The answer gives us clues to leading AI in our Org's -
The Unclear Path Ahead
There’s a lot of uncertainty over the role of AI in industrial transformation today. What is becoming clear however, is that companies need to be on the bus of implementation.
Therein lies the problem for Executives contemplating how to deploy AI into their organizations. What do we use it for? Where to we focus it? The answer isn’t straight-forward. We’ve all see the stories of AI’s that were given tasks and performed miserably, possibly hallucinating information or synthesizing it in inappropriate ways. We see the AI images anyone can generate, or the LinkedIn posts we know were written by AI. A feature of an AI world is the so-called “AI slop,” content generated easily and without much care that is obvious enough to discredit the source through its use. And we also see the statistics on effectiveness of corporate AI initiatives, where the majority are reported to not offer much ROI.
So how to think about and focus AI adoption in your company?
The 2 Power Users and 1 Dunce Using AI in Our Organizations
In his book, Adopting AI: The People-First Approach, change management expert turned AI advocate, Paul Gibbons advocates that we avoid thinking of AI as a technology or tool, and consider it more as an intelligence. “This pivots the strategy question from ‘Which problems can this new technology help solve?’ to considering ‘how could more intelligence help us?’” This framing includes the full potential of AI to get leaders to rethink roles and structures rather than keep a limited use-case mindset.
Thought of as an intelligence, we see three type of users come to mind, and we’ve seen examples of users in all of these realms.
The first is “The Expert” using AI as an assistant to automate or complete less complex tasks within a workflow, but where the expert judges the quality of the work using his expertise. This is a common approach that people naturally pick up with AI, especially when they are uncertain of its capabilities and don’t want to trust it with much. Many experts eventually realize that how they codify the context, question and manner of completion determines a lot of the quality of the result, and as they get comfortable in their explanations and prompts, they test the AI on more complex tasks.
The second is “The Student” using AI as a learning guide to retrieve and explain concepts and enabling methodologies in a domain. Here AI can be a powerful teacher, and greatly accelerate rates of learning.
The final role we see is the one that becomes problematic. We call it “The Sorcerer’s Apprentice,” and it usually occurs when someone who is less than an expert, outsources their work entirely to an AI, yet does not have the contextual task knowledge or experience to determine if the resulting output is acceptable in quality. Much like the character in the Disney movie who loses control of the magic brooms, this person cannot judge the output quality and therefore does not know how to instruct or redirect AI.
This is the source of “AI slop,” people trying to skip learning and mastery.
Actioning the Insight
Now in reality, these three roles are possible in each one of us, and we often find ourselves in each context. What the frame gives us is a way to evaluate activities with AI. Do we have the expertise to determine quality of the output or are we instead hoping the at the AI will give us quality output just by asking for it, but not knowing it? Are we putting ourselves in learning mode with an intent to develop new understanding of a topic or are we asking the AI to simply use it’s own “understanding” in place of ours. An AI might write good copy for you, but if you don’t have the wisdom to determine its quality, then slop will always get through.
This simple yet clear framework, also gives a place to start with AI implementation that is agnostic of function. We think of it in 3 facets.
1. Get everyone learning with AI.
There is a democratization effect with AI because of the power it can bring to learning. Everyone can take part here. In fact, it’s a way to drive the codification of the knowledge about the business.
2. Engage your experts.
AI isn’t a situation where “the young kids know the new tools best” That’s technology thinking. You desperately need the wisdom of your best people leveraging themselves with AI. In fact, this is where the most leverage can come organizationally. Make sure that initiatives that are intended to leverage knowledge across a function has its best experts involved.
3. Beware paths that try to turn novices into experts, or outsource all the thinking to AI.
AI’s gonna write your marketing copy for you? Beware. Just by filtering out these cases or by getting an expert involved in the project, your portfolio of initiatives gets a huge quality boost.
By the way, afore-mentioned AI expert, Paul Gibbons has just written an optimistic piece on LinkedIn where he looks at what a future filled with Students and Experts might look like. See “The Fourth Stonecutter.”
Until next week,
Kendall -
Find me on LinkedIn or Book a 1:1 call
Not a subscriber yet? Subscribe here
