Hi Dave,
Trigger is one of TV’s great sitcom characters, brilliantly played by Roger Lloyd Pack. Initially meant to be a bit-part one-off character, his role grew to be a beloved mainstay.
A street sweeper by trade, he had the same broom for 20 years:
“This old broom has had 17 new heads and 14 new handles in its time.”
Well, then it’s not the same bloody broom, is it?
Except Trigger has a photo of it 20 years earlier, proving it is the same.
A much easier philosophical debate than Ship of Theseus from a couple of weeks ago.
There’s much discussion about how AI is replacing parts of industry, in a bid to scale, find efficiency, reduce cost, and let’s face it, get rid of those pesky humans who do things like ask for pay rises, have babies, or change jobs.
Part of this is marketing-driven, proposing how AI can improve and replace manual processes.
With the caveat that AI isn’t intelligent, it’s just complex automation.
We’ve seen this before, with trains replacing horses, robotics replacing manual labour, ERPs replacing clerical work, streaming replacing Blockbuster first and looking like Cinema increasingly.
Except the argument is this won’t happen in recruitment - ‘we are a people business, where relationships are key.
Job boards didn’t kill off agencies.
LinkedIn didn’t kill off agencies.
And AI won’t rip through the recruitment industry wholesale’.
The argument that AI isn’t a replacement for recruiters is predicated on two notions:
Where AI is right now
that people and relationships are important in recruitment.
On the first point, does it matter where AI is right now?.
Or is trajectory more important?
Look at the rate of change in the quality of AI output every year.
Just a couple of years ago, we were laughing about extra arms, weird teeth, and ears for eyes.
Now, videos of Will Smith eating spaghetti look a bit weird, while we package ourselves as toys.
Flashforward a couple of years, and will that improvement be linear or exponential?
Recruitment might be different to media, but what’s in common is parsing and access to data.
(It’s a whole different article about the ethics of scraping private data, but we know it happens - many buy into services that do exactly this, whether it’s Ghibli or CV)
Chatbot agents providing real-time answers on the process to hiring managers and candidates
Automated feedback based on parsing of CVs, interview transcripts and hiring manager debriefs
Automated workflows - briefings, job descriptions, adverts, employer information, CV analysis (and why not full contextual analysis accessing datapoints like other employee profiles on LinkedIn, news, media, annual reports, etc), offer letters, contracts, preboarding, onboarding
Automated sourcing and outreach
Automated updates - new vacancies for relevant previous applicants, relevant previous applicants for new vacancies
Some of these things are ready, some are coming, more need to be invented (what about on-the-fly psychometrics based on available data?), and they will improve quickly.
Great features - how will they benefit the recruitment process?
Future potential is as much of interest to buyers as what it can do now.
If we agree that the first point, “state of AI”, isn’t relevant to the discussion, then what is left are people and relationships.
I’m a strong advocate for looking at problems as opportunities, and I’ve learnt a lot about how I can improve my work from talking to and trying to help long-term job seekers.
They’ll tell you their issues with recruitment, both agency and employer-side, are these:
lack of transparency
vagueness- comms not giving insight into what jobs actually are
ghosting
pushy comms
unreasonable recruitment processes
low-ball offers
a feeling they are a number in a lottery
Speak to employers and recruiters, and I expect they’ll report much the same with candidates and agency recruiters.
When I come across these issues, I’m keen to understand their root, how I may be guilty of the same, and what solutions can improve experiences for everyone.
Of course, these problems aren’t the whole industry, but we need to consider that candidate resentment continues to rise, while there are many employment areas with too many capable job seekers, and not enough vacancies.
What would happen with these experiences if the AI-style solutions replace people?
Particularly for business-as-usual, common-skill, or vacancy-short roles, where the job just needs to be done, and candidates are plenty?
We can expect many people to be turned off by a human-limited process (not humanless if we assume interviews will still be roughly the same). But how many will accept a human-limited process, and will that volume fill vacancies suitably?
Might their experiences even be better than through a human-employed recruitment process?
And if experiences prove to be better on the whole and at scale, what does that say about the importance of relationships in process (not in employment, if that continues to exist after 2029, ha!).
I think it comes down to how important relationships really are.
I’ve written recently that one of the elements I most enjoy in my job is coaching candidates and hiring managers. Part of this is ‘how to do better’, whatever that might mean in context. Much of it is about helping find black swans.
I see every interaction as an opportunity to build trust.
These micro relationships (ones which often go on hiatus on rejection or a few months into their new job) are both enjoyable and an area I differentiate.
For key hires, where context is king, and the requirement specific, building relationships, trust, and experiences will continue to be a priority, because of the outcomes they bring:
better fill rates
better retention
feedback hires deliver above expectations and are often future leaders
reduced risk
more fulfilling
and many more
It’s an approach I’ve seen to be successful in medium-volume and business-as-usual work. But the investment it takes from me makes these less profitable per-hire.
So while it may work for me as a micro-recruiter who wants to simplify, it’s harder to build into a systemised business that wants to scale, where that focus on individual relationships might even get in the way.
You can see the same impact in e-commerce vs retail.
For high-stakes items, buy from people you trust, where you can see it, hold it, feel it.
For low-stakes, cost-centric items, I’d buy through a carefully designed automated online platform that has a reasonable customer experience.
For a couple of decades, while retail floundered, e-commerce thrived. Jobs change, with new opportunity.
Will AI replace recruiters? Possibly.
I treat it as a form of Pascal’s Wager. It can happen, and the consequences would be severe, so it should inform our behaviour today.
Double down on what makes our recruitment great for the people we serve, and leave them with our best experience.
Anyhow, next week I’ll argue against AI again and see what comes out of that.
Thanks for reading.
Greg
p.s. I’m enjoying Black Mirror with my youngest, by the way.