Is it not a curious amusement to catch a vintage illustration or film depicting the future? Portraying how the technology of the day would evolve to serve the same social customs and contemporary jobs contrasts glaringly with what has become reality.

Illustration by Jean-Marc Côté sourced from commons.wikimedia.org

While quaint and sweet, the one element predicted that does stand the test of time is that mankind will find ways to improve productivity. The seeds of new perspectives, techniques, technology, and approaches might arrive to great fanfare, but more often evolution is gradual.  Sometimes we need to consciously recall the way a process was done a decade earlier to realize a change has transpired.

IT has been delivering productivity gains for generations. Applied like its physical counterparts such as hoes, hammers, or tractors, IT has helped by processing mountains of data quickly. For all sorts of tasks, end-users have been using tools to sort, filter out, find items, and the like in mountains of data.

Society has evolved to interact with the tools mankind develops blacksmiths into car mechanics, book-keepers into data entry operators. The relationship between all of this technological advancement has been that the end-user in this relationship has been taking the information and deciding what to do with it. 

But, an ages-old fantasy of mankind has been to go further: to develop technologies that can produce and then analyze those results.  This might have been a machine to play chess (von Kempelen, who supposedly actually relied on a hidden midget), or Arthur C. Clarke who proposed the computer HAL that, given conflicting orders, could suffer from a psychotic breakdown.

Illustration by Johann Wolfgang von Kempelen sourced from commons.wikimedia.org

While chess playing is now a reality and does not require a hidden operator, AI is emerging in other areas: handling menial tasks, allowing us to bypass keying in our questions to search engines, catching fraud, predicting behavior, and so much more. It offers a means to teach a system through examples of acceptable and unacceptable outcomes, feed it stimuli from a variety of sensors, and recommend a course of action. Using an AI is proving adept at dealing with data coming from a complex systems’ landscape.

The influx of AI is pervading most industries. And just as AI can process thousands of elements to navigate piloting a car on a road, it can also be used to steer companies when selecting candidates. And AI is becoming common place as consumers turn to Siri, Alexa, and other programs to offset tasks, plan days, recommend movies, or suggest the shortest route from A to B.

This trend has led to a shortage of software developers in the short term, as jobs of the future will require AI software engineers. And this trend has only accelerated because of Covid and the complexion of how we work has been transformed to being more digital.

This is not just happening in highly technical departments and businesses. No, AI is affecting HR as well. LinkedIn has begun to use AI to recommend positions and candidates based upon the data it holds and the jobs being posted. It is not simply filtering keywords, but looking to match similar registrants with roles that have been filled. 

SelectSoftwareReview.com posted over a dozen different AI products that can help recruiters. Some of the systems can trawl through resumés, looking for keywords, experience, and so forth. These can quickly scan millions of profiles and feed a pipeline with qualified candidates. Some tools that can optimize for which terms one searches, and look at the context in which those terms are used to better screen the candidates selected. Such can be useful when trying to estimate whether a candidate might fit the culture of a company.

While not yet able to automate the process of interviewing, there are AI tools now that assist with the recording and analysis of interviews thereafter. Focusing on the menial, repetitive elements of this aspect of recruiting saves time. And it allows professionals to focus on the more involved, human-centric work necessary to measure up potential candidates.

AI though can also help interpret the way candidates might react. A machine can more easily pose scenarios, mimic outcomes, and then review the participants’ responses. As opposed to doing such hypothetically in an interview, an AI routine can add many more details to the simulation, and by simulating the outcome, see how the candidates follow up.

These can be run offline in the sense that they do not require staff to administer such. And, as more people, in general, have some experience with gaming, are more easily delivered to participants as opposed to sitting a Myers-Briggs evaluation.

Atop the processing speed and volume that an AI can handle, there’s something another promise we might hope to find in AI: the absence of bias. Ingrained in all of us is some form of modeling and prejudice. Theoretically, knowing the training of an AI system should counter such. As AI systems are not designed to survive, theoretically, they are not secretly thinking, how can the AI take advantage of the situation thereby clouding its judgment.

Image attributed to Tom Cowap sourced from commons.wikimedia.org

While the fear of AI running amok is ever among us, the doomsday scenario is unlikely anything we can imagine. More likely, AI is making our entire world more complex. And without embracing the technology, we will find ourselves unable to compete with those who are not intimidated. Being only human, we are limited on how much we can process. As AI gains ground there will be more calls on us to teach AI which outcomes are desired and leverage the value AI brings to the party.