According to the Harvard Business Review, most current AI projects will fail. And yet, some companies mostly succeed in their AI projects.
They succeed because they are not AI newbies. They have already valiantly tried and sadly failed in early forays into machine learning and AI software applications. They each defined a reasonable target problem, gathered all kinds of data, hired numerous data scientists, secured executive commitment and ample budget, and engaged in lengthy projects only to produce little or no business value. And these are all huge companies with all the resources needed to succeed on their first try, and yet, they didn’t.
Despite being in differing industries and having differing use cases, AI comebacks have several things in common. If your organization is just getting into AI, these lessons will help you win your freshman season.
1. AI is not IT
Problem Seen – Many companies reflexively and mistakenly look to the IT organization to initiate and manage AI projects, as they routinely do with other software-related endeavors. But AI software is not like database, middleware, and enterprise applications software, the traditional technical domain of IT. AI software requires collaborative exploration and iteration by interdisciplinary, strategically managed groups with broad business, operational, and technical skills, and a mission of business transformation.
Lesson Learned – IT has two critical roles that bookend the successful AI projects. Initially, IT database skills are needed to acquire, integrate, and govern the information assets needed for machine learning, and finally, DevOps skills are essential to the effective, reliable deployment of new AI applications.
2. Data Lake or Swamp?
Problem Seen – A common mistake made by many companies in their first AI project is to focus on data first, prediction second and business problems third. They fill a data lake to the brim with diverse information from many unrelated sources and then try to discover how AI might be able to produce business value from that data. In addition to being inefficient and ineffective, for companies using personal data from EU citizens and residents, this is also illegal.
Lesson Learned – AI projects should first define a measurable business problem, like the cost of customer attrition. Then they should identify, qualify, and certify relevant customer-touching business process data and regulation-compliant customer data. And finally, they should specify and model an actionable predictive decision that can solve the given business problem using the defined business and customer data.
3. Start at the Glass
Problem Seen – In many AI projects, the people who will use the new application have little or no visibility or input into the solution until it is implemented and deployed. As a result, the AI team builds an application that is usually disruptive – and not in a good way – to established business processes and best practices; and user acceptance of the application is thus low and slow.
Lesson Learned – AI doesn’t take jobs; it takes tasks. And successful, profitable initial AI solutions should focus on automating low value business process steps that require high human effort.
4. Prepare to Evolve
Problem Seen – Most initial AI solutions solve yesterday’s problems. That is, by the time the solution is delivered, related business conditions and priorities have changed, creating a gap between solution function and business needs. AI, just like human intelligence, is dynamic and it is never finished. Many executive AI newbies are dismayed to hear that a predictive model is correct three times out of four, not realizing that this is actually a great result in many cases.
Lesson Learned – Successful AI projects depend on agility and flexibility in goal setting, resource allocation, and ongoing optimization. Going from 75% accuracy to 95% requires continuous improvements in data quality and availability, choice of machine learning tools and techniques, and target application goals and functionality. And, predictive models can be brittle; they might work well today but, due to changing goals and business conditions, not work at all tomorrow.
To avoid an AI train wreck in your first project and beyond, study the problems seen and lessons learned by others who have already tried, failed, learned, and then succeeded. Don’t be rigid, don’t be afraid, and, most importantly, don’t believe the AI hype. Instead, be practical, be realistic, and be patient.
About the Author
Tim Negris is SVP of Marketing and Business Development at Boston-based Rulex Inc., a provider of proprietary enterprise AI technology for Supply Chain, Financial Services, Energy Management, Public Utilities, Healthcare, and IoT applications. He is a long-time software industry veteran who has played a pivotal role in introducing many new advanced technologies to the world. Tim has been a prolific thinker and writer on a wide variety of advanced technology subjects, as a regular contributor to several large tech publishers and as a guest lecturer at Stanford University.
Sign up for the free insideBIGDATA newsletter.
One of my first jobs after college was working in a large commercial bakery in Pennsylvania. By working with commercial baking tools to aid us, we were able to bake thousands of loaves of bread for supermarkets. This helped reduce hunger. Meantime, the local bakeries remained in business and continued to prosper making higher quality bread a limited volume of loaves at a time. We did not advertise our commercial baking tools as “automated bakers” with “images” of “hands” and “bodies” of human bakers, because of course they were not. But those promoting “Artificial Intelligence” seem to want to do this type of negative imagery all the time. It is counterproductive. That is starting point of primary AI “train wreck,” the entire “taking over” narrative that some (esp in the AI and technology biz) want to actually promote. No business changes society by actively going to promote DISTRUST of its goals and means.