AI has been widely adopted; however, there have been limited instances of high corporate value being returned. The consensus is that we are currently in the “piloting phase” (early adoption) and heading into a phase of scale. This is a similar pattern to technology-based platforms of the past.
Like any platform, there is a wide range of skills, disciplines, standard approaches, and controls that are needed for broad-based value to be generated, with some skills being related to mastering the technical platform and far more skills being mastered in organizational change. However, AI is a platform that is evolving too rapidly to plan or execute gradual adoption which is different from past platforms.
AI Platforms Require Skills and Controls
AI is a platform offering enormous value but requires learning new skills, approaches and change. Past platforms (mainframe, PC, Internet, mobile) all brought new skills, processes and approaches and that is also true with AI.
- There are many technical skills associated with AI which include AI platform specific orchestration, data management, prompt engineering, security, integration with other systems, multiple-platform management and orchestration.
- There are change management skills (like any other operational change) including process redesign, workflow redesign, openness to change, continuous incremental change mangement, training and incentive definition.
- To achieve AI value, both types of skills must be learned and brought together within an endeavor.
- The more individuals or groups involved, the more complex the change will be required. To achieve scale, there must be standardization within the organization that requires discipline, controls, and coordination around AI-related initiatives.
The above dimensions are indicative of a classic learning curve an organization must go through to move from individual “champions” making initial use of a technology to broader adoption by teams and finally to enterprise wide where technology is part of operational fabric. With AI we are still in the individual champions and small team adoption phase that results in productivity of the individual or team as they experiment with new approaches.
AI Learning Curves Within and Enterprise
AI Competencies Build from Individuals to Enterprise
AI will progress from individual productivity to team productivity and finally to enterprise value as technical skills are achieved and paired with operational change skills.
- Individuals and early adopters must experiment and pilot to learn the required technical skills. These usually result in individual productivity gains without broader process operational change.
- To gain departmental or team value, individual technical skills must be absorbed and adopted by the team with some level of standardization. This can be departmental productivity and/or point process transformation.
- Organizational transformations are only possible with multiple departments and/or point processes are each supplemented by AI and orchestrated as part of broader endeavor. This requires the learning and AI applications to be done across multiple teams and the orchestration to be done from a top-down perspective.
This incremental approach to adoption is depicted in the following diagram as the scale and complexity of the change grows for and individual to the enterprise.
Framework for AI Adoption in Enterprises
AI Value will be tied to Controls and Discipline Across the Enterprise
AI will evolve as learning of new skills and experimentation will lead to narrow use cases and longer-term operational transformation. Each incremental will also need more discipline, standards and controls to support the movement from individual to scale.
- Short term the focus must be on attaining skills (both technical and change management). Limited controls on individuals to encourage experimentation and adoption.
- Adoption for individual and team productivity has lower organizational value but is a necessary step to build skills and identify use cases. Standard AI platforms, tools, security, techniques and data management are introduced within the team.
- Point solutions (customer service, inventory mangement, credit risk management) bring higher operational value as AI supplements and accelerates specific business processes and gives broader skills and experience to the organization. More rigorous security, data mangement and adherence to standards are required. Financial management and metrics are also introduced to monitor impact.
- Business transformations will be achieved only after individual and process specific success with AI has been experienced but will have to be executed as part of strategic endeavors driven by the executive team across multiple functional areas. Cross functional change requires yet more rigour in security, data management, operational management and financial management. Incentives tied to overall organizational productivity (rather than department) are also required to achieve unified operational discipline.
- Personal productivity and point solutions have limited scope and individuals involved and are typically initiated and executed on a project or program level (bottom up). Transformations reach across the organizations and must be initiated and coordinated by the executive team (top down) but executed in smaller chunks (bottom up). Continuous change management requires alignment and coordination of executive and operational management with governance processes that frequently review progress (metrics), realign vision and identify next target for incremental improvement.
The Prometheus Endeavor Digital and IT Management Team develops useful frameworks and insights to help organizations master transformational change.
One Response
Excellent food for thought, Pat. The necessary changes organizations will need to make to transform and ultimately reach an enterprise-wide, business-integrated operational level will be significant and costly. These will, in turn, require powerful motivation and justification. Visionary thinking and strategic advantage are top of mind, but other external factors are (have been) just as compelling. For example, consider the pandemic and the subsequent global supply chain disruptions it caused. Consider these examples that, as a Phase 1 adopter, I got from ChatGPT:
“Two of the strongest examples of rapid enterprise transformation were Amazon during its shift to cloud-native operations, and the widespread enterprise digitization that occurred during the COVID-19 pandemic across companies such as Microsoft and Shopify. In both cases, strong external pressure forced organizations to adopt new operational platforms with centralized coordination, standardized tooling, and executive sponsorship. Amazon standardized APIs and reorganized teams to handle the explosive growth of internet commerce, while pandemic-era organizations rapidly deployed cloud collaboration, digital workflows, and remote operating models to maintain continuity. In both examples, enterprise productivity gains came not just from technology adoption, but from coordinated organizational change and governance.”
These are concrete examples that may help define the “how to.”
Gonzalo