Quick fix vs. transformation fundamentals
The Ether with Casey SimpleAdd bookmark
The Ether with Casey Simple
We’ve been doing transformation projects for the last 15 years as a way to approach how we implement new technology. We used to take multiple years to get them done and we've gotten them down to generally within a year. The projects are agile, and that really has been the fundamentals that most project teams and IT departments have been using to do technology transformation.
AI: Artificial Intellegence or Alogrithmic Intelligence
Then AI comes along as this shiny, new toy with these powerful possibilities that can help solve system and data and process problems. There’s a temptation to drop AI solutions on top of these problems, because it can really offer a quick fix without having to change the underlying issues that are causing the problems that you're using the AI fix for in the first place.
I've seen this happen in projects many times, and eventually you do have to get to the fundamentals that have caused the need and the desire for the AI solution.
The other problem that can happen in this sort of quick fix orientation for AI solutions is that the implementation is often done in silos. This can make for really disjointed solutions that end up creating more problems somewhere else. I saw this in a project where an expense team created a really great AI solution to improve their audit in tax reporting capabilities, and it caused a tremendous slow down for employees when they were actually entering their expense reports. It was quickly scrapped after months of development and implementation.
So what's the alternative? I think the slow, progressive, large transformation project where you identify all the root problems and you fix all those first, is probably not always the best thing to look for first. Sometimes that can take you so long that you never get to the shiny new toy.
There's actually a place for both the quick fix and the transformation fundamentals, and you can combine them in a way to use the best of both. If you have an idea for a quick fix AI solution, before beginning you should consider a couple of key questions. The first one is, how does this fix affect other activities in the full flow of the process? This means stepping back and doing a good fundamental process flow review. The second question is, could this fix be accomplished in a non-technical way? Are there policy issues that are creating the problem? Is the process outdated for the desired outcomes? Those are the kinds of questions you can consider. Again, this requires a good fundamental problem review.
What you've done are two of the initial steps of a transformation project: you've looked at the full flow of the process and you've done a problem review. At this point it will be easier to decide if the quick fix with the shiny new toy is appropriate, or if you really do need a full transformation project to occur.
The AI solution may still be a part of the transformation project, but now you've got a clear understanding of how it can benefit the long term effectiveness of the particular area you're considering for the project.
My advice in this area is to be aware of the shiny new toy, but don't forget that transformation fundamentals can aid you in choosing the right AI projects, and how to implement them effectively and quickly.