Click here for the original LinkedIn article
Tomorrow, Wednesday September 3, Alexis Fink and I are leading a webinar on Driving Organizational Change with Data in the Age of AI. We welcome you to join us for a lively conversation on the potential and boundaries of AI to transform the work of analytics and organizational development (OD) – and the domains where human expertise will always be superior to AI.
This article reviews some of the arguments we will touch on, and provides resources for a deeper exploration of this and related challenges facing HR and business leaders today.
AI disrupts human tasks, but evolves human expertise
As some astute commentators have noted, many features of what we today call “artificial intelligence” have been around in a number of forms for many years. Prior to the launch of ChatGPT and the global explosion of people using generative AI models, we had machine learning and other forms of highly sophisticated computer-based modelling that provided many of the features of genAI behind the scenes, invisible to us end users. So in one sense, the current genAI models are like a coming out – or coming of age – party for a technology that now is clear to everyone.
Yet this current version of the technology is fundamentally different in key ways, including that it can make inference leaps using algorithms and machine learning which are more opaque than ever. We simply do not know the inner workings of these models. Which makes it really hard to forecast precisely how the technology will evolve.
Nonetheless, we can make some pretty accurate predictions about the limitations of the technology for the near term, at least for the next few years and likely many more into the future. Specifically, until and unless the technology reaches the point of actual sentience, the true artificial intelligence imagined by generations of science fiction writers, it will be severely restricted to operate in spaces defined by static data. Which means that it won’t be able to master the nuances of how individual humans, and organizational systems, operate in real time and space.
The future of HR and other organizational functions under AI
As AI takes over more and more lower-level cognitive tasks previously done by humans, the work of all corporate functions will evolve, potentially dramatically. HR, finance, legal, IT, etc. all will be transformed with many roles disappearing or evolving substantially from what they are today.
As noted in my recent article (AI’s impact on analytics, OD and Strategic HR), a central responsibility that will remain for the corporate functions is taking the outputs of AI in their domain, and applying them in the business. Which means focusing on how the organizational system integrates those outputs successfully or unsuccessfully – due to the actions of the other humans in the system.
Mastering those nuances of the people side of organizations will remain a bastion of human expertise in all corporate functions for the simple reason that genAI models will struggle to provide better advice than the best humans. But in order for the humans to maintain their competitive advantage over genAI, they will need to evolve their expertise to incorporate new insights created by genAI, and to maintain control of the tasks genAI struggles to complete independently.
This means that the corporate roles that interact directly with the business – such as the HR business partner (HRBP) – will evolve with AI rather than be obliterated by it. Yet the humans in those roles will have to evolve their own capabilities to stay relevant and be effective.
The role of analytics in org change as AI advances
As Alexis noted in her recent article (Uncertain Times Call for Analytics), in times of great volatility and uncertainty like right now, analytics is more important than ever for business-facing corporate functional roles – including HRBPs – to be effective. But rather than the common perception of analytics equating to high powered statistics, Alexis emphasizes the following main points:
- Analytics is a mindset, not a set of fancy math skills
- Applied the right way, analytics challenges you – and the people you work with – about the organization’s beliefs and assumptions
- The best way to use analytics is to approach the problem as a scientist trying to find evidence of what’s going on, rather than a lawyer seeking to prove a point
- So using analytics to address the pressing challenges facing the business means searching for evidence of root causes – which does not usually require fancy math
Statistical analysis of data using high powered models can and should still be part of the diagnosis where appropriate. Yet AI’s advances mean that developing and testing such models will become easier and easier for non-statisticians. Deep knowledge of the drivers of human behavior from the social sciences will still be critically important. But understanding those does not require statistics expertise.
Moreover, most challenges with business processes and operating model effectiveness are rooted in the work and organization design. And the type of analytics needed to diagnose those is qualitative – exactly the domain of expert business-facing functional roles. So the most critical tasks of diagnosing the sources of organizational under-performance, designing solutions, and implementing them will remain uniquely human – while the human skill sets needed will be enhanced as AI advances.
For more details, please join Alexis Fink and me for our webinar: Driving Organizational Change with Data in the Age of AI, and the virtual workshop we will be leading with Maura Stevenson, PhD by the same name Driving Organizational Change with Data in the Age of AI.
Details on doing organizational-level and systems analysis of the root causes of business under-performance using qualitative techniques are available in my two books Strategic Analytics, and Workforce Analytics (co-authored with Martin Edwards, Dana Minbaeva and Mark Huselid.