Does AI deliver value on its own?
First of all, AI is not clearly defined. OECD (link, link) defines it “as a machine-based system that can, for a given set of human-defined objectives, make predictions, recommendations, or decisions influencing real or virtual environments”. Especially the part “machine-based system” highlights the fact of technology, it does not deliver a distinct function, it helps to solve business problems. Cognilytica identified 7 support patterns:
- Hyper-personalization e.g. Spotify playlists or Amazon product recommendations
- Recognition e.g. face detection in cameras or voice-person matching in smart-speaker
- Conversation & Human Interaction e.g. smart speaker or support chatbots
- Predictive Analytics & Decisions e.g. adaptive case management in health systems or car navigation systems
- Goal-Driven-Systems e.g. chess computer or traffic routing systems
- Autonomous Systems e.g. vehicle or production line steering systems
- Pattern & Anomalies e.g. order forecasting or fraud detection for bank transactions
With respect to these patterns, AI is more like a digital production asset. You can produce better, individualized products or even invent new kinds of products e.g. smart-speakers. In this manner, AI drives product transformations, especially in the way how the product lifecycle happens. The changed product lifecycle is a cornerstone of digital aged economics. The product lifecycle speeds up and at the same time becomes more individualized. The classical engineering design still exists but becomes a new quality. In the past, engineered products were delivered after a perfection phase – don’t misunderstand it with delivering bad quality. Today, products more and more, expectedly finished at the customers’ end. Digital aged companies deliver reduced functionality combined with a (machine learning) model as the individualization (AI) heart, continuously collect data and adapt and extend functionalities according to the respective user. Physical aged companies stick on the concept e.g. maybe hundreds of pre-optimized shipped functions.
Out of that, AI will reshape companies understanding of portfolio- and asset rating. In addition to these topics also the customer interaction will dramatically change. The easiest example is today’s handbooks. A common handbook explains every functionality, gives restrictions and facilitates the companies product liability measurement. In the future, companies have to give control back to the consumer and the product itself needs to become more responsive (direct or indirect). Concepts like XAI (link) or the rules states by the OECD (link) will lead to this.
Summarized, AI absolutely delivers value on its own and can be productized, but this highly depends on the companies (engineering & product) culture. It is like the glue between classical and digital economy. And yes, it is not the one and only product, also glue exists for different purposes. May we require 7 groups of AI products (enabler).