Newest forms of AI putting HR pros on notice: It’s a brave, new world
The buzz about ChatGPT has been swirling for a few months now. The artificial intelligence (AI) chatbot capable of composing answers to all kinds of questions effortlessly and instantly is being touted as the iPhone moment for AI and even likened to the Industrial Revolution. Much of the talk has centered on its use in academic settings, where some consider it a new way to cheat, and others want to help students understand how it can be used properly.
But the potential goes far beyond classrooms and certainly could shake up the employment realm. In fact, HR professionals are excited about passing off some time-consuming tasks, but also worried about a chatbot putting them out of a job. Regardless of how people think now, the impact of ChatGPT and new AI products sure to follow are going to take some serious—and human—thinking.
Potential uses
ChatGPT is capable of “learning” and summarizing a wealth of information on all manner of topics. The more it’s used, the more it will learn.
Every day, HR professionals field questions from employees about benefits, company policies, and more. They recite the same information over and over and might appreciate an AI solution to such tedious tasks.
Recruiting and hiring also is always a challenge. Maybe the bot can help develop a brilliant plan. In addition, onboarding approaches, job descriptions, training strategies, and more may be made easier with an AI solution.
Whatever its use, HR thinkers are floating ideas. In January, HR analyst and author Josh Bursin blogged about his early exploration of ChatGPT and called its performance “astounding.”
“I’ve asked it questions like ‘what are the best practices for recruiting’ or ‘how do you build a corporate training program’ and it answered pretty well,” Bursin wrote. “Yes, the answers were quite elementary and somewhat incorrect, but with training they will clearly get better.”
HR software company PeopleForce is also among the many in the HR world looking at ChatGPT. The company published a blog entry in February on what may be the top HR functions. One use: writing job descriptions. PeopleForce asked for a job description for a senior JavaScript developer. Within seconds, the heavy lifting was done and only the location, salary, and work schedule needed to be filled in.
PeopleForce also asked for a list of interview questions for JavaScript developer candidates. The questions ChatGPT generated included: “Can you tell us about a particularly challenging project you worked on and how you approached solving it?” and “How do you handle version control in a team setting?”
PeopleForce also asked ChatGPT for a solution to a personnel problem. It told the bot of an employee who was dissatisfied with communication with the manager and asked for a solution.
ChatGPT returned results including scheduling a meeting with the employee and manager to discuss concerns. It also suggested HR could encourage open and honest communication between both parties and offer communication training or workshops to both the employee and manager.
Other results included suggesting that HR provide clear guidelines and expectations for communication, such as regular check-ins or setting communication goals. It also suggested that HR follow up regularly with both parties to ensure the issues are being resolved and to provide more support if needed.
Worrisome aspects
Imagination is the only limitation for how to use AI, but troublesome questions remain. How much can it be trusted?
When Bursin asked about recruiting best practices and building training programs, he was able to evaluate the information ChatGPT produced. But how should HR pros judge the results if they’re asking questions they don’t already know how to answer?
Bursin’s January blog also points out that the bot has been trained on the internet, which he notes is “a jumble of marketing, self-promotion, news, and opinion.” So, ChatGPT and similar technologies need “to source valid, deep, and expansive domain data.”
If the data and algorithms used are faulty, the results can lead HR the wrong way. Bursin also notes that an AI engineer he knows thinks ChatGPT will be biased because of the data it consumes.
No matter the shortcomings, the technology behind ChatGPT isn’t going away and no doubt will have an impact on how HR professionals do their work. Bursin ended his January blog entry by saying “let’s think about this as one of the brightest stars in our future and try to prevent it from getting out of control.”
