In the recent past, artificial intelligence has come to the limelight by showing its massive abilities to tackle mundane as well as complex tasks. From complex facial recognition to identifying fraudulent transactions, AI possesses the power to solve challenging problems faced by individuals and organisations on an everyday basis. Most economic sectors, including transportation, retail, advertising, and energy, are being disrupted by data digitization on a large scale, as well as the developing technologies that utilise it. Computerized systems are being employed in government services to increase accuracy and drive objectivity, and AI is influencing democracy and governance. However, as with every other good technology, AI also has a fair share of challenges that is hindering its usage at a large scale.

Data analytics helps marketers learn about their customers with target precision, from the movies they watch on Netflix to their favorite scoop of chocolate ice cream. Data is ubiquitous, essential and beneficial -- except when it's not. Experts warn that data analytics is at an inflection point. Growing concerns about security risks, privacy, bias and regulation are bumping up against all the benefits offered by machine learning and artificial intelligence. Layer those concerns on top of worries about the coronavirus pandemic and how it has rapidly changed consumer behavior, and the challenges become clear.

That could cost you – in fact, the algorithms they use may be biased against you, or at least your travel plans. This is according to a study by George Washington University published last month. The report found that passengers being picked up or dropped off in lower-income communities or in sectors with minorities were being charged more per mile. "Uber determines demand for rides using machine learning models, using forecasting based on prior demand to determine which areas drivers will be needed most at a given time," reads the study by Aylin Caliskan and Akshat Pandey. "While the use of machine learning to forecast demand may improve ride-hailing applications' ability to provide services to their riders, machine learning methods have been known to adopt policies that display demographic disparity in online recruitment, online advertisements, and recidivism prediction."

What truly is the cause of algorithmic bias, can we truly point any fingers at who's to blame? The bias within algorithm and machine learning systems today emerges as a result of many circumstances. Deep learning models function in a manner that draws from the pattern recognition capabilities the neural networks. Therefore, it could be said that deep learning models cannot be directly biased by design, and any emergence or cause of bias is external to the architecture and design of the neural network. The outputs produced by machine learning models and AI systems are simply a reflection of the training datasets they are exposed to.

However quickly artificial intelligence evolves, however steadfastly it becomes embedded in our lives -- in health, law enforcement, sex, etc. -- it can't outpace the biases of its creators, humans. Microsoft Researcher Kate Crawford delivered an incredible keynote speech, titled "The Trouble with Bias" at Spain's Neural Information Processing System Conference on Tuesday. In Crawford's keynote, she presented a fascinating breakdown of different types of harms done by algorithmic biases. As she explained, the word "bias" has a mathematically specific definition in machine learning, usually referring to errors in estimation or over/under representing populations when sampling. Less discussed is bias in terms of the disparate impact machine learning might have on different populations. "An allocative harm is when a system allocates or withholds a certain opportunity or resource," she began.