As AI algorithms play a bigger role in decision making, how do qualities like ethics, compassion, and inclusion get programmed into the code? On this edition of Bytemarks Café, a talk about the gathering of thought leaders in Hawai'i to discuss how to move the technology agenda. The event is called TechForce 2019, and its aim is to bring together leaders from key sectors to accelerate tech readiness in our islands. On this edition of Bytemarks Café, a discussion about a novel new project that projects a 3D hologram from Hawaii to American Samoa. The project is called Holo Campus, and is the delivery of University of Hawai'i lectures over the trans-Pacific fiber optic broadband network to the Pacific Islands.

It admittedly sounds a little like Big Brother, that a robot can tell significant things about your personality merely by looking into your eyes. Yet, that is the hiring territory that we are fast approaching – although we may not be sitting across from androids in interviews anytime soon. The use of artificial intelligence in making HR decisions is, while fraught with peril, not without its promising aspects. In an era when it is increasingly difficult for businesses to unearth the best job candidates, we may yet see the day when technology makes it possible to separate good from bad in the blink of an eye. Despite caveats about security and privacy, relying on AI would appear to be a method far superior to digging through a pile of resumes or asking ice-breaking questions like, "What's the last book you read?" Hiring good people – people who are talented, agreeable and work well with their co-workers – goes a long way toward nipping workplace conflicts in the bud.

G4 instances provide the industry's most cost-effective machine learning inference for applications, like adding metadata to an image, object detection, recommender systems, automated speech recognition, and language translation. G4 instances also provide a very cost-effective platform for building and running graphics-intensive applications, such as remote graphics workstations, video transcoding, photo-realistic design, and game streaming in the cloud. Machine learning involves two processes that require compute – training and inference. Training entails using labeled data to create a model that is capable of making predictions, a compute-intensive task that requires powerful processors and high-speed networking. Inference is the process of using a trained machine learning model to make predictions, which typically requires processing a lot of small compute jobs simultaneously, a task that can be most cost-effectively handled by accelerating computing with energy-efficient NVIDIA GPUs.

Finextra spoke to Amit Bhute, SVP & global head of the banking and financial services practice and Soumyendu Kishore Pal, co-head of the capital markets practice at Virtusa about what is driving digital transformation in financial services today. Bhute says that the rise of digitally native neobanks has showcased a significant gap in the market. "Banking legacy infrastructure that was built between the 1960s and 1980s is crumbling and is now unable to meet the demands of increasingly real-time and data-intensive customer demands. Banks have realised that sticking to the status quo is no longer enough and simply meeting regulatory demands will not support growth." Traditional banks must focus their digital transformation strategies on innovation, otherwise new revenue will captured by the digitally savvy neobanks.

Swiping is no longer the only way to find matches on Tinder. In a choose-your-own-adventure style series set to be rolled out next month, users will be able to match with other dating hopefuls by clicking their way through an interactive narrative. 'Swipe Night,' as Tinder is calling it, will air on October 6 and is designed to match users based on the choices they make during a short ''first-person apocalyptic adventure.' All of the episodes will be'live', so-to-speak, with each being available for viewing only between the hours of 6 pm and midnight during a respective users' local time. The series will consist of short five-minute videos during which users are periodically given seven seconds to choose what happens next.

Data augmentation (DA) is fundamental against overfitting in large convolutional neural networks, especially with a limited training dataset. In images, DA is usually based on heuristic transformations, like geometric or color transformations. Instead of using predefined transformations, our work learns data augmentation directly from the training data by learning to transform images with an encoder-decoder architecture combined with a spatial transformer network. The transformed images still belong to the same class but are new, more complex samples for the classifier. Our experiments show that our approach is better than previous generative data augmentation methods, and comparable to predefined transformation methods when training an image classifier.

This is the web version of Data Sheet, Fortune's daily newsletter on the top tech news. Sign up here to get it delivered to your inbox. The number of transistors packed onto a modern chip inside your phone or PC runs into the billions but it's still sometimes amazing to comprehend the computing power you can easily hold in the palm of your hand. When I met Intel vice presidents Gadi Singer and Carey Kloss on Wednesday, they showed me a new circuit board the company has created for speeding up artificial intelligence apps. The board is the size of an SSD drive, made to plug into a standard PC or server.

National Workshops to Explore Long-range Strategies 7. Collect suggestions from the national planning workshops, distilled in to 93 actions, assess all via five (5) Real-Time Delphi's 8. Final Report for Public Discussion 10.

A collaborative group that includes Salesforce, Deloitte, the World Economic Forum (WEF) and the United Kingdom Office of AI today introduced guidelines for government officials to procure artificial intelligence systems. The guidelines that counsel vendors to ask particular questions before selling their AI to government agencies is being called by the WEF the first for any national government worldwide. In the works for about 10 months, the guidelines made by the previously mentioned organizations were brought together by the WEF's Centre for the Fourth Industrial Revolution and its AI and ML team. The Centre for the Fourth Industrial Revolution hosts fellows from nations around the world to focus on major initiatives. Governments the world over are increasingly using AI to do things like predict the needs of citizens, help with health care screening, or do things in criminal justice like power predictive policing, track suspects with facial recognition, or determine if people deserve pretrial bail hearings.

Linking facts across documents is a challenging task, as the language used to express the same information in a sentence can vary significantly, which complicates the task of multi-document summarization. Consequently, existing approaches heavily rely on hand-crafted features, which are domain-dependent and hard to craft, or additional annotated data, which is costly to gather. To overcome these limitations, we present a novel method, which makes use of two types of sentence embeddings: universal embeddings, which are trained on a large unrelated corpus, and domain-specific embeddings, which are learned during training. To this end, we develop SemSentSum, a fully data-driven model able to leverage both types of sentence embeddings by building a sentence semantic relation graph. SemSentSum achieves competitive results on two types of summary, consisting of 665 bytes and 100 words. Unlike other state-of-the-art models, neither hand-crafted features nor additional annotated data are necessary, and the method is easily adaptable for other tasks. To our knowledge, we are the first to use multiple sentence embeddings for the task of multi-document summarization.