This past weekend I was in Chicago to attend an IIM-Africa conference. I sat in on a keynote by Dr. David Marco, the highly respected Data Warehousing expert, on using AI in Information Management. I found it fascinating and wanted to share some personal takeaways with a broader audience.
Firstly, his use of the term “Information Management.” Organizations like ARMA, AIIM, and Deep Analysis often throw the term Information Management around, but when we do so, we tend to ignore the world of structured data. Information Management in most circles folks like ourselves navigate refers exclusively to managing unstructured data. Conversely, others (such as Dr Marco) use the term Information Management to encompass the world of Big Data, Warehousing, and ETL and typically exclude unstructured data. On the one hand, this is just an issue of semantics and seems unimportant. (My colleague Dan recently discussed the problem of confusing terminology, and I highly recommend you listen/read that discussion). On the other hand, it spotlights many of the issues we face in the industry, for as in broader Computer Science work Information Management refers to all the Data maintained in IT infrastructures. Ultimately we are all part of the solution and part of the problem but have a tendency to go our own way.
Secondly, both sides of this Information Management equation have much more in common than we sometimes think. The technologies and techniques may differ, but the goals, concerns, and approaches continuously overlap. This struck close to home, as in listening to Dr. Marco’s presentation, I knew I could deliver the same speech with a few tweaks in terminology. His breakdown of the powers and limitations of AI was the same as I would define them. His emphasis on the need for clean and accurate data, his concerns over poorly managed, duplicate, and redundant data, the need to create and manage precise metadata, etc, were all exactly as I would describe the same. And to top it all, an emphasis on the need for strong human skills to organize, structure, and categorize data before AI can deliver its benefits resonated profoundly.
So what’s the moral of all this – well, I would say that two divided communities need to come together and explore our commonalities and differences more often. The potential of AI means we both need to move beyond explaining how AI works and better explain what needs to happen before it can deliver anything of value. Most notably, how to meet the global challenge of vast mountains of messed up piles of data that need expert human guidance to bring order to the chaos; until that happens, the true potential will never be unleashed. Sure, there will be small pockets of advancement, but beyond the investor hype, it will struggle in enterprises. No investor wants to hear about the sheer scale and complexity of the work involved in Information Management and Governance, but enterprises urgently need to listen to it.
There has never really been any doubt that the structured and unstructured data worlds could learn much from each other. But there has been little interest and even a reluctance to bridge the divides. But it does need to happen. Silly though it may sound, if every Datawarehousing professional read a book on enterprise content management and conversely every document management professional read a book on data management, we would likely see the ignition of a period of beneficial change, and that moving forward, we could tackle the challenges and leverage the undoubted benefits of AI and digital modernization together.
Finally, as the quirky photo suggests, I was in Chicago for a specific reason this past weekend; I was inducted as an honorary fellow of IIM-Africa, a lovely gesture and an honor gratefully received.