Creating Foundational Capability: A Credible Path to AI Breaking down the journey to harnessing AI in digitally transforming your organization. For large and small companies alike, a credible path to harnessing artificial intelligence (AI) for competitive differentiation is murky, covered in silver-lined pot holes. AI has been underdelivering on promised benefits for years; nevertheless, it is attracting more investment dollars than ever. Given the quick pace of advancements in computational science, there are serious implications for any organization, both in terms of potential vulnerabilities and growth opportunities. To clear the noise, the MIT Sloan Management Review took the pulse of 3,000 global executives. This study brings data to decipher where companies are in their AI journeys:
75% of global executives believe AI will enable companies to move into new business. Almost 85% of executives believe AI allows their companies to obtain or sustain competitive advantage.
Only one in 20 companies has extensively incorporated AI in their offerings and processes.
Fewer than 39% of all companies have an AI strategy in place.
Maturity Model: From Description to Prescription Part of the challenge in defining the enterprise-wide opportunity for AI is the complexity of the technical area, as data science is not yet mainstream. Without getting “technical,” AI’s benefits and value to an organization remain nebulous and only within arm’s length for trained data scientists. Predictive Analytics and Machine Learning (PAML) will be the connective tissue between Business Intelligence and AI.
© Market Strategy Group, 2017
October 2017 www. mkt-strat.com
For the citizen data scientist or organization without AI capability, understanding the maturity model from description to prescription is essential. It’s a journey that requires developing a fundamental data analysis capability and builds iteratively toward establishing a multi-functional and anticipatory prescriptive capability. DESCRIPTIVE
DIAGNOSTIC
PREDICTIVE
PRESCRIPTIVE
Artificial Intelligence Natural Language Processing
Machine Learning • •
Deep learning Predictive analytics
• • •
Classification & clustering Information extraction Translation
Expert Systems
Speech • •
Speech to text Text to speech
• •
Medical diagnosis Financial advice
Planning, Scheduling & Optimization
Robotics • •
Autonomous learning Pattern recognition
• •
Supply chain planning Autonomous pricing
Vision • •
Image recognition Machine vision
Instead of focusing on singular business applications by vertical or horizontal, organizations must “look under the hood” at their internal data science capability sets to determine where they are and how to establish a credible path to AI value. We suggest a capability review to get your teams “AI ready” – apologies in advance, it’s neither sexy nor a silver bullet. Instead, it’s good, old fashioned perseverance: 1. Establish Foundational R and Python Skills – These are the most commonly used programming languages with basic support/community activity. Choose one of these languages and institutionalize. 2. Refresh the Basics of Descriptive and Inferential Stats – There is no time like the present for a refresh in statistics, before you take on advanced predictive and machine learning techniques. 3. Get Your Data Right – Data exploration, cleaning and preparation will separate good data from bad data, and great results from poor results. This process is labor-intensive, spend time in solutioning. 4. Introduce Machine Learning – Learn the introductory set of PAML algorithms and methods – decision tree, regularization, regression, clustering, Bayesian, dimensionality reduction, instance-based. 5. Implement Advanced Machine Learning – Explore deep learning, machine learning with big data, advanced ensemble modeling, text mining and distributed computational infrastructure. 6. Scale Through Clicks, not Code – Find appropriate infrastructure and automation tools to do more with less; it takes time to develop data science unicorns, get more from the citizen user in the interim. Once that foundational capability has been established, leaders across the business can assess where you are today and plan for a digital future and leverage multi-functional autonomous intelligence. In addition to thinking in broad strokes, organizations should roadmap their AI vision and highest-potential applications. About Market Strategy Group Market Strategy Group sees business differently. To us, winning isn't about going from Point A to Point B. It's about aligning three distinct dimensions — business direction, people & process, and markets & customers — to drive growth. Getting this right is critical to your bottom line. And it’s more than an operational challenge, it is often a strategic issue. We have helped organizations get to an answer custom to your unique situation. We should chat if this is an issue: Alex Kruzel (alex.kruzel@mkt-strat.com).
© Market Strategy Group, 2017
October 2017 www. mkt-strat.com