neural networks, such as knowing how a program should traverse a sequence, but not what local actions should be performed at each step. To this end, we present an end-to-end differentiable interpreter for the programming language Forth which enables programmers to write program sketches with slots that can be filled with behaviour trained from program input-output data. We can optimise this behaviour directly through gradient descent techniques on user-specified objectives, and also integrate the program into any larger neural computation graph. We show empirically that our interpreter is able to effectively leverage different levels of prior program structure and learn complex behaviours such as sequence sorting and addition. When connected to outputs of an LSTM and trained jointly, our interpreter achieves state-of-the-art accuracy for end-to-end reasoning about quantities expressed in natural language stories.

Could you ever imagine yourself in a heated argument with a chatbot? Like, really passionate, deeply reasoned position-taking--argument, counterargument, countercounterargument, countercountercounterargument. Could you imagine a chatbot convincing a jury of a defendant's guilt? And if you could imagine it, what would that mean? Questions like these are at the core of a paper published recently in AI Matters by Samira Shaikh, a computer science-slash-psychology researcher at the University of North Carolina-Charlotte.

The Library of Congress has made 25 digital catalog records available for anyone at no charge. Imagine you wanted to find books or journal articles on a particular subject. Or find manuscripts by a particular author. Or locate serials, music or maps. You would use a library catalog that includes facts – like title, author, publication date, subject headings and genre.

Robot Academy is an online platform that provides free-to-use undergraduate-level learning resources for robotics and robotic vision. The content was developed for two 6-week Massively Open Online Courses (MOOCs) that Corke taught in 2015 and 2016. This content is now available as individual lessons (over 200 videos, each less than 10 minutes long) or in masterclasses (collections of videos, around 1 hour in duration, previously a MOOC lecture). Unlike a MOOC, all lessons are available all the time. While the content is typically designed for undergraduate-level students, around 20% of the lessons require no more than general knowledge.

Google's AI boutique, DeepMind, known for dispelling human delusions of intellectual superiority by soundly beating the world's top Go players with computer code, has found that instilling its software agents with something like imagination helps them learn better. In two papers published this week – "Imagination-Augmented Agents for Deep Reinforcement Learning" and "Learning model-based planning from scratch" – the AI biz's brain boffins, based in Britain, describe novel techniques for improving deep reinforcement learning through what can generously be described as imaginative planning. Reinforcement learning is a form of machine learning. It involves a software agent that learns by interacting with a specific environment, usually through trial and error. Deep learning is a form of machine that involves algorithms inspired by the human brain, called neural networks.

Gaining competitive advantage and improving business processes are among the top goals of digital transformation strategies, according to the report, "The Digital Workplace Report: Transforming Your Business," which is based on a survey of 850 organizations in 15 countries. About three quarters of the organizations surveyed (64 percent) use analytics to improve customer services, and 58 percent use analytics to benchmark their workplace technologies. "The digital workplace is transforming how employees collaborate, how customers are supported, and ultimately how enterprises do business," the report said. "A successful digital transformation strategy also must have clear and measurable goals from the start and must receive continued support throughout its implementation from heads of business units across the enterprise," the report said.

System makers Fujitsu and Huawei Technologies reportedly are both planning to develop processors optimized for artificial intelligence workloads, moves that will put them into competition with the likes of Intel, Google, Nvidia and Advanced Micro Devices. Tech vendors are pushing hard to bring artificial intelligence (AI) and deep learning capabilities into their portfolios to meet the growing demand generated by a broad range of workloads, from data analytics to self-driving vehicles. Microsoft, Google, IBM and others are creating AI business units and building out products and services that can leverage the technologies. Chip makers also are making the move. Intel last week unveiled its latest generation Xeon server chips that, among other improvements, deliver 2.2 times the performance for deep learning training and inference tasks than their predecessors.

SoftBank and General Motors are among a group of companies that invested $159 million into U.S. driverless car start-up Nauto on Wednesday. Palo Alto, CA-based Nauto makes cameras that sit inside of the car pointing at the driver, and outside of the vehicle looking at the road. They can track driver behavior in real time, and know if they are distracted. It uses computer vision, which is a form of artificial intelligence, to capture and process all of the data. All of this data and insight will help the development of self-driving cars.

Around 47% of the AI companies acquired since 2012 have had VC backing. Corporate giants like Google, IBM, Yahoo, Intel, Apple, and Salesforce are competing in the race to acquire private AI companies, with Ford, Samsung, GE, and Uber emerging as new entrants. Over 250 private companies using AI algorithms across different verticals have been acquired since 2012, with 37 acquisitions taking place in Q1'17 alone. That quarter also saw one of the largest M&A deals: Ford's acquisition of Argo AI for $1B. Baidu has been particularly aggressive in its AI acquisitions in 2017, with 3 M&A deals so far this year, including its acquisition of Amazon Alexa Fund-backed Kitt.ai this quarter.