CEO’s Letter: I, Robot (NOT)

HRO Today

By Elliot H. Clark

With all due respect to Isaac Asimov, the continuing reliance on technology has yet to produce the predatory examples of “I, Robot,” or the scary dominance of machines of “Terminator” (which threatened to come back and kept on doing so in sequels, sadly), or the terrifyingly logical and murderous Hal of “2001: A Space Odyssey.” In truth, software robots, manufacturing robots, and even those Roomba vacuum cleaners are pretty boring (unless you see a YouTube video of a cat riding one). For HR, the questions are: What are robots? And what are they not?

HR applications of technology are different than automobile assembly plants, which deal with inanimate production products. The frothy excitement that is seen in the HR press about “bots” and their future is pretty speculative and at some level, in spite of the promise of technology, I just don’t care. And as you will see below, that is the point.

In truth, the robots return the favor. They don’t care about me, anyone else, or anything. That’s because they can’t care.

Robots have been mislabeled as “artificial intelligence.” The one word in the label that’s true is “artificial.” They are man-made creations and they are not, have never been, and as currently configured, will NEVER BE intelligent. What the advertising gurus that develop messaging for software don’t understand is that the word intelligent has a specific meaning. It connotes the ability to judge, to improvise, and in humans, the ability to emote or to care. If we continue mislabeling, the Age of Reason will have offi cially ended because we seem to no longer recognize the difference between the ability to reason and the inability to reason.

Today’s software bots engage in machine-based learning. They can observe and replicate. They can learn driving by analyzing traffic “transactions.” You don’t need to tell them to stop at red lights. They observe millions of traffic pattern data points and just begin to replicate. They can do it at breathtaking speed. However, they learn by the same premise that led to the old proverb: “monkey see, monkey do.” Does that sound intelligent?

It was bots that landed Amazon in trouble with the EEOC. I know many of the HR professionals at Amazon and they are some of the best in the business. The bots had flaws that led to replication of patterns of bias that Amazon would ethically abhor, but the company had no idea bias was taking place inside the black box of technology.

I am, to be clear, not opposing the rise of technology in HR. I am, however, strongly advocating for calling it what it actually is: machine learning, not intelligence. Many transactional areas of HR can and should be automated, but when there is a function that requires caring and requires sensitivity, it must involve humans. The human brain is still the most sophisticated, electrically-based computer on Earth.

The debate will rage over the next decades over where the line should be drawn. There are some companies experimenting with using chatbots online and on the phone to manage first-line triage in employee relations call centers. Some HR leaders see this as great; others are horrified. Consider these examples: If an employee is calling to verify dental coverage, a bot could likely handle the questions. But if an employee is calling to report an incident of harassment, this is likely too sensitive for an online software program to manage.

Where do you stand on the deployment of technology? Everyone has a different perspective and this will be an important topic for HR to address as a community. We know HR cares about how this impacts the workforce. And caring still makes us better than the bots, and we should celebrate our flawed humanity every chance we get.

Posted May 6, 2019 in Enabling Technology

Leave a Reply