How will Gen AI change the future of recruitment? | Webinar with Hung Lee

About this Video

• Video Title: How will Gen AI change the future of recruitment? | Webinar with Hung Lee
• Channel: Arctic Shores
• Speakers: Hung Lee, Robert Newry, Jasper
• Duration: 01:05:19

Introduction

This webinar discusses the impact of generative AI on recruitment processes. The speakers present data-driven research on candidate usage of generative AI tools and their effects on various assessment methods. The main purpose is to inform recruiters about the changing landscape and offer strategies for adapting to the evolving situation.

Key Takeaways

• Generative AI usage among candidates is rapidly increasing: 72% of students regularly use generative AI tools, significantly impacting recruitment assessments.
• Simple prompt engineering techniques significantly improve AI performance on assessments: Even basic prompts can boost AI scores on verbal reasoning and other tests. GPT-4, without prompt engineering, outperforms many human candidates.
• Time limits don't effectively prevent AI usage: Candidates can quickly input assessment questions into AI tools and receive answers within time constraints.
• AI can outperform humans on certain reasoning tasks: Generative AI models can reason and even sometimes reason better than humans on verbal reasoning tests, exceeding the scores of 98.8% of human candidates in some cases.
• Self-report personality assessments are vulnerable to AI manipulation: AI can easily generate responses that maximize scores on socially desirable traits, significantly outperforming humans in manipulating responses.
• Task-based assessments show greater resilience to AI manipulation: The interactive and visual nature of task-based assessments makes them more resistant to AI manipulation than traditional text-based assessments.
• A redesign of recruitment processes is necessary: The speakers advocate for a shift away from language-based assessments to task-based methods and a focus on continuous assessment updates to maintain validity.
• In-person interviews might become more common: As digital assessments become increasingly susceptible to AI manipulation, in-person interviews may offer a more robust evaluation method.