Nowadays, it seems like everyone is talking about Large Language Models (LLMs), such as ChatGPT, and virtually every conceivable topic. From cooking to Bible studies, to Dungeons and Dragons, it seems there's nothing LLMs can't do.
Are AI tools like ChatGPT a shortcut to getting the job done or a path to better hiring?
Large Language Models (LLMs) like ChatGPT are sparking debate. Are they a shortcut to laziness or a catalyst for productivity? In recruitment, the big question is: do they contribute to cheating or more efficient hiring?
Consider internet-based employment tests at the turn of the millennium. Despite concerns, this became a billion-dollar industry. With ChatGPT, we're seeing a similar development – new technology initially creating concern, but then becoming part of progress.
Cheating in Recruitment, or an Efficient Tool?
Neil Morelli from Codility defines cheating in recruitment as candidates using tools like ChatGPT to falsely demonstrate qualifications they don't possess. With this definition in mind, let's explore how ChatGPT is currently used in the recruitment process:
-
CVs and Cover Letters
Using ChatGPT to create CVs and cover letters, even if seen as a form of cheating, can be argued to improve the hiring process by emphasizing the importance of thorough assessments later in the selection process. -
ChatGPT in Interviews
There are cases where candidates use ChatGPT during interviews, which, while not widespread, could become more common. This innovative, albeit misleading, use of technology also shows a candidate's resourcefulness. -
ChatGPT in Pre-employment Testing
Traditional tests, often immune to cheating via ChatGPT due to their format, pose a challenge for AI. Research from Stanford highlights ChatGPT's limitations in accuracy, especially in visually based pattern recognition tests, which are common in pre-employment testing. -
ChatGPT in Knowledge and Skill Tests
For actual knowledge and skill tests, the threat of cheating via ChatGPT is more significant. Coding tests, for example, could be particularly vulnerable. However, embracing ChatGPT as a tool instead of a threat could transform how we evaluate candidates' ability to effectively use LLMs in their work.
“a no or low-knowledge candidate who wouldn't be successful otherwise is now using this to basically impute knowledge that they don't have and signal that they're qualified for a job that they're not qualified for.”
– Neil Morelli of Codility, an IT coding assessment provider
The Workforce of the Future and ChatGPT
The core issue in the debate about ChatGPT in hiring is not about replacing human skills but leveraging the synergy between humans and machines. This partnership can enhance what each party can do independently, adding value to the workforce. Despite fears that GPT enables unqualified candidates to cheat their way through the system, a well-designed hiring process with multiple evaluation steps can mitigate this risk. Additionally, ChatGPT offers significant advantages in the development of candidate evaluation and feedback frameworks, as highlighted by Alan Bourne from Sova.
Conclusion: How Should We View Candidates Using ChatGPT in Recruitment Processes?
The fear that ChatGPT will lead to a workforce filled with unqualified individuals is probably exaggerated. Instead, we should view LLMs as another tool in human-technology collaboration, capable of generating positive and meaningful outcomes in the workplace. As with all disruptive technologies, the key lies in finding the balance between embracing innovation and maintaining integrity in processes such as recruitment.
