The Siren songs of AI: The Innovation and FOMO bias
I May Be Biased
I recently read in a BCG survey that 74% of all AI initiatives fail to reach the proof-of-concept stage. In a recent MIT study, only 5% of those initiatives that reach proof of concept actually deliver realized ROI. Think about what this means for all the organizations that are building AI for the future. How much time, energy, and resources are put towards the initiatives, not to mention the planned savings, the technology savings are all being forecast and built into operating budgets for the next few years. Several years ago, I regularly conducted Design Thinking workshops in which participants had to innovate and solve a problem. For a time, every solution and prototype was an app; “there was an app for that”. It became so ingrained in the participants' minds that we had to exclude building an app as a solution. Today, I see similar conversations and expectations regarding emerging AI technology.
A combination of biases is at play in our quest for AI transformation. The Innovation bias, or shiny object bias, and fear of missing out (FOMO) bias, also known as Social comparison anxiety. Together, these biases become an enchanting siren song of transformation, drawing us closer to the rocks. But what are we supposed to do? Stand by and let other organizations reap the spoils of transformation and cash in on the promise of AI? Should we be like Odysseus in Homer's Odyssey and tie ourselves to the mast to keep us from engaging in AI transformation?
The solution lies in betting on technology and applying a systems-thinking approach to transformation. This system involves three essential pillars that the technology must be embedded in: strategy, process, and people.
The strategy:
Every organization must ask itself a few questions about using AI technology: One, does the strategy align with the organization's overall strategy, or is it just chasing a trend? Two, does the AI technology fit or solve the problem correctly? Is implementing a chatbot the answer when we could be using mass customization to address customers' needs? Three, is the expected return worth the investment and disruption? Will the technology actually deliver the savings necessary to fund the transformation?
The Process:
We all heard the saying, "It's only as good as what you give." Organizations must ask themselves how efficient and effective their processes are. They must ask themselves what the condition and quality of the data we are going to ask the AI to analyze are. Before any organization can optimize or realize the benefits of an AI transformation, it must address gaps and inefficiencies in its processes, eliminate unnecessary steps, and optimize processes and systems to create value.
The people:
Every AI transformation is a cultural change; it affects how people work, think, and operate day to day. For an AI transformation to be successful, employees must understand the reason for the change and realize that doing things the same way will cost more than the pain and disruption of adopting the new way. Next, employees need to have agency over the transformation. They need to have time to understand it, use it, explore it, and accept it. They need to see the value of the tool and understand that AI technology can enable them to be more creative and impactful at a scale and speed never before seen.
Is your AI technology transformation embedded in a systems thinking approach?
Does your AI transformation align with the organization's strategic goals?
Are your processing systems and data efficient enough for the AI technology to utilize effectively?
Do the people utilizing the AI technology understand why it's being implemented? Do they know how to use it?
As always, I welcome your thoughts and comments.
David