Artificial intelligence is not a competitive advantage
Artificial intelligence (AI) is probably the technology that has the best ability to create completely exceptional results. It can be difficult to get started, but once an AI solution picks up speed, the results can be almost infinite scaling. But still, I do not see it as a competitive advantage.
So I do not think it is the ability to develop AI solutions that will be a competitive advantage in the future. AI algorithms are more and more off-the-shelf technology. Maybe a slightly cumbersome one at times, but very much an off-the-shelf item. So the real competitive advantage lies not in the AI, but in the data operations. So what are data operations? It is data management with two purposes:
Ensuring the right data in a high quality
Ensuring continuous flow of new data at a responsible price
This is both a difficult and a necessary task in the future for many companies. When AI needs to realize its potential, it can only happen when the necessary data is available. If AI is to be a positive business case, then the necessary data must also be obtained at a low price and in a quality that enables the AI to be good enough for the task it’s made for.
A good comparison is how supermarket chains compete. Of course, they must have good products and a great brand, but if they do not have a good operation on their stock and purchase of goods, then they will never be able to compete, as the price of products will be too expensive. The same will apply to AI. The one who at the lowest price obtains data in good quality will be the one who has the easiest way forward. Skilled developers with an understanding of machine learning and data scientists will of course also be of great importance, but if the data is not in place, they will find their jobs difficult.
A particular challenge with the field of AI is that you rarely know exactly what data you will end up needing. Even when you know exactly what problem you are trying to solve, it can be difficult to know what data is needed in advance. It is often an experimental process. So the task in the future is to have a general data operation that includes data you don’t necessarily see a use for right now. It can be anything from manufacturing data, finances and data on purchasing behavior. It may seem like a superfluous task, but if we do not, we will often regret it. We will look back on ourselves as the caveman who found a lump of gold and threw it away as it had no purpose right here and now. If you do not want to be a data caveman, then you need to get your data in order now.
