Are you ready for your AI transformation?
I’m going to go out on a limb and say, “probably not.”
Listen, have you just completed the potentially lengthy (and somewhat dilatory) process of digital transformation? If so, your organization may be completely on-board with your drive towards internal democratization of data. You might preach to the proverbial choir when you wax philosophic about the value of the horizontal application of processes and data transparency. You may have seen some external successes with various proofs-of-concept/technology and thought, “if only I worked in a company that properly supports innovation/R&D…!”
Take a breath. Your heart is in the right place. Your head, too. But to be honest, the third-party company you are considering for an intra-company artificial intelligence initiative is probably just a little full of BS. I’m not saying that you cannot create an intelligent proof-of-concept or develop a point solution that leverages AI. But the horizontal application of artificial intelligence across your entire organization? Color me skeptical.
“I’m not saying that you cannot create an intelligent proof-of-concept or develop a point solution that leverages AI. But the horizontal application of artificial intelligence across your entire organization? Color me skeptical.”
You may desire to be a technology pioneer, but you have to ask yourself the following questions about your company:
- Are we gathering all the data that we could use to drive value for our users/customers?
- Is our data well-structured, categorized, and classified?
- Is any part of your business completely unaware of what is being pursued by another part?
- Have we evaluated the applicability of AI platforms, understanding that their “black box nature” may be more deleterious than a bespoke solution?
If the answer to any of the above is “no” or “sort of,” proceed cautiously. Data hygiene, engineering, and normalization can be costly and time-consuming. New technology initiatives should absolutely be preparing for the AI-driven future but implementing such features or solutions may be premature.
Editor’s Note: AI Technology: Machine Learning vs Artificial Intelligence
AI technology refers to both “general AI” (sentient machines — science fiction) and “narrow AI” (thinking tools without actual cognition). Machine learning is a type of narrow artificial intelligence designed to identify patterns that humans would otherwise be unable to identify. Machine learning is now used across many industries, from banking to healthcare. Machine learning (also called deep learning) builds upon data science algorithms to glean actionable insights from exceptionally large data sets, increasing accuracy over time. While the terms artificial intelligence and machine learning are frequently used interchangeably, the distinction is essential.
About 5-7 years ago, a host of AI services companies were fitfully born. They correctly perceived that the market had an appetite for machine intelligent solutions. One of my go-to phrases when speaking to clients about the Fourth Industrial Revolution was that “what looks like magic to your competitors in 5 years is actually the result of today’s good planning.” I still believe that to be true. But the stats are not encouraging – Fortune estimates the failure rate for A.I. projects at between 83% and 92%.
Editor’s Notes: AI Workloads: Where Does an AI System Live?
Cloud solutions, such as Google Cloud and AWS, have proven to be uniquely impactful in the AI industry. Cloud solutions have the resources necessary to build complex analytics and AI algorithms. The machine learning model relies upon high-performance computing, frequently processing large data sets with advanced analytics algorithms. Still, the success of an AI-based solution will ultimately require the proper implementation. Today’s enterprise AI capabilities rely upon data scientists with accurate training data and an eye for customer experience.
In my experience ushering organizations towards a future driven by intelligence, I’ve come to find that many simply aren’t ready, organizationally speaking. And all the flashy vision videos, R&D marketing, and shout-outs on design blogs can’t change a client org’s data hygiene. Your company may have digitally transformed, moved data to the cloud, and ensured that your data is usefully parsed and categorized, and that is to your credit. But the fact is: most businesses are still emerging from digital transformation and need to mature (or evolve) to AI-readiness. Sure, the promise of AI is compelling, and the capabilities it may one day enable are attractive, but it’s worth being honest about your company’s developmental phase. Some industries have already embraced intelligence holistically – finance, medicine, pharma, and some cutting-edge energy and resource companies, specifically. To prepare for an intelligence-driven future at your organization:
- Hire a data scientist or data engineer (you know you want to).
- Normalize structure across the organization– or specifically in the industry/vertical/ department in which you plan to implement.
- Coordinate with other leaders in your company to drive data and initiative transparency.
- Popularize the idea that failure is just one step in an iterative process that winnows the field of possible solutions.
- Architect new solutions with a clarity of understanding around how intelligence will be woven into future releases.
AI at your company may yet be some distance in the future, but planning for, exploring, and researching is something you can (and should) start today. If you’re deciding upon a partner, pragmatism should be a key characteristic of whomever you choose.