Not long ago I was asked by a senior leader, “why aren’t we doing anything with AI?”
My initial thought was, “But we’re using machine learning in our forecasting.” But, that’s of course not what he was talking about. He was talking about Generative AI, Large Language Models, ChatGPT. All the cool stuff.
As often happens with new technology, there’s a lot of “Fear Of Missing Out” going on out there.
Generative AI is going to change the way we do things. There is no question. But for most companies, it’s more about being ready than about building new capabilities.
Here are a couple of questions to ask after that sense of panic subsidies:
Many if not most companies are digging out of some serious data/analytical debt. The fundamental descriptive analytical questions often don’t get answered because there are problems with data availability, quality, or accessibility. If you’re going to do the cool stuff, it’s important that the not so cool stuff gets done first.
This is different from the first question. The key issue is whether there are tools in place that provide easy access for end-users to get the information they need to make the decisions they need to make. The right visualizations and dashboards should be at people’s fingertips, for example. And the confusing clutter of irrelevant data should be removed.
The fastest way to turn the business against GenAI is to do random projects that don’t drive the business. AI and GenAI are tools. Asking “what can we do with them?” is like asking, “what should I make with this hammer?” Understanding the biggest pain points will surface the most important/impactful analytics projects. These may not be AI projects at all - sometimes simple diagnostic capabilities can make a huge difference right now.
The other day, I asked someone (not a data or technology person), “how do you feel about natural language processing?” They looked at me like I was crazy, which was what I was aiming for. My question was a little bit like asking a home owner how they felt about the nail gun used to build their house. NLP is now just a part of the apps we use, rather than something every company has to build. In many cases, this same effect is already happening with GenAI - chat capabilities and image generation built right into the tools we already use.
Having a defined plan and path forward inoculates you against shiny new technology hype cycles. When you have a plan, you can be thoughtful about how the novel technology and capabilities fit into your roadmap. If you prioritized the right use cases based on their importance to the business, you may decide that the new technology isn’t the right focus yet.
In the end, the question of whether to use GenAI, or any other cutting edge technology, boils down to this one fundamental idea: “Your problems are still your problems. The technology just offers potential solutions … or it doesn’t.” Finding a problem just so you can use the technology only promotes less important use cases over the most important ones. That said, there’s no doubt in my mind that there are real problems out there that GenAI is going to solve.
If you would like to understand how Proprioceptive can help assess your analytics capabilities and build a roadmap for the future, go to www.proprioceptive.io or shoot me a note at jeff.sigel@proprioceptive.io.