Humans have always had a complicated relationship with new "technologies." From awe to fear, centuries ago, Plato even worried that writing would adversely affect people's memories. Modernity has had a particular curiosity regarding artificial intelligence (AI). From Terminator-style killer robots to emotive humanoids, the mention of AI brings to mind the many silver screen renderings of some future civilization. More likely than any of these, however, is the reality that AI will probably turn out to be another commonplace technology that, while novel at first, will end up integrated into our everyday lives.

An artificial intelligence "chatbot" from Taco Bell now lets you order a meal in a smartphone text exchange that might look something like this: TacoBot: Hello there, I'm your TacoBot, I can help you order a meal for you or your team. TacoBot: Sounds good... do you want to keep adding stuff? Brands like Taco Bell and tech companies large and small are betting that more and more people will start using this "conversational" way of interacting online instead of clicking through on-screen menus. If the trend catches on -- as firms like Facebook and Microsoft expect -- it could transform the digital landscape by allowing smartphone users to find information or make purchases with simple text messages, bypassing apps and search engines. Among the companies already developing or launching chatbots are the Wall Street Journal, CNN and retail giants Sephora and H&M.

Aliyun, the cloud computing unit of Alibaba Group Holding, on Wednesday officially introduced an artificial intelligence (AI) robot named Ai that's meant to compete with overseas AI players. The Ai system was developed on the basis of neural networks and social computing, according to a press release Alibaba sent to the Global Times on Wednesday. The Ai system is good at predicting and understanding human emotions, the company said. Neural networks are based on animals' central nervous systems, and they are used to mimic functions involving many inputs. Social computing can be defined as the intersection of social behavior and computing systems.

The threat from such machines is real enough for 100 states to come together and debate the matter of their ban for three consecutive years now. The use of autonomous machines could potentially change the vocabulary of warfare, just like gun powder and nuclear arsenal upon their entry into the battlefield. In April 2013, NGOs associated with successful efforts to ban landmines and cluster munitions got together in London and issued a call to governments urging the negotiation of a treaty preventing the development, deployment and use of what are known as'Killer Robots' in popular parlance. In July 2015, some of the world's leading Artificial Intelligence (AI) scientists including Apple co-founder Steven Wozniak, Skype co-founder Jaan Tallin and Professor Stephen Hawking signed a letter with nearly 21,000 signatures asking for an outright ban on these autonomous weapons systems (AWS). "Autonomous weapons will become the Kalashnikovs of tomorrow," states the letter.

Welcome to California Inc., the weekly newsletter of the L.A. Times Business Section. Expect financial markets to face headwinds today after the Federal Reserve reported Friday that U.S. industrial production fell more than expected in March. This is the latest sign that economic growth slowed significantly in the first quarter. On the plus side, though, many economists still forecast a rebound in growth as the year plods ahead. Tax deadline: Monday is the deadline for most Americans to submit their tax returns.

We live in an age of disruption -- and that's a good thing. Old systems will collapse as entrepreneurs figure out how to optimize and reinvent inefficient businesses, products, and services to provide consumers (us) with all things better, faster and cheaper. According to the Olin School of Business, 40% of today's Fortune 500 companies will be gone in the next 10 years. This post is a quick look at three industries (healthcare, finance and insurance) that are ripe for disruption this decade due to big data and artificial intelligence. Clearly big data and AI will change almost every industry this decade...but none more than these. Healthcare is so massively broken, that its disruption will come easy and happen fast.

TOKYO--Japanese industrial robot maker Fanuc Corp. said Monday it has developed technology to connect robots to networks that it hopes will give it an edge as it races to build the factory of the future. The technology is meant to enable factory owners to easily customize their robots and systems by downloading any needed applications, much like a consumer downloads smartphone apps on iOS or Android platforms.

This work is a further study on the Generalized Constraint Neural Network (GCNN) model [1], [2]. Two challenges are encountered in the study, that is, to embed any type of prior information and to select its imposing schemes. The work focuses on the second challenge and studies a new constraint imposing scheme for equality constraints. A new method called locally imposing function (LIF) is proposed to provide a local correction to the GCNN prediction function, which therefore falls within Locally Imposing Scheme (LIS). In comparison, the conventional Lagrange multiplier method is considered as Globally Imposing Scheme (GIS) because its added constraint term exhibits a global impact to its objective function. Two advantages are gained from LIS over GIS. First, LIS enables constraints to fire locally and explicitly in the domain only where they need on the prediction function. Second, constraints can be implemented within a network setting directly. We attempt to interpret several constraint methods graphically from a viewpoint of the locality principle. Numerical examples confirm the advantages of the proposed method. In solving boundary value problems with Dirichlet and Neumann constraints, the GCNN model with LIF is possible to achieve an exact satisfaction of the constraints.

We consider the problem of learning a non-negative linear classifier with a $1$-norm of at most $k$, and a fixed threshold, under the hinge-loss. This problem generalizes the problem of learning a $k$-monotone disjunction. We prove that we can learn efficiently in this setting, at a rate which is linear in both $k$ and the size of the threshold, and that this is the best possible rate. We provide an efficient online learning algorithm that achieves the optimal rate, and show that in the batch case, empirical risk minimization achieves this rate as well. The rates we show are tighter than the uniform convergence rate, which grows with $k^2$.

To whom correspondence should be addressed; Email: artitw@gmail.com Customer temporal behavioral data was represented as images in order to perform churn prediction by leveraging deep learning architectures prominent in image classification. Supervised learning was performed on labeled data of over 6 million customers using deep convolutional neural networks, which achieved an AUC of 0.743 on the test dataset using no more than 12 temporal features for each customer. Unsupervised learning was conducted using autoencoders to better understand the reasons for customer churn. Images that maximally activate the hidden units of an autoencoder trained with churned customers reveal ample opportunities for action to be taken to prevent churn among strong data, no voice users.