Bias in AI and Machine Learning: Sources and Solutions - Lexalytics


"Bias in AI" refers to situations where machine learning-based data analytics systems discriminate against particular groups of people. This discrimination usually follows our own societal biases regarding race, gender, biological sex, nationality, or age (more on this later). Just this past week, for example, researchers showed that Google's AI-based hate speech detector is biased against black people. In this article, I'll explain two types of bias in artificial intelligence and machine learning: algorithmic/data bias and societal bias. I'll explain how they occur, highlight some examples of AI bias in the news, and show how you can fight back by becoming more aware.

Amazon and Google unwittingly approved speaker apps that eavesdropped on users and stole passwords

Daily Mail - Science & tech

Researchers successfully sneaked malicious apps behind the defenses of two major smart speaker companies in a test on their security practices. Experts at Security Research Labs say the apps were design to target personal data like voice-recordings and passwords of both Google Home and Amazon Echo users by posing as software that reads horoscopes through voice-commands. The apps were only removed once researchers made the company aware of their test. All eight of the apps designed by the researchers were able to bypass Amazon and Google defenses and were approved by the companies' moderation teams - a lapse that experts say invites even greater scrutiny on smart devices' privacy and safety standards. 'As the functionality of smart speakers grows so too does the attack surface for hackers to exploit them,' write the researchers in their report.

Costa Rica Puts Time and Attention into AI Development - Nearshore Americas


Artificial Intelligence (AI) is having a broad and deep impact on the way services are exported globally. Be it for good or bad, there is no getting away from the reality that AI is an agent of disruption. One of the perennial front-runners of Nearshore outsourcing, Costa Rica, appears to be adapting to the AI opportunity faster than most countries in the region. Local companies are intensifying their AI development operations and a number of AI technologies are gaining traction there – all of which will influence Costa Rica's positioning in the next-generation of services delivery. The Latin American nation of nearly five million has long been seen as a tech epicenter of Central America ever since Intel chose it to open the biggest microchip factory in the region in 1997, with an initial investment of US$800 million.

A Hive Mind – An Open-Source-Driven AI Platform -


As firms are coming under increasing pressure to adopt artificial intelligence (AI) and machine learning programs, working out how to integrate them into existing systems in the least disruptive and most efficient way is a pivotal consideration. Peter Simon, lead data scientist, financial markets at DataRobot, explores how firms can satisfy regulators' calls for interpretability, the sustainable advantages of setting up AI programs and why the often-overlooked route of an open-source code base could be key to maximising efficiency.

What Do Machine Learning and Hunter-Gatherer Children Have in Common?


Gul Deniz Salali is a British Academy research fellow and lecturer in evolutionary anthropology/medicine at University College London. She studies modern hunter-gatherer communities and teaches evolutionary medicine at UCL.

How Digital Transformation Is Revolutionizing These 5 Industries


Depending on the industry, digital transformation comes with its own unique trends, challenges, and opportunities. While enterprises in every sector are leveraging next generation technologies, there are five that are regarded as bellwether industries in digital transformation. Each of these industries share a common ambition to modernize processes. From artificial intelligence (AI) and the Internet of Things (IoT) to cloud computing and blockchain, these sectors are leveraging all types of innovative solutions. After speaking to these experts, Enterprise Digitalization developed a progress report that provides the current 2019 status of digital transformation in each of these industries.

Neural network image super-resolution and enhancement


Let's Enhance uses cutting-edge Image Super Resolution technology based on Deep Convolutional Neural Networks. Before appearance of this technology it was impossible to dramatically increase photo or image size without losing quality. Your best option in Photoshop, called Bicubic Interpolation - made your image unsharp and blurry. Those of you on the math side of the things could argue – however you increase image resolution – there is no new information about the image - you just can't add extra quality! This is not true in a case when Neural Network and AI is used.