Putting ChatGPT to the HR Test: Here’s What We Discovered

Short Summary:

• ChatGPT aids in automating HR tasks, freeing up time for leaders.
• While useful, it’s not a replacement for human judgment in HR decisions.
• Its limitations include outdated information and potential misinformation.

In an era where artificial intelligence is making significant strides, ChatGPT stands out as a remarkable tool for assisting with human resources tasks. Launched by OpenAI in November, ChatGPT has quickly garnered millions of users and generated much buzz about its revolutionary potential. Despite its obvious advantages, experts in HR and technology are cautious about relying on it exclusively for critical HR functions.

The Propositions and Promises

According to Bill Gates, AI is as revolutionary as the creation of the microprocessor and the internet. He asserts that it will fundamentally change industries and shift how people work, learn, and communicate. This tech giant’s enthusiasm is echoed by companies like Morgan Stanley and PwC, who are already incorporating chatbots into their work processes.

“The development of AI is as fundamental as the creation of the microprocessor, the personal computer, the Internet, and the mobile phone.” – Bill Gates

Unsurprisingly, Resume Builder recently found that 90% of business leaders believe job candidates with ChatGPT experience have an edge. This illustrates the growing acceptance of ChatGPT as a tool that can add value to routine tasks, thereby potentially enhancing productivity.

ChatGPT in Action: The HR Test

To explore ChatGPT’s efficacy in HR, a scenario was constructed to handle a common leadership challenge: addressing an underperforming employee. Olivia, a team leader, needed to discuss missed deadlines and other issues with Liam, a remote worker. The challenge was to draft an initial email and prepare for a critical performance review.

ChatGPT crafted an email draft in seconds, which, while not perfect, was impressively close to what might be considered a final version. The AI then generated an agenda for the meeting, leaders with a valuable starting point. Additionally, the AI anticipated three possible explanations that Liam might give for his poor performance, showing its potential for role-playing scenarios and pre-meeting preparations.

Capabilities and Limitations

ChatGPT can automate numerous HR tasks, such as:

• Crafting job descriptions
• Drafting emails to job candidates
• Summarizing research reports
• Generating employee surveys
• Preparing performance reviews

Despite these capabilities, the AI shows limitations. ChatGPT is based on data up to 2021 and has been known to generate erroneous or misleading information. For example, in a test by Mineral, the AI could handle straightforward HR questions but faltered with nuanced queries. Susan Anderson, Chief Services Officer at Mineral, noted that ChatGPT’s answers often lacked detailed accuracy, and in some instances, the AI fabricated references.

“In almost all cases, the answers ChatGPT provided were missing necessary nuanced details, and in some cases, ChatGPT completely fabricated reference material that did not exist at all.” – Susan Anderson

Prompt Engineering: art or hype?

The quality of responses from AI like ChatGPT heavily depends on the questions asked—an art known as “prompt engineering”. While some see it as a critical skill, others believe it is a passing phase. Ethan Mollick from Wharton School argues that as AI gets better, the need for specialized prompt engineering will diminish.

“‘Prompt engineering’ is not going to be a big deal in the long-term & prompt engineer is not the job of the future.” – Ethan Mollick

Nevertheless, well-crafted prompts can lead to more accurate and tailored outputs, making tasks such as writing job descriptions, interview questions, and compensation data searches more efficient.

Concerns and Ethical Considerations

Generative AI capabilities promise to democratize access to information, making it easier for diverse groups to communicate. Yet, these technologies also risk perpetuating biases present in their training data. This duality makes it essential to use AI responsibly and critically.

Cybersecurity also poses a significant concern. Cyberhaven’s Ting suggests avoiding sharing sensitive or confidential information with ChatGPT because the learning AI stores user inputs. This means any shared data could become part of its accessible information pool, potentially leaking confidential corporate data.

“We’ve seen doctors putting patient information into ChatGPT to draft letters to insurance companies, automating what used to be a tedious task but also violating HIPAA in the process.” – Howard Ting

The Future of HR with AI

While ChatGPT can be a powerful aid, it cannot replace the human touch in HR. Evaluating a candidate or addressing sensitive employee issues requires judgment, empathy, and experience that AI cannot replicate.

That said, ChatGPT can significantly streamline initial stages of HR processes. By combining the tool’s initial outputs with human expertise, HR professionals can obtain reliable and accurate answers more quickly. However, it’s essential to review AI-generated material for accuracy and relevance.

Closing Thoughts

ChatGPT serves as an effective starting point for various HR tasks, but its limitations ensure that human oversight remains crucial. Leaders can leverage AI to free up time for more strategic and human-centric activities, enhancing overall productivity and focus. However, ethical considerations, privacy concerns, and the importance of accuracy should never be underestimated.

As we continue to explore the potential of AI in HR, it’s clear that tools like ChatGPT are here to stay. But just as crucial is the need for skilled professionals who can guide, interpret, and refine the contributions of these emerging technologies. For more insights into AI in writing and HR, visit Autoblogging.ai.