Hi Keryn... maybe we should re-visit your question now this topic is even more in the zeitgeist.

Actually, that suggests it will pass, but this will be transformative beyond our imaginations!!

In reply to Steve Bridger:

Hi Laura  - how did your dissertation go back in 2021, which you mentioned in this thread?!

I've had a similar experience recently where there's just something...eerie/uncanny about some applications which have been recently coming through which has made me suspect that they're AI-generated at least in part.

In the absence of anything to accurately detect it, I'm shortlisting the applications on face value. Although it's potentially more time-consuming if the candidate is wholly unsuitable, I figure it's hard to hide behind a face-to-face interview and if an AI-generated application helps someone who has trouble with writing job applications then I'm kind of okay with it (I remember asking people for help when writing applications/personal statements when I was earlier in my career, so I figure it's not too different!)

In reply to Christopher:

I'd agree with this approach in principle. It can often be difficult to put onto "paper" what you're thinking and if people are using it as a tool to do so then I don't necessarily have too much of an issue with it as long as they're then capable of backing it up when in a face to face interview. It's going to be pretty obvious at this stage if someone doesn't have the knowledge or understanding to back up what they're submitting in their assessment answers.

It's also not going to be possible to put a stop to it as well so it means that businesses are going to need to make sure those carrying out face to face interviews have the skills needed to delve deeper and ask the right questions.

It may mean we see more of a "20 questions" type interview moving forward (something I've never been much of a fan of) but we'll just need to adapt our styles a bit to make it work.

In reply to Gemma:

Yes, I also think Christopher 's approach is a sound one to take at this time.

Incidentally, since the end of July we receive several comments from fake community profiles which use AI text generation tools which can be quite hard to spot (and remove). The clue is that the text is often rather bland and not in a human voice. This is a good summary I have snipped from somewhere (sorry - mislaid the source!)

In reply to Christopher:

Forgot to add - I've been thinking about using telephone interviews a bit more routinely (I used to do them, then had to stop because they were just too time-consuming) as part of my shortlisting.

It helps that the majority of my recruitment workload tends to be stuff that's administrative/customer service so telephone interviews lend themselves well in a lot of those cases and give the opportunity to ask direct questions about the application without the rigmarole of getting them in front of me.

In reply to Steve Bridger:

Sorry Steve - just spotted this. We've been phrasing our application questions along the lines of "Using specific examples from your previous experience..." which has helped a bit and, as Gemma says below, really focusing on probing questions at interview.

In reply to Christopher:

Use specific assessments (yes i know i would say that) and ensure they are ones that it is hard for an AI to deal with (for example simulated call handling using audio)

But equally especially for junior roles don't assume AI when there can be a simpler answer - when looking for apprentices, ones from the same school can have the advantage of a careers teacher who knows what you ask, helps several pupils with the application and so you get what is basically that teachers stock answer regurgitated by several people

In reply to Steve Bridger:

This is useful, Steve. The more I explore AI the more the consistency and repetition is obvious. What's quite concerning is its ability to generate quite plausible sounding references that are completely fabricated when fact checking.

In reply to Julie:

Julie,

Really the problem of AI generated fake references is no different from them writing themselves or getting their mate to write it.

27 years of recruitment showed me literally hundreds of fake references - I never trusted any reference unless I had also spoken to the person sending it preferable going via the main switchboard and confirming their position before transfer.

As so many are just a confirmation of dates of employment and title - they aren't often worth much anyway in the decision making process.

In reply to Ian:

Good point, Ian. I was a tad ambiguous as I was referring to academic references. If you fact check ChatGPT's output by asking for sources they look incredibly plausible yet are almost non-existent when checked.

In terms of job references, completely agree with you in terms of their usefulness.

Lots of people have their friends and family help with putting together their CV and cover letter, so using ChatGPT seems like a logical extension of this.

From my experience, though, as soon as someone opens their mouth during an interview you can easily spot whether they had help (or how much help).

In my time I have come across many candidates with amazingly written CVs or cover letters unable to provide a single coherent answer during a job interview.

Absolutely, you've touched on a crucial point. As we enter this new era of technology and artificial intelligence (AI), it's important to recognize that AI will play a significant role in our lives, becoming increasingly integrated into various aspects of work and daily routines.

The real challenge lies not in the use of these tools, which can undoubtedly help individuals communicate more effectively and efficiently, but rather in the genuine understanding of the purpose and the underlying knowledge of the subject matter. Technology should serve as an aid, not a replacement, for our ability to comprehend, analyze, and critically think about the information presented.

Building a solid foundation in the subject matter remains essential. While AI can assist in data analysis, language processing, and information retrieval, it's the human capacity for creativity, empathy, and ethical decision-making that will continue to set us apart. AI is a tool that can enhance our abilities, but it's our unique human qualities that enable us to navigate the complex world of ideas and concepts with depth and meaning.

In this evolving landscape, embracing technology while maintaining our commitment to continuous learning and a deep understanding of the topics at hand will be key to thriving in this new era.