Springtime for Process Managers
Process management, a discipline that emerged in the 1970s and 1980s, sought to bring the principles and techniques from science, engineering, and manufacturing to the realm of knowledge work. The core idea is to map out all productive activities in meticulous detail, enabling the analysis and improvement of each step in the process and their interactions. Despite the clear appeal, and billions flowing to consultancies that promise help with implementing it, process management has struggled to gain widespread adoption. Many contemporary organisations either fail to adopt it completely or apply it in a manner so superficial and abstract that it becomes meaningless.
To a naive observer, it may appear irrational for organisations to ignore process management, as it would seem to offer productivity enhancements. While there is some truth to this perspective, there's more to the story - the primary constraint on knowledge work is cognitive capacity. Historically, humans have been the sole source of this capacity, and they tend to resist tight management. Therefore, companies have often opted to focus on recruiting capable individuals and avoiding excessive management interference. The guiding principle for some of the most successful organisations of the past decades has been to "hire the best people and get out of the way." This approach is rational, if not getting out of the way means that you wouldn't be able to hire the brains in the first place.
How things have changed ...
Recent advances in artificial intelligence have given rise to models with remarkable cognitive abilities. While these models cannot compete with human minds in terms of flexibility and generality, their utility is undeniable. For instance, GPT-4, a leading large language model, may lack sophisticated planning and symbolic manipulation capabilities, but its text generation prowess surpasses that of any human. Similarly, image-generation models like Stable Diffusion or Midjourney produce images of exceptional quality at unprecedented speeds, despite lacking an understanding of the requests they process.
These AI models offer immense potential for productivity improvements, but unlocking that potential is challenging. AI models cannot replace the entire human mind. Individuals who effectively use AI for productivity enhancement follow a common pattern: they break down their work into smaller tasks and delegate specific tasks to AI models. This approach is, in essence, a form of ad-hoc, on-the-fly process management that does not require coordination with others.
The issue lies in the fact that much of the knowledge economy is driven by collaborative work much more than individual work completed by a single human. And when it comes to collaborative work involving multiple people and teams, AI has yet to make a significant impact on productivity. For that reason, it is likely that for years to come AI technology will continue to demonstrate impressive abilities in demos and narrow implementations without causing the transformative shift in the economy that many anticipate.
Eventually, though, competitive pressures will push organisations to explore ways to utilise AI for productivity gains. To achieve this, they will need to implement detailed process management across the organisation. While this may initially find greater acceptance in less individualistic cultures, it is likely to become a universal trend.
As we move towards this new reality, it is essential to consider several practical implications:
Get used to being a "node" in a process network, the proverbial cog in the machine. Don't take it personally.
Develop your process management skills and seek opportunities to map out processes in your organisation and identify opportunities to integrate AI into your organisation's workflows.
Tools and methods that help organisations manage processes better are going to become increasingly important. That's a promising area in business to invest in.