The FT’s Andrew Hill is a thorough journalist, so I was interested to read his recent piece asking, “Does HR still need humans?” This is an intriguing question. Much of HR is low-level transactional. So, looked at in a certain way, the answer is: No, HR no longer needs humans, or soon won’t.

Meanwhile, on AI, the directive from the C-suite is clear and dual: equip the workforce with AI skills and reduce the workforce by introducing AI. As an aside, Gallup’s 2025 State of the global workplace report points to flagging employee engagement. Could it be (in part) that employees are not really ‘into’ working on projects that are seeking solutions that will replace them? I cannot think of many endeavours that would be more depressing. Can you?

The US Census Bureau says that almost 10% of US businesses are now using AI in the production of their goods and services. Most businesses are yet to incorporate AI into their processes.  

Elsewhere, a banking leader whose department work to counter fraud says she expects headcount to fall even as the business grows. But what if fraudsters find a way of gaming AI-driven processes and the bank lacks the skilled human power to pivot and fight back? After all, it’s not like we haven’t had an ‘all-hands-to-the-pump’ moment in recent living memory. Covid taught us how important it was for people to work together and be adaptable.  

Are we saying things would be more stable with AI embedded in systems during a Covid-level event? I suggest the opposite: that AI would get in the way as humans scrabbled to find new ways of living and working. Many scenarios are possible. When the fever dream of the everything machine breaks, investors will ask which are more probable.

The FT article outlined five learnings, all of which I agree with and briefly summarise here: 

1. AI implementation must be handled strategically ie AI provides the ideal pretext to fix broken, outdated or inefficient processes. 
2. The lowest hanging fruit are in support and administrative functions such as the automation of answering simple, high-volume queries. 
3. Senior executives must be engaged and participate in AI rollouts.
4. AI blurs the boundaries between functions, surfacing previously unseen inefficiencies.
5.  AI demands changes in how work is done.
    

Number four is perhaps the most interesting, and where the most productivity gains are likely to be had beyond merely reducing the workforce. AI will surely make it easier for departments to work together and see patterns and possibilities anew with so-called ‘vibe analytics’. Although, might I suggest a better name for that: Analytics! True analytics were often just too difficult before, due to differing cultures, siloes, competing terminologies and technical needs. So, it’s here we could see real organisational changes and gains with machines helping humans to get beyond their tribal limitations and set the stage for their ingenuity to flourish.

To summarise:

Does business need HR? Yes, or something like it.

Does low-level transactional HR still need humans? Hardly – though humans will have to be on hand when there are qualitative shifts in demand.

Does strategic HR still need humans? Yes.

Does business need AI? Yes, but only in a minority of use cases so far touted.  

 

Image license:
Yutong Liu & Digit / https://betterimagesofai.org / https://creativecommons.org/licenses/by/4.0/