There is increasing concern about the use of ChatGPT style agents creating job application content. Recent reportsdescribe how business owners are struggling to differentiate applications that have been generated through AI and those that have been crafted in more traditional ways. The legitimate concern, of course, is that people can “fake it but never make it”; as once they start the job the AI cannot do the job for them – not just yet anyway!
However, AI is not a panacea for all problems and the same business leaders are spotting trends and patterns in AI usage. For example, when applicants are trying to use AI to “leverage their skillset”, the chatbots are producing generic statements like “my skills align with your organisation’s objectives and goals”. As a result a tit-for-tit war of attrition will likely ensue with application parsing software trying to identify AI-generated responses. Before, presumably, the next generation of AI will generate content that can not be identified by AI. And so the battle will go on ad infinitum.
Of course, these are not the only applications where AI is a cause for concern. Some high profile examples of disasters include;
- A legal firm using ChatGPT to research precedents in a suit against the airline Avianca. However, the model hallucinated six previous cases that did not exist.
- A chatbot that encouraged business owners to break the law by encouraging them to take a cut of workers tips, fire workers who report sexual harassment and serve food that had been eaten by rodents.
- McDonald’s scrapped an AI experiment, three years in the making, allowing AI to take drive-thru orders after videos of confused and frustrated customers were posted on social media.
The impact and progress of content generated by AI has been staggering, but so has our over reliance on it. Even for Fortune 500 companies, over 10% of blog content has been detected as potentially AI generated. While AI can be a useful tool, its uncritical adoption in job applications risks undermining the integrity of the process, leading to mismatches between skills and roles, and ultimately, harming both individuals and organisations. At the macro level it can lead to the types of unmitigated disasters highlighted as above.
How AI Can Be Used For Good In Job Applications
But AI can be used for good in job applications. At Skillsminer, we help our users write CV’s using AI – but with a difference – any content we generate is welded to our skills and occupational taxonomy. It’s a form of RAG modelling, but we won’t go into this here. This vastly reduces the risk of hallucinations, ensures users are not left confused or frustrated by the outputs and encourages users to utilise skills they have already validated.
This serves as a benefit for both applicants and businesses. Firstly, writing a CV from scratch is not easy, many people suffer from writer’s block or the objectivity required to write about their skills and experiences. Using a system that prompts the user about these hidden skills is empowering. We always encourage people to refine the content, as no matter how good the AI, it will always need human input. And for businesses, the war for talent is real, and uncovering hidden talent that may otherwise go under the radar, both internally and externally, reduces the pressure on the demand valve, which in certain sectors is at bursting point.
AI is still in its experimental phase. So, I thought I’d use this article as an experiment. It may be obvious reading this article that I am not a professional writer. So, I used AI to both “rewrite the full article in a blog style” and to write from scratch “a blog on the use of AI in job applications”. The TLDR; the power of content generated AI is staggering, but, even when refining an existing document, it is laced with hallucinations (for example, assuming that the phrase “fake it but never make it” is a real phrase – it’s actually something I made up for this article). But when it is used unchecked, it is misleading, incorrect and vague. And that is dangerous in all applications not just job applications.
So, my advice, if you are going to use AI in your job application, make sure you validate its outputs, or, better still, use a tool like Skillsminer, but then I would say that. It’s the same message for businesses, harnessing the good that comes from AI applications in uncovering hidden talent can help you win the fight for talent.