by Johannes Humbert – LinkedIn | Twitter

Companies cannot handle artificial intelligence like conventional software. We need a fundamentally different understanding of how artificial intelligence can be used in business.

Artificial Intelligence (AI) cannot be handled and purchased like conventional software, because otherwise it would lose its enormous potential as a competitive advantage. In order to generate a competitive advantage - and not just to keep eye level with the competition - it is necessary to understand AI as a strategic product.

As an analogy, we can look to the electrification of factories beginning in the 1880s. At that time, electricity did not initially register any efficiency gains over steam engines. It was only when it was realized that machines no longer had to be located close to or further away from the steam generators according to their energy requirements, but could rather be positioned according to the work processes, that the economic quantum leap that we associate with that phase today occurred.

What do we mean when we talk about AI?

AI today means technologies such as machine learning, deep learning and reinforcement learning. These are, at their core, highly developed statistical methods. They give us information about the behavior of customers or evaluate language on a human level. Only with the help of AI can a computer understand from the context of the conversation whether "bank" is something to sit on or an organization that manages and keeps money.

In contrast, conventional IT systems are designed to handle processes that are already understood more efficiently or according to new patterns. When we talk about digitization today, it's disruptive process innovation at best. But often it remains just mapping analog processes in digital formats.

The return on investment (ROI) of AI is superior to conventional IT solutions.

Competitive advantages can be achieved through both conventional IT and AI-based systems. But the benefits derived from an AI application are usually significantly higher. Actual process innovations can be realized more easily and sustainably.

Only with AI it is possible to uncover rules and behaviors that would otherwise have remained hidden. Moreover, AI acts as a controlling entity that can actively shape and customize the flow of processes.

With Natural Language Understanding (NLU), for example, it is possible to read out concerns, urgency, mood as well as information from unstructured communication such as emails or WhatsApp. This efficiency in information processing makes it possible to serve customers individually in real time, taking full account of information from all available channels and systems without delay. No question remains unanswered, no prior event unconsidered. Communication becomes scalable.

Equipped with historical data and enhanced with methods from the field of reinforcement learning, such a solution will not only be able to ensure effective communication without increasing marginal costs. It will also continuously optimize itself and automatically adapt to changing conditions.

This dimension of efficiency and automation is simply not feasible with conventional systems.

Why does it make little sense to buy AI-based systems in the same way as conventional IT?

Fundamentally, any competitive advantage based on technology is temporary. If everyone else is building a chatbot, you have to go with it, but it probably won't change your company's productivity.

Artificial intelligence that can be used to realize a true competitive advantage always requires a partially scientific project design. It's about gaining insights that couldn't be foreseen before, not about casting the existing and familiar into new formats. AI is a shaping technology that will fundamentally change the way businesses operate. In contrast, SaaS is always a mere commodity.

Management teams that consequently aim to technologize their area of responsibility with AI-based software will almost necessarily do so in a program that allows them to rethink and reshape core areas of actual value creation.

In part 2, read about the issues to consider when purchasing AI versus traditional software.

Previous Post Next Post