don’t know what they are doing and don’t know when they are dead wrong. People have to catch the errors and retrain the system for improvement. AI algorithms adjust their hidden layers by trial and error. If, in this work, they figure out a “hidden pattern,” then we may not ever know how it did, any more than the computer does, given the number and sheer complexity of the layers. Much of AI calculations go on in a “black box.”
AI’s blindness to its own workings is as bad as its brainlessness. It is a huge problem for compliance with law, especially in the European Union, where people have the right to know why the algorithm did what it did—why, for example, their application for a loan or insurance or a job was rejected.
But sometimes interesting trends do emerge. One bank determined that among its customer base, those who filled out the loan application in all caps were riskier—that is, defaulted at a higher rate—than those who used both upper- and lower-case letters (correctly, we assume). This is an example of AI exposing a hidden pattern, but it takes a human to interpret and act on it. And the correlation probably has nothing to do with causation. What to do with this information is up to the bank. Should the system be configured to accept only those applications that use upper- and lower-case letters? Should applicants be warned not to use all caps? Bank personnel will have insights on this matter, not the AI.
There is a set of “unsupervised learning” algorithms that conduct statistical analysis of data to identify relationships among data entries, such as clusters, associations, regression, time series, etc. These are actually standard data-mining tools of the data scientist, not a mysterious form of insight.
So, if you ever meet an AI vendor who claims their algorithms think better than you do, jack up the BS sensor.
Myth 5: AI Means Easy Money
Just letting AI algorithms loose on your business’s data will not result in automatic cost cuts, revenue enhancement, and correspondingly higher profits. Reform of the business process must accompany the use of the AI tool.
Adopting AI successfully, profitably, is as much about adjusting your standard operating procedures and accommodating the people who manage them as it is about the new tech. If it can’t be done profitably, it shouldn’t be done at all. Chapter 4 will help you with that important work.
Some pro-AI futurists say that AI will allow everything, every task and every job, to be automated, so firms will barely need any workers and will be rolling in dough. AI-enabled “singularity” will see auto-generating cycles of self-improvement and relentless acceleration. Zealots of AI-powered “superintelligence” say that it will rid the world of poverty and strife, and that a new humanity, a living hybrid of man and machine, will bring the miracle of happiness to all. Such authors are smart to predict the arrival of singularity in a couple to a few decades—just long enough for people to have forgotten the nonsense when the date passes.
Singularity technophiles are like religious sects, who, over the centuries, based on their reading of the Bible, pinpointed the time and place of Jesus’ return to earth. Similar gatherings go on in the United States today. The result is always the same. The heavens are not torn asunder, and the Messiah does not descend in Glory. The group recalculates, pushing off the date by a few decades, or centuries, in an attempt to save face.
Bad or unrealistic AI deployments can cost you big-time. MD Anderson, the University of Texas’s cancer center, burned through $62 million trying to get IBM’s Watson (AI services and toolkit) to automate their cancer diagnoses and treatments. IBM pocketed at least $39 million for Watson’s data processing, and PricewaterhouseCoopers another $21 million to develop and manage the business plan. Believers at MD Anderson claimed that leukemia was all but cured, but after $62 million, not one patient had been treated. The project was canned as quietly as possible. The whole venture was shady to begin with: A multimillionaire from Malaysia, Low Taek Jho (often called Jho Low), supposedly put up $50 million.[6] This is the same Low Taek Jho who then allegedly helped to defraud the Malaysian government of a few billion US dollars from the 1MDB fund, managed by Goldman Sachs. Ugh.
If there is such a thing as easy money, all too often it comes with high costs in other ways, no?
The last two myths we can dispense with in short order.
Myth 6: AI Drives Businesses
AI does not drive your organization; you and your coworkers do. AI will not take over your business or your life, but you would be wise to make use of it, as it best fits your needs. It can enhance your knowledge about your customers and your own coworkers. It can streamline some operations, help automate mind-numbing work, and lighten the load in some tasks.
AI processes data, and data on their own have no decision-making power. Analyzing data can tell you things about your customers, your suppliers, your partners, and your coworkers, but only in part. I think all would agree that working with people is the most complicated part of any business.
Never believe that by getting this or that AI system, you will be able to put this or that business function on autopilot and tune out. That rarely if ever ends well.
Myth 7: AI Will Control Your Mind
The opposite of this myth is true. AI systems at Google and Facebook process oceans of data and classify you as this or that for advertisers willing to pay them for the results, but it’s up to you to buy their goods and services.
To integrate AI successfully into your business, you will have to work with AI vendors or your own team of coders, data scientists, and project managers, but you should never defer to AI outputs—at least, not totally or unconditionally. By all means, take them into account, but remember that people make the decisions and bear the responsibility, not machines.
Many people may have a deep-seated longing to have the perfect servant—one that anticipates your every need and whim, one that never complains, is always quick to respond, and provides pure convenience without a hint of trouble or nuisance. Iron Man has his Jarvis and Friday. The commercial success of Amazon’s Alexa, Google’s Home and Assistant, Apple’s Siri, and Microsoft’s Cortana, despite their severe limitations, indicates just how many people share that desire.
But AI will never tell you what it is that you really want. No AI system will know you better than you do yourself. Unless you lie to yourself. And some people, unfortunately, do that very well.
Having done away with these myths, let’s agree on AI common sense. Working with AI requires all your intelligence and diligence. Below are seven important rules for getting it right.
Rule 1: Data Is the Mother of AI
We don’t want to take this metaphor too far, but think of an AI system as a family. Data is the mother, and if Mom isn’t happy, no one is happy.
Data is where every AI system begins. AI depends on data quality and quantity. “Garbage in, garbage out,” is still the rule. From biased data come biased results, bad business decisions, and big potential legal problems. You will need to bestow a lot of love on your data. You need to compile it, integrate it, and shatter the silos that prevent you from bringing it together. You will need to prepare it, repair it in places, and maintain it.
You will have to work with both “structured” data—the type your algorithms can search and query easily—and “unstructured” data—pretty much everything else—the kinds
