There is a tendency to see artificial intelligence as the latest technology fad, either a buzzword that canny entrepreneurs exploit or the starting point for dystopian nightmares. Both are potentially accurate descriptions of much of the discussion surrounding AI, but both miss the most important point: AI is almost certain to become the most critical feature of the digital economy, assuming a role akin to electricity in the industrial revolution. If that prediction is correct -- and few disagree -- then mastery of AI and leadership in the field could determine the future economic and military balance of power. AI is shorthand for an amalgam of computer processes that permit machines to evaluate and learn about their environment on their own. It includes automated intelligence, assisted intelligence, augmented intelligence and autonomous intelligence.

To remain livable and economically competitive, rapidly growing cities need to embrace high-tech solutions to solve their many practical problems. However, how willing are citizens to sacrifice their privacy for the benefits of smart cities, and can government regulations keep up with new tech? This places a significant burden on vital infrastructure, such as transport, housing, energy supply, health care and waste management. Livability in the megacities of tomorrow will largely be determined by the smart solutions being developed today. The term "smart city" is popular among policymakers worldwide.

"Fail fast" is a well-known phrase in the startup scene. The spirit of failing fast is getting to market with a minimum viable product and then rapidly iterating toward success. Failing fast acknowledges that entrepreneurs are unlikely to design a successful end-state solution before testing it with real customers and real consequences. This is the "ready, fire, aim" approach. Or, if the blowback is big enough, it's the "ready, fire, pivot" approach.

Data can appear lifeless and dull on the surface – especially government data – but the thought of it should actually get you excited. Data is a very interesting and powerful thing. First off, data is exactly the stuff we bother to write down – and for good reason. But its potential far transcends functions such as tracking and bookkeeping: Data encodes great quantities of experience, and computers can learn from that experience to make everything work better. For example, take agriculture – and the federal studies that advance it.

A strong declaration from a historically antagonist foe should put chills in the hearts of Americans preparing themselves for the world ahead: Russian President Vladimir Putin says the nation that leads in AI will be the ruler of the world [1]" … The ruler of the world! "The development of artificial intelligence has increasingly become a national security concern in recent years. It is China and the US (not Russia), which are seen as the two frontrunners, with China recently announcing its ambition to become the global leader in AI research by 2030. Many analysts warn that America is in danger of falling behind, especially as the [current US] administration prepares to cut funding for basic science and technology research." Elon Musk, one of America's foremost technology advocates, predicts that countries seeking leadership (and domination) from artificial intelligence will be the basis for World War III[2].

To train an inference network jointly with a deep generative topic model, making it both scalable to big corpora and fast in out-of-sample prediction, we develop Weibull hybrid autoencoding inference (WHAI) for deep latent Dirichlet allocation, which infers posterior samples via a hybrid of stochastic-gradient MCMC and autoencoding variational Bayes. The generative network of WHAI has a hierarchy of gamma distributions, while the inference network of WHAI is a Weibull upward-downward variational autoencoder, which integrates a deterministic-upward deep neural network, and a stochastic-downward deep generative model based on a hierarchy of Weibull distributions. The Weibull distribution can be used to well approximate a gamma distribution with an analytic Kullback-Leibler divergence, and has a simple reparameterization via the uniform noise, which help efficiently compute the gradients of the evidence lower bound with respect to the parameters of the inference network. The effectiveness and efficiency of WHAI are illustrated with experiments on big corpora.

Computing the partition function $Z$ of a discrete graphical model is a fundamental inference challenge. Since this is computationally intractable, variational approximations are often used in practice. Recently, so-called gauge transformations were used to improve variational lower bounds on $Z$. In this paper, we propose a new gauge-variational approach, termed WMBE-G, which combines gauge transformations with the weighted mini-bucket elimination (WMBE) method. WMBE-G can provide both upper and lower bounds on $Z$, and is easier to optimize than the prior gauge-variational algorithm. We show that WMBE-G strictly improves the earlier WMBE approximation for symmetric models including Ising models with no magnetic field. Our experimental results demonstrate the effectiveness of WMBE-G even for generic, nonsymmetric models.

Fingerspitzengefühl: A German word used to describe the ability to maintain attention to detail in an ever-changing operational and tactical environment by maintaining real-time situational awareness. The term is synonymous with the English expression of "keeping one's finger on the pulse". The problem with fingerspitzengefühl traditionally, in addition to pronouncing it, has been it is hard for an individual to scale up. In a world of sensors, AI and mobile devices, having real-time situational awareness is far easier than ever before. In fact, today the challenge is not how to do it, but what to do with the massive volume of data that can be provided.

It always starts with porn. What first revealed the internet's power to distribute information? Porn has historically been a reliable canary in the coal mine, so the "deep fakes" video Vice found in late 2017 has lawmakers paying attention. Using free machine-learning platforms, people on Reddit superimposed the face of Wonder Woman's Gal Godot on a porn actress's body in a creepy, almost-convincing sex video. Researchers use "Real-time Face Capture" on Russian President Vladimir Putin.

US President Donald Trump has stepped up his war of words over trade tariffs, threatening to "apply a tax" on imports of cars from the European Union. Mr Trump said other countries had taken advantage of the US for years because of its "very stupid" trade deals. The trade wrangle began on Thursday when Mr Trump vowed to impose hefty tariffs on steel and aluminium imports. That brought a stiff response from trading partners and criticism from the IMF and WTO. EU trade chiefs have reportedly been considering slapping 25% tariffs on around $3.5bn (£2.5bn) of imports from the US, following Mr Trump's proposal of a 25% tariff on imported steel and 10% on aluminium.