Where do all these factors come together in the decision intelligence effort? Well, that’s what you and your organization have to figure out for yourselves. Certainly, there are guidelines on how to do it as well as tools at your disposal that I present in greater detail later in this book. However, remember that the first part of bringing humans more fully into the decisioning roles is in deciding the particular blend of human versus machine processes that are needed.
Or, to put it another way — and in keeping with Kozyrkov’s earlier analogy — this is the part for the microwave chefs to work their magic in making the recipe. The job of the microwave builders is largely finished. You know you have the right recipe when you see the proof in the pudding, so to speak.
Chapter 3
Cryptic Patterns and Wild Guesses
IN THIS CHAPTER
Yahoo! put the first Hadoop cluster — arguably, the first truly successful distributed computing environment designed specifically for storing and analyzing huge amounts of unstructured data — into production back in 2006. It’s that date which, for most practical purposes, marks the onset of the big data gold rush and the hunt to discover unknown information buried in known data sets.
The results were largely perceived to be worth the effort and generally enlightening — even though most big data initiatives fail to this very day. Even so, the call for data driven businesses, to the chagrin of business leaders and managers everywhere, became the mantra in business and investment circles worldwide. Organizations were soon convinced that using data analytics meant the same thing as harvesting answers. The thinking was that the answers generated were perfect right out of the box and were produced by means far beyond mere human abilities. Gut instinct and human talent were summarily discounted and dismissed as little more than wild guesses. However, the reality was and is quite different, as analytics have limits, big data and AI projects have high fail rates, and business executives very often let their gut instincts override algorithm outcomes.
Fear of AI began to soar as people expected machine masters to leap from science fiction and rule the real world. But that’s a far cry from what has happened so far.
The notion that data analytics was somehow churning out answers at a record pace gave way to the broadening realization that what analytics actually was producing were insights. Humans were still needed to glean understanding and perhaps inspiration from those insights and to turn them into decisions and actions.
It turns out that machines aren’t the new masters of the human race, after all. And they don’t provide the final answers humans seek. But that’s more the fault of humans than the machines. People were so busy asking questions of the data that they forgot to look where the work was headed. Organizations often found themselves working in circles or solving problems that yielded no tangible benefits for the questioner.
Machines Make Human Mistakes, Too
People commonly believe that machines are unbiased and more perfect than humans. Data analytics, automation, and machine learning (referred to as AI by marketers everywhere) are often presented as though the machines are capable of sorting out the data and reaching a perfect and fair conclusion on their own.
This simply isn’t true. It’s imperfect humans — not perfect machines — who make the technologies, set the rules, design the models, and select the training data. That means subconscious or intentional human influence can seep into every step: the rules, the programming and models, and the data selection. In short, the creation mirrors the creator. Machines are influenced by humans who build them and therefore frequently make many of the same mistakes humans make. Examples are numerous and varied. They include institutional biases, such as the (infamous) example of the continued use of redlining, a discriminatory practice in bank lending and other financial services that draws a figurative redline around minority neighborhoods so that those residents either can’t get loans approved or can’t get them at fair terms.
Such biases in computing are insidious and not entirely new. For example, a computer algorithm used in 1988 to select medical school applicants for admission interviews discriminated against women and students with non-European names. Similarly, Amazon ended a recruiting algorithm in 2018 that proved to be biased against women.
AI can ease such problems or make them worse. In any case, reverting to human-only decision making obviously isn’t the answer.
Be proactive: Use AI specifically to seek and measure discrimination and other known decision flaws throughout an entire decision-making process. Yes, this is using AI to make other AI and humans transparent and accountable.
Recognize the problem: Use algorithms to identify patterns, triggers, and pointers in actions, language, or rules that lead toward discrimination in order to set safeguards against discriminatory precursors in machine and human behaviors.
Check the outcomes: AI operates in sort of a black box, where humans can’t quite see what it’s doing. Yet AI cannot yet explain itself or its actions, either. But that’s okay — you can still check and grade its homework. For example, when checking for fairness in data-based or automated recruitment and hiring, look to see whether the outputs meet current legal standards such as the 80 percent rule — the rule stating that companies should be hiring protected groups at a rate that’s at least 80 percent of that of white men. Software developers should also perform disparate impact analyses — testing to see if neutral appearing functions have an adverse effect on a legally protected class — before any algorithm is used by anyone. If your software is from a third party, ask to see the results of the analysis and a detailed explanation of how the product works.
Do the math. Statistical analysis has been around for a long time. You can perform an old-fashioned and routine statistical test to reveal disparities arising from unintentional biases based on gender, race, religion, and other factors. Be sure, however, to automate the math rather than do it manually, because an automated process scales better, speeds results, and is likely more accurate.
