About the Author
Today you can open any number of paper and online resources in the business and technical realm and find a story or position on Artificial Intelligence (AI) and Machine Learning (ML). This is especially present in the multitude of technology companies that sell into our respective organizations, searching for their slice of the annual Information Technology (IT) budget and spend.
In many ways, the largest problem for all industries leveraging high technology is often the hype of the latest and greatest solutions and how they usurp the traditional solutions of the past. There is also a great deal of marketing and messaging aimed at changing the perception of solutions despite the reality that the foundations are steeped in techniques that have a long and distinguished history.
This was so thirty years ago, where the mantra was “… the PC will result in the extinction of the mainframe platform…”, yet, today the mainframe platform is still a material part of many organizations. Have mainframe solutions changed and adapted? Yes, and they continue to be a solid solution for many businesses.
The same situation goes for artificial intelligence and machine learning. Like other technologies, they have foundations which also continue to evolve and adapt. However, in every evolution (some of which are branded as revolutions), there are people, organizations, and solutions that seek to profit from the intrigued buyer, who is curious about the nature of the “new” and is perhaps unwary.
Here at ECI Software Solutions, we spend a significant amount of resources identifying, analyzing, integrating, and operationalizing technologies to service all aspects of our cloud solutions. Those efforts broadly address our products and ongoing operation of our solutions including systems management, security, compliance, or operational control. In these efforts, there have been many technologies that profess AI and ML foundations, but come up short on the delivery of the promise of such technologies.
In a recent issue, Forbes Magazine published an article, 13 Industries Soon To Be Revolutionized By Artificial Intelligence. Within each industry, there are rational and real perspectives as to the impetus for those industry-specific revolutions. Many can call these 13 perspectives foundational to the AI discussion as they highlight the requirement that drives the need for AI and ML. These include:
Some of the affected industries hit very close to home for ECI, including manufacturing, business intelligence, retail, and construction, to name a few. So, naturally, we ask, “Do AI and ML have the potential for being applied to our business?” The answer is “yes” however, pragmatic thought immediately follows, “But, only if it is real and accessible.”
So, what is “real”? Earlier, I mentioned that some organizations utilize hype to exploit the “new” buyer for profits. Here are some tips on how to not fall into this trap.
Like with many other relatively new evolutions, it’s terribly unclear what the tech industry’s definitions are for AI and ML. Depending on the vendor, the definitions change to suit their respective needs. So, for ease of discussion here are the Dictionary.com definitions:
Artificial Intelligence: The capacity of a computer to perform operations to learning and decision making in humans, as by an expert system, a program of CAD or CAM, or a program for the perception and recognition of shapes in computer visioning systems.
Machine Learning: A branch of Artificial Intelligence in which a computer generates rules underlying or based on raw data that has been fed into it.
As such, the evaluation of any technology professing AI or ML foundations have some tenets that they must keep to truly be classified as such.
So, while there is huge promise for this segment of technology, the goal should be to make sure you receive the gift and not just the wrapping paper. How do we do this? Here are three tips for considering technology in your business when it comes to this space:
1) Ask deep questions around how the primary AI/ML functions are managed. This is especially key when it comes to the use of your team to identify scenarios and situations that define actions for the system.
If there is no direct linkage to the system expanding or enhancing those definitions itself, you actually have a system that is dependent on human interaction to achieve its main purpose. This is not AI.
2) Watch out for solutions that have a crowd-sourcing approach to identifying and distributing material learning into your solution. Crowd-sourcing involves data sourced from the aggregation of organization and human experience. It may also involve system and technical data.
This technique is especially prevalent in the cyber-security space. Crowd-sourcing is a very beneficial technique for gathering security events and resultant protections and remediations, but it is not AI.
3) Dig into the “how” around the solution when evaluating, especially in the sales cycle. If you get a constant “that is proprietary” answer, the hair on the back of your neck should stand up.
In my experience, there is nothing proprietary unless you are asking for the underlying code and technical details. Communication of the layman’s “how” does come easy when dealing with companies that have something that is real.
While this is not an exhaustive list of things to consider, be wary and inquisitive. Once again, there is huge promise in this space. The real results will come from effectively identifying and choosing solutions that deliver on the promise.
Stay on top of industry trends and insights.
Subscribe to the Big Ideas for SMBs blog.
About the Author