What makes AI so different from other Technological Advances and why should anyone care? Simple, technological advances continue to progress on an incremental basis, but AI is a Quantum Leap forward, because it enables systems to learn and adapt autonomously with precision and performance, that equals incredible speed to value. Think of a pyramid with data and operations systems at the bottom, AI in the middle, and enhanced business decision making using AI derived insights/information at the top. AI empowers organizations to make faster and better decisions to identify risks, serve customers, and optimize processing. For example, making purchasing recommendations by past buying history, online research conducted/interests helps companies sell more product. This is because learned algorithms can handle large amounts of data to make decisions typically requiring human processing and reasoning at light speed with accuracy, a force multiplier.
Saying no to AI would be like saying I would prefer to move my entire family from Maine to California by horse and buggy instead of flying. Now image if you are in the product shipping business and instead of moving people, you are moving the product. Within months your company would be out of business. “Growth”, everything in life must serve a purpose or it dies. A Shipping company using a horse and buggy to deliver goods to folks that want their goods in hours/days will die because it has no purpose, it is a self-fulfilling prophecy and factual reality. Such is AI, and to ignore its existence is to put yourself/organization at a competitive disadvantage or out of business.
So, you are probably thinking, tell me something I don’t know. I am afraid you are too smart for that, so let’s dig into something you do know; this constant talk of AI...and what it can do or will be able to do can be very confusing, especially when pressed against the landscape of your business or discipline. Now that we have determined you want your business to go on and even thrive, you must embrace AI..., but how to do so without chasing red herrings or increasing your risk tolerance?
Firstly, the world of AI is fueled by one source and that is “data”. “GIGO”, you may have heard this expression, Garbage In- Garbage Out. No matter how good the AI is, if the data is bad, you have a big problem. Scenario Planning, Decision Making and Risk Mitigation all become flawed and place an organization into a sea of vulnerability. Before addressing what can be done about it let’s look at the Four Biggest Risk Factors of AI mass adoption according to a March 2023 article by Deloitte.
Deloitte tells us that the following Risk Factors will be a force to be reconned with as AI becomes widely adopted:
1. Uncertainty-is the information outdated? Is the data being manipulated with false narratives?
2. Explainability-Doesn't provide the real sources of the data
3. Bias-Necessity of being aware of inherent bias (i.e., learned bias – green is good)
4. Environment Impact
a. Privacy & Security-as AI needs large amounts of data, inadvertently getting a hold of personal or other confidential information is a risk
b. Unemployment and Economic Disruption
So, what is the solution? Unfortunately, there is no quick fix, and it harkens back to Data Management. If data sources are not trustworthy or there is misinformation in a data source, then the analysis and conclusion/s will also be flawed. Biased information, such as commentary/opinion versus fact will likewise need to be reconciled. In turn, as AI advances, we can begin to see the need for rationalizing and testing the validity of the data to ensure more valuable results. Any organization that has taken the cleaning of data seriously knows this is an ongoing process but has the benefit of having more reliable data, which in turn when AI is applied to it makes for more informed decision making. Where things get complicated is when organizations have multiple data sources, like from ERPs, CRMs, Point Solution after Point Solution, massive data lakes and the like, all of which must be mined and reconciled for value, and therein lies the challenge.
In dealing with this challenge, we are seeing more organizations utilizing a platform above the operational systems that has cleansing and data organizational filters, with easy to employ APIs for direct connection to the platform. This may not be perfect, but it is a material step in the right direction; unfortunately, only a couple of suppliers have mastered that upper layer where they can absorb both operational and financial data for enhanced decision making. Even so, these suppliers are still figuring out how to maximize AI and how to mitigate Deloitte’s four risk factors as detailed above.
Two things are for sure in this new world of AI:
1. The cleaner and more reliable your data, the more effective AI will be, and
2. AI is here to stay so organizations need to find a way to harness it or go the way of the horse and buggy
Gartner says by 2025, 10% of all data produced globally will be from AI and by 2030 the AI market will reach $15.7 Trillion with an expected increase if global GDP of 26%. These are incredible statistics even if they are only 50% correct. The question is how your organization will take advantage of this phenomenon without getting caught up in the noise, leaving you stuck in a place like the horse and buggy.
If that wasn’t enough KPMG predicts that AI will boost global labor productivity by as much as 7% over time. This is so because they say, “leaders are embracing AI to drive material efficiencies for their business and help workers do their jobs more effectively”. As you know, AI does not create new knowledge, it uses existing knowledge to achieve specific goals and that is why it will be used to work with us, not in place of us. That doesn’t mean there will not be labor reshuffling, but one thing is for sure, organizations will have to be ready to help their organizations adapt to it. In short, “Adapt and Overcome or Risk Being Overcome”.
Some Key Factors to Consider on your journey:
1. It all starts with defining clear objectives and knowing why you want to achieve them. Your objectives will need to consider not only where you want your organization to go but also where your industry is headed and that may include looking back (crunching years of data) in order provide a clearer path forward.
2. Assuming you have done this before, a Re-evaluation of your People, Process and Technology. Let’s address one at a time.
a. People: As AI begins to replace more rote tasks how will your organization re-purpose the people that were performing them or not. What new skills will your organization need and outside the Fortune 500 most organizations will not be able to afford a small Army of Data Scientists. This is an area where organizations need to think strategically, perhaps a blend of insourced and outsourced resources.
b. Process: Perhaps the low hanging fruit of AI, automating repetitive tasks, reducing errors and freeing employees for strategic work. AI can also uncover inefficiencies through process mining and provide data-driven insights for optimization. Moreover, predictive maintenance for things like when equipment is likely to fail will save down time which can be synergistically tied to supply chain optimization for enhanced demand forecasting and inventory management all resulting in cost reductions and labor savings.
c. Technology: Even though Al is not going to require replacement of current systems, it will require having access to the data from many of those systems. The problem with most data is its lack of cleanliness and reliability, and how to access it. As you know 99% of all systems do not have the ability to communicate with each other without a great deal of human/manual intervention; this is where a clearing house of sorts would be a tremendous benefit, but as discussed there are only a couple of suppliers capable of doing it.
3. Developing A Plan for Change
Before going live to production, you are going to want to beta test long enough to work out the kinks and that means developing a “use case”. The best use case you can come up with should include one with a healthy ROI. Why? You don’t want your organization to experience inertia because leadership as well as the other members of the organization lose interest; remember the horse and buggy cannot survive in this new world. One meaningful statistic to keep in mind is that the average data scientist is spending 60% to 70% of their time just cleaning the data before embarking on a given project..., so choose wisely. You would also be wise to keep in mind Deloitte's Four Factors to mitigate risk. So, you may be asking, is this all worth it, hence the ROI. Example: What if an Insurance company could reconcile underwriting, claims and financial data to better predict claims and thus enhance their underwriting standards to increase profitability. I think leadership would sign up, hands down. This is why one must pay attention to all these areas of consideration.
4. Go Live
This is a scary and exciting part of the journey. Putting that use case into production and living with the results. Here you must be ever vigilant and do your best to avoid GIGO and have a monitoring process in place to manage those four risk factors. Like anything in life, conditions change, environments will not remain stagnate so you must adapt and overcome to succeed. I would submit this is one of most exciting times for business, science and education and we are only limited by our imagination. Moreover, you can be sure that there will be plenty of start-ups looking to help you on your journey and in fact that may very well be necessary, so choose wisely.
5. Ongoing Maintenance and Continuous Improvement
This may require a combination of insourcing and outsourcing to maintain the right balance of ROI and Risk mitigation for your organization. You may wish to think of this part as something akin to continuous project management. The one thing for certain, no one organization will have all the answers and only a select few will be able to afford an army of data scientists; and perhaps even that may be risky as it is likely to create insular thinking at a time when out of the box thinking will likely have the greatest payday. This area will require several key KPIs, and continuous validation and you can bet new technologies will emerge to assist in the process.
Even though earlier forms of AI have been around for a while like predictive analytics, new AI deployments like machine leaning and chat bots are a Giant Leap forward and organizations will need to ready themselves to survive, let alone thrive.
Good luck and enjoy the journey during these exciting times.
Mr. Marinelli has more than 25 years experience in both Industry and Consulting, where he has worked for such leading companies as The Walt Disney Company, Verizon Communication, Deutsche Bank A.G., Siemens A.G., and Guardian Life Insurance Co in various operations management roles at both the divisional and “C” levels. Mr. Marinelli also serves as the CEO of Marin Advisory Services, an affiliate firm providing consulting and staffing with a focus in financial regulation and technology projects. Mr. Marinelli specializes in expense management, back office re-structing and modernization, vendor management and out-sourcing services. Over his career he has also worked on nearly all aspects of operational management and organizational re-structing which included system evaluations, and implementations , process improvements , contract management, cash management, out-sourcing , change management , expense management , organizational design, and function , policy creation.
Comments