Artificial Intelligence (AI) is a commonplace term; everything is “AI powered” these days. A closer look however reveals that there is no “AI” anywhere to be found in most cases Parties so wrongly labeling their system with the term “AI” get away with it, telling us how misunderstood Artificial Intelligence generally is. Moreover, the boundaries between Artificial Intelligence, Machine Learning (ML) and Data Analytics have unfortunately been blurred in common conversations amongst business leaders and the general public leading to wrong and inflated expectations.
However, what strikes me the most and what we should collectively care about is the missing “D” in the conversation:Data. Apart from data scientists or people training systems nobody seems to care about data, yet that element is the most challenging component if you consider it for just a moment. That is not to say that developing systems and algorithms is easy, far from it, but in this day and age the tooling for most common problems is available. What we often lack is data. Or to be specific, sufficient, relevant and unbiased good quality data relevant to our challenge with the required rights to use. And that’s where things become challenging. Insufficient data does not allow us to train a system adequately and poor or wrong data (naturally) leads to poor, biased or misleading results (or all of the above). From a user’s perspective poor results are annoying, which in itself is a challenge, but even worse biased or misleading results are largely in transparent to the end-users and can have wide ranging consequences. Just take a look at this article as one of many examples that discuss the terrible impact that a combination of a blind belief in AI and use of biased data can have https://www.technologyreview.com/s/612775/algorithms-criminal-justice-ai/ .
Don’t get me wrong, analytics, AI and machine learning are great and powerful technologies that are of profound value helping us address a wide range of challenges and deliver stunning applications of the technology. However, as engineers, managers, leaders and citizens we have an obligation to ensure that the technology is applied responsibly and that we keep on asking the tough questions. Awareness of the challenge and asking questions is a good starting point!