"I don't think Homo sapiens-type people will exist in 10 or 20 years' time," Getnet Assefa, 31, speculates as he gazes into the reconstructed eye sockets of Lucy, one of the oldest and most famous hominid skeletons known, at the National Museum of Ethiopia. "Slowly the biological species will disappear and then we will become a fully synthetic species," Assefa says. "I believe [we] can inspire the Ethiopian youth to actually get really engaged in AI and feel like it's their thing." "Perception, memory, emotion, intelligence, dreams -- everything that we value now -- will not be there," he adds. Assefa is a computer scientist, a futurist, and a utopian -- but a pragmatic one at that. He is founder and chief executive of iCog, the first artificial intelligence (AI) lab in Ethiopia, and a stone's throw from the home of Lucy.

Question answering (QA) has been a longstanding challenge in the field of artificial intelligence. Numerous research works have pushed forward techniques for building QA systems. Many existing approaches achieve high performance on benchmark datasets. However, most of the questions in those datasets only require surface-level reasoning, and do not reveal the full-scale complexity and challenge of the question answering problem. Recently, the AI2 Reasoning Challenge (ARC) has been proposed [Clark et al., 2018], which is designed to pose a challenge to the QA community. On the ARC Challenge Set, several state-of-the-art QA systems, including leading neural models from the well-known SQuAD and SNLI tasks, only perform slightly better than the random baseline. This striking observation has demonstrated that QA is still far from being solved. Why it is so difficult to answer the questions in the ARC Challenge Set? 1) ARC consists of natural science questions, namely questions authored for human exams. All of these questions are drawn from real exams; 2) In order to encourage progress on hard questions, a Challenge Set has been partitioned from ARC.

We introduce a method for following high-level navigation instructions by mapping directly from images, instructions and pose estimates to continuous low-level velocity commands for real-time control. The Grounded Semantic Mapping Network (GSMN) is a fully-differentiable neural network architecture that builds an explicit semantic map in the world reference frame by incorporating a pinhole camera projection model within the network. The information stored in the map is learned from experience, while the local-to-world transformation is computed explicitly. We train the model using DAggerFM, a modified variant of DAgger that trades tabular convergence guarantees for improved training speed and memory use. We test GSMN in virtual environments on a realistic quadcopter simulator and show that incorporating an explicit mapping and grounding modules allows GSMN to outperform strong neural baselines and almost reach an expert policy performance. Finally, we analyze the learned map representations and show that using an explicit map leads to an interpretable instruction-following model.

This paper analyzes asymptotic performance of a regularized multi-task learning model where task parameters are optimized jointly. If tasks are closely related, empirical work suggests multi-task learning models to outperform single-task ones in finite sample cases. As data size grows indefinitely, we show the learned multi-classifier to optimize an average misclassification error function which depicts the risk of applying multi-task learning algorithm to making decisions. This technique conclusion demonstrates the regularized multi-task learning model to be able to produce reliable decision rule for each task in the sense that it will asymptotically converge to the corresponding Bayes rule. Also, we find the interaction effect between tasks vanishes as data size growing indefinitely, which is quite different from the behavior in finite sample cases.

Stochastic gradient descent is the method of choice for large scale optimization of machine learning objective functions. Yet, its performance is greatly variable and heavily depends on the choice of the stepsizes. This has motivated a large body of research on adaptive stepsizes. However, there is currently a gap in our theoretical understanding of these methods, especially in the non-convex setting. In this paper, we start closing this gap: we theoretically analyze the use of adaptive stepsizes, like the ones in AdaGrad, in the non-convex setting. We show sufficient conditions for almost sure convergence to a stationary point when the adaptive stepsizes are used, proving the first guarantee for AdaGrad in the non-convex setting. Moreover, we show explicit rates of convergence that automatically interpolates between $O(1/T)$ and $O(1/\sqrt{T})$ depending on the noise of the stochastic gradients, in both the convex and non-convex setting.

Neural networks are machine learning algorithms that provide state of the accuracy on many use cases. But, a lot of times the accuracy of the network we are building might not be satisfactory or might not take us to the top positions on the leaderboard in data science competitions. Therefore, we are always looking for better ways to improve the performance of our models. There are many techniques available that could help us achieve that. Follow along to get to know them and to build your own accurate neural network.

It's impossible to collect all the relevant data to answer any particular question, so there is necessarily uncertainty in our analysis. As such, we need to quantify the uncertainty and from that judge our results. Traditional statistical methods (also called frequentist methods) such as hypothesis testing and confidence intervals often don't address this appropriately. For example, we typically want to know the probability that a parameter falls in some range, but this type of analysis is unavailable from a frequentist perspective. Developing a statistical model with frequentist methods is often out of reach for typical data analysts so they are left asking "What test do I apply to this data?"

The game was developed by Revived Games, published by Acid and led by a person named Ata Berdiyev. Valve spokesman Doug Lombardi says Berdiyev had previously been kicked off the platform under a different business name.

When I was a graduate student in cognitive science, I spent countless hours poring over videos and transcripts of natural language, looking for patterns in the data that could help me better understand how people learn words, concepts, and categories. We support the company's mission to make AI beneficial to everyone by helping educate Googlers and others on how to build machine learning (ML) models that look for patterns in data in order to solve a variety of problems. Back in February, our team shared our internal Machine Learning Crash Course (MLCC) with the world to help more developers learn to use ML. Since then, we've heard from many people who are hungry for more ML education. In particular, you want to learn from teams who have built and deployed ML models.

A blood-thirsty video game that encouraged players to take part in a school shooting has been pulled by the publisher after it triggered a furious backlash. 'Active Shooter' was marketed on its ability to allow players to take on the role of a gunman on a murderous rampage inside a school, as well as a SWAT team member trying to stop the bloodshed. As the lone gunman, players would be shown a tally of the number of civilians and police officers they managed to kill during their simulated shooting spree. Anti-gun violence charity Infer Trust described the game as'horrendous' and in'bad taste' given the recent mass shootings in the US. An online petition calling for the game to be scrapped gained more than 194,700 signatures.