It has not been long since we have entered into a new era in the relationship between humans and machines. While it has been almost a decade since AI forced us to think hard about our future, over the past year it seems that we have begun a more serious questioning about the future of everything.
As hard as I try not to write yet another piece about generative AI (I am favoring this term over the narrower name “ChatGPT”), the pull is simply unavoidable. The more I read about examples of how this tech is being used, and the more I mess around myself, the more transformative I feel these tools are.
Indeed, generative AI is already creating new jobs. Just a few months ago when people talked about a “prompt engineer,” they were talking about a punctual coder. Now it is probably the hottest new job on the planet.
No sir, this is not second life, this is not the metaverse, this is not employment games, this is not blockchain. It is much, much more.
And so as with every other discipline under the sun, the wires in the world of hiring are abuzz with talk of how generative AI is going to transform hiring.
Creating “Complimentarity”
Recently, I wrote an article for this site called “Will ChatGPT Kill Recruiting As We Know It?” I concluded that while generative AI solutions such as GPT are amazing, we have a long way to go before they replace humans or radically change hiring. I had pointed out that the tech was best seen as a sidekick to help impart wisdom and help work more efficiently. Hiring would just have to adapt and find its way just like everything else.
But as we move forward, it is becoming increasingly harder to tell applicants from machines. With AI-generated cover letters and resumes and AI-generated responses to coding challenges, how can we really know who we are hiring these days?
And this is just the beginning. It is only a matter of time before conversational AI-powered bots will take over more subjective parts of the hiring process, opening a situation that makes Pandora’s box look like a matchstick holder.
When it comes to the benefits, many are obvious, but others require a bit more contemplation. Which is why I started to wonder if there is actually an inverse effect of generative AI, whether it will actually increase the value of the things it can’t do.
I asked GPT about it, and it went like this:
Me: “Is there a term that summarizes a situation where new technologies actually increase the value of older ones?”
GPT: The term “complementarity” is often used to describe a situation where new technologies increase the value of older ones. In economics, complementarity refers to the extent to which the use of one product or technology enhances the value of another product or technology.
For example, the advent of smartphones and mobile internet technology has greatly increased the value of older technologies such as email and instant messaging. Similarly, the development of e-commerce has complemented traditional brick-and-mortar retail, as customers can research products online and then go to a physical store to make a purchase.
Overall, complementarity is an important concept in understanding how new technologies can build upon and enhance the value of existing technologies, rather than replacing or rendering them obsolete.
In this meditation and exchange it dawned on me that generative AI is actually going to save talent assessment. Not that talent assessment needs saving in a classical sense; we have decades of proof that assessment has tremendous ROI.
Saving assessment is all about its “complementarity” with generative AI.
Assessment is the science of accurately and reliably measuring individual differences. These individual differences are what make people unique. When applied to a work paradigm, individual differences are the building blocks of job success and organizational fit.
Once one breaks down a job into the individual differences (i.e., traits, competencies, etc.) that are required for success, assessments provide detailed, scientifically sound signals about an applicant’s potential. This formula also works for predicting a match between an applicant’s values and those of an organization, which is important because such matches are indicators of engagement, motivation, and tenure.
A Sea of Sameness
But in today’s world, generative AI’s ability to create resumes, cover letters, complete unsupervised coding challenges, and otherwise allow applicants to embellish themselves obscures the signal on individual differences. By allowing everyone to appear better, and creating a sameness amongst applicants, it is effectively creating a playing field that is too level to support effective decision-making.
This phenomenon is upping talent assessment’s stock significantly. When it comes to gaining an objective, reliable, and accurate signal about a job applicant, talent assessment offers many formats that are impervious to the use of generative AIs like GPT.
These assessment formats include:
- Formats where questions are not factual and have no clear right or wrong answers, such as personality and work values assessments. These measure individual differences that represent the type of soft skills and attitudes that are becoming increasingly more important to employers.
- Formats where novel questions appear quickly, are often timed, and make use of nonverbal item types. They are usually cognitive assessments, which are highly sensitive to individual differences.
- Formats that use live interactions that are often scored and evaluated by humans. This is often used for job simulations, assessment centers, and technical interviews, all of which are extremely powerful measures of individual differences.
If you don’t believe me, just ask ChatGPT about it. I did and here is what it told me:
“One potential application of conversational AI in talent assessment is the use of chatbots to conduct initial screenings and interviews with job candidates. Chatbots can ask a standardized set of questions and provide consistent feedback, which can save time and reduce bias in the early stages of the hiring process. However, chatbots are not yet capable of fully replacing human evaluators when it comes to assessing a candidate’s soft skills, cultural fit, and other more nuanced factors.”
“Conversational Talent Assessment”
It seems unlikely that generative AI will replace quality psychometric measurement anytime soon. Thus far, mortal AI-based tools certainly have not been able to do so for many of the same reasons that generative AI can’t.
Consequently, falling back on and investing in good ole fashioned talent assessment offers an island of sanity and clarity that will continue to help employers cut through the fog and identify a meaningful signal about job applicants. When embedded into more sophisticated platforms that make use of AIs of many types (including generative AI), talent assessment’s complementarity helps the whole become more than the sum of its parts.
When thinking about how your organization will evaluate talent in this brave new world of smarty-pants machines, live by the idea that the more things change, the more they stay the same. By this I mean there is no need to deviate from the tried-and-true formula of defining the individual differences needed for job success and organizational fit and then choosing assessment tools that have been proven to fairly, accurately, and reliably measure these things.
The new element to this formula is ensuring that you choose formats that are impervious to but have complementarity to generative AI. In doing so, we can create a win-win situation that highlights the value of collaboration between humans and machines.
When I struggled to find a name for this type of talent assessment, ChatGPT was kind enough to come up with one for me:
Given the potential of Conversational AI to revolutionize the hiring process, we propose the term “Conversational Talent Assessment” to describe this phenomenon. By leveraging the power of NLP and machine learning, Conversational Talent Assessment can help companies identify the most qualified candidates efficiently and effectively. However, it’s important to recognize that this technology is not a panacea for all hiring challenges and should be used in conjunction with other methods to ensure that companies are making well-informed and fair hiring decisions. As Conversational Talent Assessment continues to evolve, it’s likely that we will see even more innovative and exciting applications of this technology in the field of talent assessment.
You can expect the future to bring lots of conversation about “Conversational Talent Assessment.” Just remember you heard the term here first!