top of page

AI Adoption: What Your Change Management Framework Should Take Into Account?

I don't think I've ever seen a more elegant theory than the Theory of Constraints (TOC) by Eliyahu Goldratt. The principles are so simple and yet so powerful that whenever you discuss how systems work you end up seeing TOC written all over it.


Understanding TOC makes you high-octane when you need your system to change and the change is dramatic, like AI adoption, for instance, and its impact on ways of working, interacting with each other, interfaces, and data.


And here's why.


A Theory of Constraints Perspective: Applying AI to Your Bottleneck


According to ToC, the throughput of any system is determined by a single constraint. Here's an illustration of your workflow with every square being a task or a team or a person that pushes your value through its delivery pipeline.


The capacity of each task/team/person is different - so is the size of the square. Each task/team/person is different - so is the color and pattern of the squares. Some of them happen in parallel, some overlap, but all need the previous step to be complete before the value moves on to them.


Bottleneck in a workflow by Theory of Constraints
Bottleneck in a workflow by Theory of Constraints

And there's always this one box with the lowest capacity that turns into a blocker keeping the throughput at its lowest on its stage. When you think about adopting AI, or any other technology for that matter, this bottleneck is the most obvious place to apply it in order to increase the throughput and potentially accelerate the delivery.


But here's how AI is different from other technologies and tools.


AI has the potential to increase the capacity of the blocker so much that it turns into a reversed bottleneck, wrecking the workflow before and after it.

Reverse Bottleneck in a Workflow by Theory of Constraints
Reverse Bottleneck in a Workflow by Theory of Constraints

Optimization that is focused on isolated segments of the workflow is a way to fail when adopting AI - you need to consider system-wide implications.


Even if it's just your people randomly using ChatGPT on their own, their ways of working are never going to be the same.

How AI adoption influences Way of Working


So what can you do to start seeing that helicopter view?


AI Adoption Through the McLuhan Tetrad: Planning for Effects


Marshall McLuhan's media effects tetrad offers a great starting point for the teams to look at a bigger picture and map out the transformations AI might cause:


AI as a medium in Marshall McLuhan's tetrad


  • Enhancement: AI augments human capabilities, automating routine tasks and analyzing vast datasets, thus enhancing our decision-making processes.

    • What are the routine tasks that eat away most of your time and feel like a waste?

    • Where do you have the biggest lack of insights?

    • What data analysis requires too much time and effort and thus is never actually done?


  • Obsolescence: Certain roles and skills, particularly those involving routine, predictable tasks, will become obsolete in the face of AI's efficiency.

    • Does it mean you will have to reskill or let go certain number of people?

    • How do you encourage creativity and learning to take place of the mundane activities?

    • How do you communicate these changes to your customers?



How AI adoption influences Tech Stack


  • Retrieval: AI revives previously sidelined analytical methods by applying advanced algorithms, bringing historical data into a new light.

    • How do you ensure that your old data is clean enough so you don't have "garbage in - garbage out"?

    • How do you check you data for biases?


  • Reversal: At its extreme, an over-dependence on AI can lead to a diminished appreciation for human intuition and judgement, potentially reversing the value we place on human expertise.

    • How do you prevent your employees from "falling asleep at the wheel" and seizing to check the AI-generated results?

    • How do you leverage tribal knowledge and company specific expertise when using AI tools?

    • What will your system of counterweights be to maintain the balance between human an AI?



How AI adoption influences Value Delivery Flow


This approach can help you frame the "philosophy" behind AI adoption outlining the overarching attitude your organization has towards the technology and its implications.


But then you need to communicate it to everyone on the team to motivate and engage them into the transformation. How do you do that?


Human Adaptation to AI: Insights from Dr. Travis Proulx


Current reality, already being transformed by AI, is not the world we've expected to find ourselves in just a year ago.


We are in a state of constant cognitive dissonance every time we deal with AI because it transforms the workplace in unexpected ways and at a breakneck speed.

Most of your employees will freak out.


They are afraid they will loose their jobs, or that they become completely irrelevant in the new world, or that it will be hard for them to reskill and they will loose their expert leverage. All of those fears should be addressed carefully and with a lot of patience because change (especially the one you can benefit from) can only happen if there's enough employees supporting it.



How AI adoption influences People Management


A few years ago Dr. Travis Proulx from Cardiff University has researched human responses to inconsistencies - or basically how we respond when we face something that violates our understanding of the world.


These five primary methods humans use to resolve inconsistencies that were described by Dr. Proulx can help you design the approach to accommodate for all of the compensatory behaviors and mitigate risks of AI adoption failure:


  • Assimilation:

    • Adapting AI into existing frameworks without causing disruption,

    • Introducing upskilling programs to improve AI literacy,

    • Enabling teams to use AI features in their core tools.

  • Accommodation:

    • Restructuring organizational frameworks to integrate AI capabilities,

    • Compare and contrast the outputs from traditional approaches and AI,

    • Train teams to differentiate where AI tools are best for the job and what are the limitations of the technology.

  • Affirmation:

    • Reinforcing the organization's core values amidst the transformation,

    • Aligning usage of AI with your company's ethics.

  • Abstraction:

    • Developing new working principles that align with AI's advanced capabilities,

    • Establishing new contracts with employees, partners, and customers that take AI adoption into account.

  • Assembly:

    • Constructing new systems that include workflows, tools, policies, and people to fully leverage AI's potential,

    • Becoming an AI-first company.


Each of these methods can help you design a roadmap for smoothing the transition into an AI-centric work environment, ensuring that human workers and AI systems coalesce into a harmonious and productive entity.


And I can help you with the narratives to communicate the changes both internally and externally. Let's talks!



コメント


Let's Talk

Thanks for submitting!

bottom of page