Are we ready for the AI recruitment revolution?

I haven’t written about artificial intelligence for a while, but with all the noise we’ve been hearing about ChatGPT recently, I guess it’s about time I got in on the action???. Will it totally transform how we communicate? Will a chatbot be writing these posts in future, while I sit back and enjoy a coffee? Or is AI just going to replicate and amplify our own errors and biases??

Who knows – at the moment it’s impossible to predict what will come next. For now, like many other people keen to see what all the fuss is about, I have dabbled with ChatGPT. In some ways, it’s very impressive. The ability to type a question about whatever you are interested in and get a pretty much immediate response feels quite engaging. In a way, it’s a bit like the chatbot experience I wish I could have when I contact companies today.

As for where it falls short? Well, one of the most obvious downsides is that the system’s knowledge base only goes as far as September 2021. But as the technology advances, it doesn’t seem impossible that one day it will be able to hoover up and spit information back out at us in real time. More of a problem is the fact that it is not always accurate. Not such a big deal for me to be told I work for a Dutch competitor of T-Mobile Netherlands (the very idea…), but for the Australian mayor?falsely accused?of serving time in prison for bribery, well – I can see why he took legal advice.

While my own experimentation has been a bit of fun, really I’m more interested in what AI can do for businesses, and whether it will ever be capable of carrying out tasks that we assume need the human touch. Let’s take recruitment for example. Does anyone, anywhere have a 100% track record on getting it right? I certainly don’t, and I have been hiring people for more years than I care to think about.

I use the “Five E” concept for external hires, which probes experience, education, energy levels, empathy and ethics. The last two – so-called soft factors – are both the most crucial and difficult to nail during an interview. References can help here, along with feedback from the other people your interviewee encounters, like your company’s security people, reception staff or personal assistants.

Despite all this, if I’m honest I’d say I haven’t managed to get quite the right fit between the candidate, me and the rest of the team maybe once in every five or six times. Probably the worst example was the guy who had good energy and education, but was clearly lacking in the ethics department, as he was eventually caught helping himself to the company’s money…

So, could AI help me make better hiring decisions and avoid future mistakes? Let’s take a look at some areas where it could add some value.

• Scanning the job market to help build more diversity into your workforce
• Seeking out evidence of the right skills, achievements and attributes, for example past performance and behavior. Our digital footprints say a lot more about us than we might think.
• Doing a deep analysis of CVs and cover letters?

The feature I’d love most however would be one that predicts whether a candidate is truly interested in taking up a new opportunity. My most frustrating experiences, hands down, have all been when I found a fantastic person who then changed their mind (or used the offer to get a better package from their current employer). On the other side of the coin, I’m sure jobseekers would be interested in finding out what they should be looking for and any opportunities to avoid. It would save time and effort for everyone.

As for the minuses, I think cultural and team fit will be hard for AI to predict because what we’re really talking about is chemistry, which is deeply complex and personal in so many ways. I also have no doubt that smart candidates will soon learn some tricks to get past inconvenient AI roadblocks. And while your company might benefit from this kind of innovation and nerve, you probably need some more reliable individuals in there too.?

Equally, I’m not sure whether AI could accurately and reliably predict a person’s potential. This is so important, but is impossible to judge based on a past track record. Some of my best hires have been individuals who had limited experience of what I was looking for, but had the energy, confidence and curiosity to take on a completely new challenge. And in fact it is the leaders who gave me the chance to stretch my own capabilities who enabled me to become who I am in business.?

On a similar note, when are the right boxes really the right boxes? We live in a world where attending a particular school or reaching a specific level of educational attainment is more or less a fast-track into certain jobs and firms. But at the same time, some of the world’s most successful people (Steve Jobs, Michael Dell, Bill Gates, I could go on) either didn’t go to college, or didn’t complete their studies. Without the right programming, your super-efficient AI recruitment assistant could lose you some really excellent people.

As for me, if a CV-scanning bot had pushed me out of the pool because I didn’t meet all of the prescribed criteria, who knows, maybe I would still be serving pancakes in a Dutch restaurant…?

All in all, it’s a very interesting and challenging area, as a recent article in the?New York Times?describes. Just as progress on even more powerful and accessible systems than ChatGPT is ramping up, regulation is coming to meet it. In New York itself, a new?law will require companies using AI as part of their recruitment process to disclose this to candidates, and to have the system independently checked for bias. And across the pond in Europe, the planned?Artificial Intelligence Act?has designated hiring as a “high-risk application”.

It's not yet clear whether the?open letter?calling on AI labs to press pause on their efforts to go even further with the technology will have the desired effect. But there’s no doubt in my mind that if we are to make AI work for us, whether it’s in hiring or any other aspect of our daily lives and businesses, that we have to carefully plan and control its impact.

Let’s see what happens next.

Photo credit:?jittawit.21, iStock