Greater compute power and power efficiency has made deep learning algorithms ubiquitous in our world. Deep learning has found its way into self driving cars, convenience stores and hospitals. Yet the fight for top talent in the space remains fierce and is a bottleneck for reaching new industries and solving tough challenges. To complement Udacity's previous AI courses, the online education startup is partnering with YouTube star Siraj Raval for a new deep learning nanodegree foundation program that will be co-taught with Udacity's Mat Leonard. Foundation Programs are going to be a major focus for Udacity in the coming year.

Microsoft announced this morning that it has acquired Maluuba, a Toronto startup focused on using deep learning for natural language processing. Deep learning is an approach to artificial intelligence currently in vogue that has driven incredible gains in the field over the last five years. As Microsoft wrote in the blog post announcing the purchase, "We've recently set new milestones for speech and image recognition using deep learning techniques, and with this acquisition we are, as Wayne Gretzky would say, skating to where the puck will be next -- machine reading and writing." The Verge covered Maluuba in the summer of 2016, when the startup shared the results of an AI system that could read and comprehend text with near human capability, outperforming similar systems shown off by Google and Facebook. Along with acquiring the company, Microsoft has also established closer ties with Yoshua Bengio, a pioneer in the field of deep learning who served as an advisor to Maluuba, and will now become and advisor to Microsoft's AI division.

Advances in artificial intelligence are changing the way automakers and their suppliers think about autonomous technology. Boosts in brainpower are not only accelerating the timeframe to bring self-driving technology to the marketplace, they're also broadening the scope of companies' ambition. "We're not just talking about autonomous cars," said Stefan Sommer, CEO of ZF Group. "We are talking about autonomous everything." At the CES technology show in Las Vegas, Sommer unveiled his company's newest product, an electronic control unit that contains artificial-intelligence software tailored for self-driving vehicles, including not only cars but also trains, buses, forklifts, trucks, tractors, and mining equipment.

On Tuesday, the White House released a chilling report on AI and the economy. It began by positing that "it is to be expected that machines will continue to reach and exceed human performance on more and more tasks," and it warned of massive job losses. Yet to counter this threat, the government makes a recommendation that may sound absurd: we have to increase investment in AI. The risk to productivity and the US's competitive advantage is too high to do anything but double down on it. This approach not only makes sense, but also is the only approach that makes sense.

We describe DyNet, a toolkit for implementing neural network models based on dynamic declaration of network structure. In the static declaration strategy that is used in toolkits like Theano, CNTK, and TensorFlow, the user first defines a computation graph (a symbolic representation of the computation), and then examples are fed into an engine that executes this computation and computes its derivatives. In DyNet's dynamic declaration strategy, computation graph construction is mostly transparent, being implicitly constructed by executing procedural code that computes the network outputs, and the user is free to use different network structures for each input. Dynamic declaration thus facilitates the implementation of more complicated network architectures, and DyNet is specifically designed to allow users to implement their models in a way that is idiomatic in their preferred programming language (C++ or Python). One challenge with dynamic declaration is that because the symbolic computation graph is defined anew for every training example, its construction must have low overhead. To achieve this, DyNet has an optimized C++ backend and lightweight graph representation. Experiments show that DyNet's speeds are faster than or comparable with static declaration toolkits, and significantly faster than Chainer, another dynamic declaration toolkit. DyNet is released open-source under the Apache 2.0 license and available at http://github.com/clab/dynet.

The promise of compressive sensing (CS) has been offset by two significant challenges. First, real-world data is not exactly sparse in a fixed basis. Second, current high-performance recovery algorithms are slow to converge, which limits CS to either non-real-time applications or scenarios where massive back-end computing is available. In this paper, we attack both of these challenges head-on by developing a new signal recovery framework we call {\em DeepInverse} that learns the inverse transformation from measurement vectors to signals using a {\em deep convolutional network}. When trained on a set of representative images, the network learns both a representation for the signals (addressing challenge one) and an inverse map approximating a greedy or convex recovery algorithm (addressing challenge two). Our experiments indicate that the DeepInverse network closely approximates the solution produced by state-of-the-art CS recovery algorithms yet is hundreds of times faster in run time. The tradeoff for the ultrafast run time is a computationally intensive, off-line training procedure typical to deep networks. However, the training needs to be completed only once, which makes the approach attractive for a host of sparse recovery problems.

We are now witnessing the increasing availability of event stream data, i.e., a sequence of events with each event typically being denoted by the time it occurs and its mark information (e.g., event type). A fundamental problem is to model and predict such kind of marked temporal dynamics, i.e., when the next event will take place and what its mark will be. Existing methods either predict only the mark or the time of the next event, or predict both of them, yet separately. Indeed, in marked temporal dynamics, the time and the mark of the next event are highly dependent on each other, requiring a method that could simultaneously predict both of them. To tackle this problem, in this paper, we propose to model marked temporal dynamics by using a mark-specific intensity function to explicitly capture the dependency between the mark and the time of the next event. Extensive experiments on two datasets demonstrate that the proposed method outperforms state-of-the-art methods at predicting marked temporal dynamics.

Deep learning promises to do more than just reshape city streets, thanks to autonomous vehicles. It can help put expensive infrastructure where it's needed most. It can automate the generation of zoning laws that ensure more liveable, walkable cities. And it poses new challenges, too. In Episode 5 of our AI Podcast, we spoke to two leading advocates for smarter, more liveable cities -- Lynn Richards, head of the Congress for New Urbanism, and Charles Marohn, head of Strong Towns -- to learn more about what AI will mean for the places where we live and work. To hear the whole conversation, tune into this week's AI Podcast.

As of last month Alphabet Inc.'s AI division, Google DeepMind, has open-sourced their new machine learning platform DeepMind Lab. Artificial Intelligence is the technology of the moment, constantly debated and attracting massive attention from investors. Despite warnings from influential figures including Professor Stephen Hawking, Google's decision to open up their software to other developers is part of a mass movement to advance the capabilities of AI. Facebook open sourced its own deep learning software last year, and Elon Musk's non-profit organisation OpenAI recently released Universe, an open software platform that can be used to train AI systems. So, why have Google, OpenAI and others made these platforms public, and how will this affect the adoption of Artificial Intelligence and machine learning as a whole?

In its first aquisition of calendar 2017, Microsoft announced plans to buy Montreal-based deep-learning startup Maluuba for an undisclosed amount. Maluuba has done work in natural-language understanding and reinforcement learning. Harry Shum, Microsoft executive vice president of the company's Artificial Intelligence and Research Group, explained a potential scenario where Maluuba's technology could help this way: While privacy and regulation will slow the pace of adoption, AI will bring some profound changes to healthcare. Yoshua Bengio, an advisor to Maluuba and head of the Montreal Institute for Learning Algorithms, also will be advising Microsoft and interacting directly with Shum as part of the deal. Microsoft created the combined Artificial Intelligence and Research Group last September concurrently with the departure of executive vice president Qi Lu, who previously led the combined Office and Bing organizations.