A3i: The Kaizen Road to Transformative Progress
A3i: Incremental Artificial Intelligence Improvements
A3i's Kaizen-inspired Approach to AI Integration
Adopting new technologies is essential, and risky, especially the new world of Artificial Intelligence (AI). Many technology projects fail for organizations, resulting in frustration, delays, and budget blowouts. "Incremental Artificial Intelligence Improvements" (A3i) is an alternative to large projects that uses AI for what it is good for, making humans better at what they do. A3i is based on the idea of making small, continuous changes instead of big, disruptive ones. In this article, we will explore how A3i works, how it differs from traditional technology change programs, the personas that should be involved, and how it can help organisations achieve their goals.
Understand the Principles of AI
The principle of AI is to create systems that can perform tasks that normally require human intelligence, such as reasoning, learning, decision making, and problem solving. By applying AI to various domains, humans can do what they do better, faster, and more efficiently. AI can help managers decipher information, leaders make informed decisions, teachers personalise learning, farmers optimise crops, and engineers design smarter infrastructure. AI augments human capabilities, collaboration, creativity, and innovation.
The AI process involves creating, training, testing, and deploying algorithms that can learn from data and make decisions or predictions based on that data. AI can be divided into two types: narrow and general. Narrow AI is designed to perform a specific task, such as preparing a monthly report or predicting productivity. General AI is the hypothetical ability of a machine to understand and learn from any kind of data, and to perform any intellectual task that a human can do (e.g. ChatGPT or Bard.)
A3i is focussed on specific tasks in Narrow AI.
AI technology architecture typically consists of four layers: data, infrastructure, platform and application. Data layer is where the raw data is collected, stored and processed for AI use cases. Infrastructure layer is where the hardware and software resources are provisioned and managed to support the data layer and the AI workloads. Platform layer is where the AI frameworks, libraries and tools are deployed and integrated to provide the core functionalities of AI, such as machine learning, computer vision, natural language processing, etc. The application layer is where the AI solutions are developed and delivered to meet the specific needs and goals of the end users or customers.
The A3i Philosophy: A Kaizen-infused Evolution vs. Large-scale Organizational Change
At the heart of A3i lies the Kaizen principle, derived from the Japanese word for "continuous improvement." Kaizen emphasises the incremental betterment of processes, fostering a culture of adaptability and perpetual advancement. A3i extrapolates this principle to the integration of AI technologies. Rather than undertaking monumental overhauls, A3i introduces AI in incremental doses, amplifying the capabilities of human professionals and optimizing existing processes.
The conventional approach to technology adoption often involves large-scale organizational change programs. These endeavours, while well-intentioned, frequently disrupt established workflows, encounter resistance, and bear the risk of becoming overly complex and costly. A3i, in contrast, mitigates the pitfalls associated with radical shifts. It fosters an environment where incremental changes become the norm, minimising disruption, and maximising specific efficiency gains.
A3i is use case focussed.
A3i in Practice: A Strategic Blueprint
A3i begins by identifying opportunities for AI application in specific use cases to enhance outcomes, reduce effort, and elevate quality. This involves techniques like process mapping and interviews, promoting a transparent 'white box' approach. After shortlisting candidates, the focus shifts to evaluating AI implementation options based on factors such as cost, feasibility, and potential impact. The subsequent phase involves crafting a synergy between AI and existing workflows, ensuring effective integration and continuous improvement. This journey emphasizes iterative learning through the Plan, Do, Check, Act (PDCA) cycle, fostering perpetual advancement and refining operating processes based on real-time feedback.
Step 1 – Identify Applications, Define Their Characteristics:
A3i's strategic journey commences with identifying areas ripe for AI augmentation. Rather than pursuing grand overhauls, A3i homes in on specific use cases, where AI can enhance outcomes, reduce effort, or elevate quality by solving an existing problem. Process mapping, work packaging, interviews, lessons learned are the techniques to isolate individual opportunities for improvement in a system. Importantly, at this juncture of AI and humans there is a need for a ‘white box’ approach where process is visible for fear of excluding business processes because of an opaque system. The fear of opaque ‘black box’ AI can be mitigated by adopting a pedagogy of Human In The Loop (HITL) design in future stages. HITL also engages employees to work alongside AI instead of eing replaced by it.
At its conclusion this first step can generate a long list of potential use cases that can fuel an ongoing hit list of improvements around the business.
Step 2 – Evaluate Options for Implementation:
With potential use cases identified, the journey moves into a strategic crossroads. In this phase, options for AI implementation are evaluated with a discerning eye. Factors such as cost, time, feasibility, and potential risks come under scrutiny. This step prioritises AI applications based on their potential impact, and the organisation's current performance against competitors or best practice. The higher the impact, or poorer the performance, the greater the outcome. For evaluations that result in many options, a second pass evaluation of the short list using cost versus risk will prioritise the order.
Step 3 – Develop the Process:
With priorities set, the journey evolves into the process development phase. Here, the intricate synergy between AI augmentation and existing workflows is crafted. It's a harmonious blend of innovation and continuity. The process development stage ensures that AI is seamlessly integrated, enhancing operational efficiency while honouring the organisation's unique ecosystem. To do this well requires a diverse set of personas, presented later in this article, that operate as an A3i Team to collaborate and answer two important questions:
1. Are we doing the right thing?
2. Are we doing it well?
An effective approach is to breakdown the process into its transformational resources and inputs, the transformation process, outputs, and the customer (SICOP). This helps lift the veil on are we doing the right thing. Most organisations develop processes over time and focus only on how to do them better. Breaking from tradition is easier to do when you understand the parts of the transformation process. Like all masterpieces, it is always possible to improve how an organisation works. A3i takes this to the next level by introducing an unrivalled processing speed for incremental improvements.
An example application is reporting. Many business processes involve the interpolation of structured and unstructured data points such as financial data combined with status to produce a report for decision making. The use case occurs when automation of reporting is not possible. AI’s ability to disseminate and analyse the unstructured data and rapidly communicate through the large language models will astound most users of what is possible. When combined with HITL, the output is a faster production time and higher quality summarisation of issues that decision makers can rely upon.
The A3i Team’s job is to train the model on the transformation components and curate the output for the same people that had to do it before. Except now, there is less time processing and producing report content, and more time refining the message.
Step 4 – Plan, Do, Check, Act:
The implementation phase is characterized by the Plan, Do, Check, Act (PDCA) cycle, emblematic of the Kaizen philosophy. It's a dance of continuous improvement, where each step is followed by a thoughtful assessment of outcomes. The PDCA cycle amplifies iterative learning, with refinements informed by real-time feedback. It's an ongoing process that nurtures adaptability and fosters a culture of perpetual advancement.
Lean process focuses on delivering exactly what’s needed, when its needed. An A3i solution releases a process back into the business, while continuing to improve the process, as the business uses it. A good rule of thumb is 40-60% of the overall A3i improvement time is spent in this step, refining an operating process with user feedback.
A3i Persona Team
Applying A3i requires a balance of expertise that is not available in a homogonous business unit. Different use cases will require different expertise. Six distinct personas serves to address the multifaceted nature of the process and the diverse range of expertise. Each persona brings a unique set of skills, perspectives, and responsibilities for a coverage of all aspects of the AI implementation journey.
A description of each persona is provided here:
The Visionary Business Sponsor: This persona ensures strategic alignment and provides the necessary resources and support for the initiative. Their role is pivotal in securing organizational buy-in and fostering an environment conducive to innovation.
The Strategic Business Partner: Collaboration with stakeholders and alignment with the overarching business strategy are crucial to ensure that AI initiatives contribute directly to the organization's objectives and competitiveness.
The Process Maven - Business Improvement Expert: This persona's expertise in process optimization and business improvement is essential for identifying areas where AI can enhance operational efficiency and effectiveness.
The Technological Sentry - Technology Architect: With a focus on technical integration and data security, this persona ensures that AI solutions seamlessly fit into the existing technological landscape while maintaining data integrity.
The AI Artisan - AI Expert: Specialized knowledge in AI technologies is critical for designing, modelling, and implementing AI solutions that align with the organization's unique requirements and goals.
The User Advocate: End-user perspectives are invaluable in refining AI applications to ensure they effectively meet user needs and provide tangible value, making this persona crucial for user-centric development.
The collaboration of these personas creates a well-rounded team capable of addressing the various challenges and opportunities that arise during the A3i implementation journey. Their combined expertise ensures a holistic approach, driving successful AI integration and ongoing refinement for optimal outcomes.