Every technology was once the technology that would change the world.
Eighteen months ago, AI applications such as ChatGPT engaged the public’s imagination and fears. Today, AI continues to ride the enthusiasm that comes with the peak of inflated expectations, but the shadows of the trough of disillusionment are deepening.
The history of technological change gives us two lessons.
Lesson One. The technology’s developers, evangelists spruiking it, and end-users all have different views on what it does. What the technology delivers in practice is rarely consistent with these views.
Lesson Two. The growth in the use of technology and its value to our lives is not threatened, slowed, or stopped because we become bored or the technology is no longer useful. The failure to make the most of technology is generally a failure of effective implementation.
In our digital economy, new technology will continue to present us with wicked implementation problems that extend well beyond the enthusiasms of developers, evangelists, and users.
Today, the ethics of AI are problems for policymakers and leaders. For example, the rapid spread of AI has increased the computational power required to train AI programs, increasing power consumption and putting unwanted strain on energy grids. The many benefits of AI come with, often unforeseen, local and global costs.
Technology adoption has always been a wicked problem for leaders, just as strategy myopia has been our Achilles heel.
Planning the ‘wicked’ out of wicked problems
By definition, wicked problems are highly resistant to resolution. Yet strategists and planners seem intent on removing the ‘wicked’ from them. Unfortunately, the wickedness of wicked problems lies in the interactions between causes, policies, politics, and stakeholders, which are often conflicting but also sometimes resonating.
The traditional business planning process involves working systematically from problem to solution through evidence gathering, data analysis, and stakeholder engagement, developing a preferred option supported by a detailed implementation plan. It often addresses a symptom of the problem and takes too long. This approach divorces the problem from the context, which is the very characteristic that makes it wicked.
Our planning mindsets and tools are not well suited to unfolding patterns of socio-technical change where the answer is not optimisation.
Digital transformation starts with technology
We all make the mistake of thinking that technology is a thing or a collection of things. It’s not. There is always an intent behind technology driven by the question, ‘How can we make it better?’, where the ‘it’ is quite specific. Consequently, technology inevitably carries a design to organise knowledge, people, and assets to achieve a precise goal.
Three characteristics separate digitisation generally and AI specifically from classic technologies. These characteristics inherently introduce wicked problems and require systems solutions.
First, they are modular. The technology is complete in itself but also stackable. Stacking creates layers that enable rapid recombination that creates new outputs and opportunities. The recombination speed is increasing, and the depth of the layers is becoming opaque. Modularity has been an engineering principle for many years; now, it is a wicked problem.
Second, digital technologies are generative. The ability to transform and reuse information as content has been an innovative boon and the source of social media disaster. AI has taken technology’s generative characteristics to new places, leading to problems in AI ethics and embedded biases. Most recently, those looking for an advantage in AI products are exploring how AI models can be used to improve another AI model. This takes technology to new, weird, and wicked places.
Third, there are people. This takes us back to lesson one. What technology delivers in practice is rarely consistent with what the designers think it will do because people interact with it. People interacting with technology is a wicked problem.
Strategy and planning in the wicked world of technology implementation should begin with a deeper understanding of the systemic effects rather than controlling its more obvious outputs.
The most basic question about new technology is how today’s ethical values and norms will apply in the world the new technology creates. All business strategies and plans follow from there.
Where does that leave planners?
The dynamism of technology and the uncertainty it creates may leave strategists and planners grasping for relevance. However, lesson two hints at where the effort should be applied. It requires a shift in mindset and language from ‘strategic planning’ to ‘strategic implementation’.
The objective is not to produce a perfectly articled strategy accompanied by a detailed implementation plan. Rather, the challenge is identifying and focusing on the core human questions and responding.
Planning and action begin with a clear-eyed understanding that policy, strategy, and implementation are blurred, interacting, and interdependent. Implementation is an ongoing evaluation of the context and issues that arise from interaction. Planning involves determining what system-level changes will be needed to capitalise on the most immediate opportunities and prevent the most pressing harm to people and performance. Options are alternative system-level courses of action that transcend or circumvent strategic risks as they are known today or immediately foreseeable.
Finally, there should be a genuine need to learn from errors rather than expect to find all the answers in advance. This has implications for the follow-on systems of assurance and quality, which are inherently underpinned by the principles of linear management models seen through a rear-view mirror. Retrospectively assessing ‘failure’ is a contemporary challenge for all organisational leaders wrestling with wicked problems.
Strategic implementation is a systems approach
The difference between traditional problem-solving and working with wicked problems has been likened elsewhere to the difference between throwing a stone and throwing a live bird. The stone’s trajectory can be calculated precisely using the laws of physics, while the bird’s trajectory is far less predictable.
AI is the latest technology to show us that the world is less predictable than we want to believe. However, as we address the issues, some handholds are firmer than most.
- Leadership remains fundamentally important. Not the process maintenance leadership that has come to dominate our organisations but rather leadership that is demanding, active, and constant in its commitment to the overriding strategic priorities of the organisation.
- People are a constant. Not the process and paper of people but a genuine commitment to understanding that the interaction between people and technology goes beyond automation. Performance comes from systems where technology augments human capacity and capability.
- Independence must be thoroughly built into the strategy and planning system. The decision-making environment should allow dissent to identify bias and magical thinking wherever possible.
- Governance must be simple. Structures, processes, rules, and procedures must be as simple and straightforward as possible so that everyone can understand them. Today’s organisations are hamstrung by process complexity, which undermines agility.
READ MORE:
Defence staff used ChatGPT thousands of times without authorisation
