Google, Apple, Microsoft, Amazon, Facebook, and other tech giants are all building their own "bots" that utilize artificial intelligence and machine learning. There are a bevy of other startups doing the same, developing conversational technology that can help humans with everyday tasks. But one small company out of Palo Alto, Calif. is building a virtual assistant that it believes is differentiated and special -- and some top investors seem to agree. Ozlo is a new Silicon Valley startup that last month reeled in 14 million investment round from Greylock and AME Cloud Ventures, a fund started by Yahoo co-founder Jerry Yang. It also just opened its first remote office in Seattle and plans to move into a permanent location next month, with room for up to 25 employees.

That's the conviction held by Russian entrepreneurs Artem Kukharenko and Alexander Kabakov, whose startup, NTechLab, recently launched a facial recognition app that nearly obliterates the concept of anonymity. Called FindFace, the app has remained exclusive to Russia since going live earlier this year. Soon, though, Kuhkarenko and Kabakov are introducing a cloud-based platform that makes their frighteningly accurate algorithm available to everyone, the pair said in a Skype interview in May. In a practical sense, what this means is that none of us is safe from an always-probing public eye. "In 10 or 20 years, there won't be a place on the earth [where] ... nobody [can] see you," Kabakov said over Skype.

Experts on artificial intelligence are following up on the first White House workshop on artificial intelligence, presented last month in Seattle, with a session that addresses a central question about the technology: What good can it do for humanity? Whenever folks talk about AI, the discussion usually turns to the dark side. Will machines surpass us, even rule over us? Researchers point out that although computers can be programmed to outdo unassisted humans in specialized tasks, such as playing the game of Go, artificial general intelligence still lags far behind human capabilities. But if there's even a minuscule risk that robot overlords will prevail, as claimed by luminaries such as Stephen Hawking and Elon Musk, why take the chance?

Alexa, Siri, Cortana and OK Google are all digital celebrities. But a new study has found that they are only popular behind closed doors. Research shows that 51 percent of consumers use voice assistants in the car, but only 6 percent will activate them in public. Alexa, Siri, Cortana and OK Google are all digital celebrities. But a new study has found that they are only popular behind closed doors.

Deep learning and artificial intelligence are set to be the next big innovations that change the way that people run their businesses. At Shopify's Tech @ Scale in Toronto -- an event that included a day full of talks exploring how developers and engineers could stay agile in a world of fast-moving technology -- one talk included expert insights on exactly how these emerging technologies would impact both businesses and their customers. "The fabric of the internet is not ready to scale for what's coming. If you think of the components we built our service on top of, the one part that is not ready to scale is security because it is the most centralized." Chris Matys, chief analytics officer at Georgian Partners, moderated a panel featuring Damian McCabe, VP of Engineering at Connected Lab; Hesham Fahmy, VP of Technology at ecobee; Inmar Givoni VP of Big Data at Kobo; and Karl Martin, co-founder and CTO of Nymi.

The White House announced a series of public workshops on artificial intelligence (AI) and the creation of an interagency working group to learn more about the benefits and risks of artificial intelligence. The second workshop Artificial Intelligence for Social Good will take place on June 7 at The Willard Intercontinental Hotel, in Washington D.C. From the website, "there has been a dramatically increasing interest in Artificial Intelligence (AI) in recent years. AI has been successfully applied to societal challenge problems, and it has a great potential to provide tremendous social good in the future. In this workshop, we will discuss the successful deployments and the potential use of AI in various topics that are essential for social good, including but not limited to urban computing, health, environmental sustainability, and public welfare." The two keynote talks will be from Lynn Overmann, Senior Policy Advisor at the White House Office of Science and Technology Policy and Eric Horvitz, former CCC Council Member and Technical Fellow & Managing Director at Microsoft Research.

In the study of human learning, there is broad evidence that our ability to retain information improves with repeated exposure and decays with delay since last exposure. This plays a crucial role in the design of educational software, leading to a trade-off between teaching new material and reviewing what has already been taught. A common way to balance this trade-off is spaced repetition, which uses periodic review of content to improve long-term retention. Though spaced repetition is widely used in practice, e.g., in electronic flashcard software, there is little formal understanding of the design of these systems. Our paper addresses this gap in three ways. First, we mine log data from spaced repetition software to establish the functional dependence of retention on reinforcement and delay. Second, we use this memory model to develop a stochastic model for spaced repetition systems. We propose a queueing network model of the Leitner system for reviewing flashcards, along with a heuristic approximation that admits a tractable optimization problem for review scheduling. Finally, we empirically evaluate our queueing model through a Mechanical Turk experiment, verifying a key qualitative prediction of our model: the existence of a sharp phase transition in learning outcomes upon increasing the rate of new item introductions.

The Google DeepMind challenge match in March 2016 was a historic achievement for computer Go development. This article discusses the development of computational intelligence (CI) and its relative strength in comparison with human intelligence for the game of Go. We first summarize the milestones achieved for computer Go from 1998 to 2016. Then, the computer Go programs that have participated in previous IEEE CIS competitions as well as methods and techniques used in AlphaGo are briefly introduced. Commentaries from three high-level professional Go players on the five AlphaGo versus Lee Sedol games are also included. We conclude that AlphaGo beating Lee Sedol is a huge achievement in artificial intelligence (AI) based largely on CI methods. In the future, powerful computer Go programs such as AlphaGo are expected to be instrumental in promoting Go education and AI real-world applications.

Continual Learning in artificial neural networks suffers from interference and forgetting when different tasks are learned sequentially. This paper introduces the Active Long Term Memory Networks (A-LTM), a model of sequential multi-task deep learning that is able to maintain previously learned association between sensory input and behavioral output while acquiring knew knowledge. A-LTM exploits the non-convex nature of deep neural networks and actively maintains knowledge of previously learned, inactive tasks using a distillation loss. Distortions of the learned input-output map are penalized but hidden layers are free to transverse towards new local optima that are more favorable for the multi-task objective. We re-frame the McClelland's seminal Hippocampal theory with respect to Catastrophic Inference (CI) behavior exhibited by modern deep architectures trained with back-propagation and inhomogeneous sampling of latent factors across epochs. We present empirical results of non-trivial CI during continual learning in Deep Linear Networks trained on the same task, in Convolutional Neural Networks when the task shifts from predicting semantic to graphical factors and during domain adaptation from simple to complex environments. We present results of the A-LTM model's ability to maintain viewpoint recognition learned in the highly controlled iLab-20M dataset with 10 object categories and 88 camera viewpoints, while adapting to the unstructured domain of Imagenet with 1,000 object categories.

For complex, high-dimensional Markov Decision Processes (MDPs), it may be necessary to represent the policy with function approximation. A problem is misspecified whenever, the representation cannot express any policy with acceptable performance. We introduce IHOMP : an approach for solving misspecified problems. IHOMP iteratively learns a set of context specialized options and combines these options to solve an otherwise misspecified problem. Our main contribution is proving that IHOMP enjoys theoretical convergence guarantees. In addition, we extend IHOMP to exploit Option Interruption (OI) enabling it to decide where the learned options can be reused. Our experiments demonstrate that IHOMP can find near-optimal solutions to otherwise misspecified problems and that OI can further improve the solutions.