Reinforcement learning in a multi agent system is difficult because these systems are inherently non-stationary in nature. In such a case, identifying the type of the opposite agent is crucial and can help us address this non-stationary environment. We have investigated if we can employ some probing policies which help us better identify the type of the other agent in the environment. We've made a simplifying assumption that the other agent has a stationary policy that our probing policy is trying to approximate. Our work extends Environmental Probing Interaction Policy framework to handle multi agent environments.

Value iteration networks (VINs) have been demonstrated to have a good generalization ability for reinforcement learning tasks across similar domains. However, based on our experiments, a policy learned by VINs still fail to generalize well on the domain whose action space and feature space are not identical to those in the domain where it is trained. In this paper, we propose a transfer learning approach on top of VINs, termed Transfer VINs (TVINs), such that a learned policy from a source domain can be generalized to a target domain with only limited training data, even if the source domain and the target domain have domain-specific actions and features. We empirically verify that our proposed TVINs outperform VINs when the source and the target domains have similar but not identical action and feature spaces. Furthermore, we show that the performance improvement is consistent across different environments, maze sizes, dataset sizes as well as different values of hyperparameters such as number of iteration and kernel size.

Massachusetts police have begun rolling out Boston Dynamics' futuristic breed of robotic dog in a first for law enforcement anywhere. According to new documents obtained by the American Civil Liberties Union, the'dog,' a four-legged robot named'Spot,' has been involved in multiple police'incidents' since being introduced in August, though the full details of those cases is not yet clear. WBUR, who originally reported Spot's introduction onto the force, said police used Spot as a'mobile remote observation device' that was designed to investigate suspicious packages or sniff out where a suspect could be hiding. State police in Massachusetts have deployed Spot on two known incidents, though it's unclear exactly what those tag-alongs entailed Spot, which comes equipped with a robotic arm for opening doors and a low-light camera can be operated autonomously or manually by remote. The robot has turned heads often throughout its development for successfully carrying out feats like opening doors with its robotic arm using using a mixture of artificial intelligence and computer vision.

Machine learning provides a powerful way to automate decision making, but the algorithms don't always get it right. When things go wrong, it's often the machine learning model that gets the blame. But more often than not, it's the data itself that's biased, not the algorithm or the model. That's been the experience of Cheryl Martin, Ph.D., who worked as an applied research scientist at the University of Texas, Austin and NASA for 14 years before joining the AI crowdsourcing outfit Alegion as its chief data scientist earlier this year. "You often hear that the algorithm is biased, or the machine learning is algorithmically biased," Martin tells Datanami.

Singapore is ramping up efforts to harness artificial intelligence (AI) across its economy through a national strategy that encompasses skills training and industry partnerships while promoting the responsible use of AI. Unveiling the national AI strategy at the Singapore Fintech Festival, Singapore's deputy prime minister, Heng Swee Keat, said AI is the country's next step in its smart nation initiative, noting that AI is already being used in several aspects of daily life. To bring the strategy to fruition, the government has identified five national AI projects to address key national challenges, and to deliver social and economic benefits to Singaporeans. These projects include intelligent freight planning in logistics, smart estates, chronic disease management in healthcare, personalised education and border security. Heng said the national AI strategy will be rolled out iteratively, to respond to the rapidly changing technology landscape and to tap new opportunities brought about by AI.

In an era of purpose-driven consumption, values--such as transparency, trust and humanness--are key drivers that unlock value in AI, new research from WP Engine finds. The firm's latest study, conducted by researchers at The University of London and Vanson Bourne, explores the present and near future of AI-driven human digital experiences on the web, and the often tenuous but also potentially rewarding relationship between consumers, brands and AI. According to IDC, worldwide spending on AI systems is forecast to reach $35.8 billion in 2019, an increase of 44 percent over the amount spent in 2018. Much of that growth will come from the application of AI online because there is a natural, evolutionary symbiosis between AI and the internet. However, it was a sudden burst of activity starting in 2013 that marks the beginning of what we might term the modern AI period, especially for digital and digital experiences, characterized predominantly by automated content creation, programmatic ad buying in 2014, and intelligent search.

A highly compact, low-energy device capable of switching the paths taken by light within photonic systems has been unveiled by physicists in the US and Switzerland. The new switch could provide a basis for artificial-intelligence (AI) systems that mimic human decisions, allowing for a diverse range of applications. "All-optical computers" use light instead of electronic signals to process information. In principle, they could be faster and much more energy efficient than conventional computers. However, it is proving very difficult to create compact and energy efficient photonic devices that can switch and process optical signals at high speeds. The new device was created by Chris Haffner and colleagues at the National Institute of Standards and Technology (NIST), ETH Zurich and the University of Maryland.

Facebook and McKinsey) based out of its offices in Amsterdam, London and Jakarta.

The update shares an announcement from the United States Patent and Trademark Office ("USPTO") and provides important details regarding a second Federal Register Notice on AI and innovation building upon the earlier request for comments focused on the impact AI poses for patent law and policy. The Ocean Tomo AI Report Update contains important details regarding the window of opportunity to respond to the notice as well as sample questions included in the notice. You can request a copy of the study update here. Ocean Tomo Industry Analyst Reports provide a comprehensive look at current industry trends and deal activity in several technology areas. As a financial advisor with a focus on technology and intellectual property (IP), Ocean Tomo has gained unique insights related to the intellectual property driving the development of a variety of technology areas.

Early December marks the kickoff of Amazon's AWS re:Invent conference in Las Vegas, and ahead of the festivities the tech giant has unveiled a slew of product enhancements. To this end, Amazon Translate, the company's cloud machine translation service that delivers language translation via API requests, today gained new languages and variants and expanded to new regions globally. By way of a refresher, Translate -- which debuted in preview in November 2017 ahead of general availability last April -- taps AI that aims to deliver more accurate and natural-sounding translation than statistical or rule-based approaches. It allows customers to define how brand names, character names, model names, and other unique terms get translated. When used in tandem with a natural language processing app, Translate also facilitates sentiment analysis.