Most companies have now realised that artificial intelligence (AI) promises enormous potential for almost all sectors and areas of application. However, when it comes to planning and implementing AI projects, those responsible are often still confronted with numerous challenges.
We present the seven most common stumbling blocks in AI projects that need to be addressed in order to make the project a success for everyone involved.
1. a common understanding of AI
There is no single, universally valid definition for AI - everyone associates different ideas with the term. It is precisely this circumstance that can become a stumbling block in the project. All project participants, including those from the specialist departments, should be informed in advance. It needs to be discussed exactly what the Artificial intelligence is capable of and what it is not. Myths and misunderstandings must be dispelled. Any worries and concerns of the employees, such as the fear of losing their job or of critical situations arising from AI predictions, should also be taken seriously. This is the only way to create a uniform understanding so that everyone is on board and pulling in the same direction.
2. the data basis: quantity and quality
Data is a company's treasure and the basis for every AI project. Very large amounts of data are needed to train models - and these are not sufficiently available in every company. Another challenge besides the quantity is the quality of the data. The content must cover all possible constellations so that the model is later equipped for the actual use case. Only if anomalies are also found in the data sets can these be taken into account in the training and uncovered in the application. If the data sets are also unbalanced in any way, this leads to an undesirable bias - the AI then makes decisions that may not be optimal.
3. too high expectations
Every project is preceded by a cost-benefit analysis - and especially when it comes to comparatively new and complex technologies like AI, there is a certain scepticism among decision-makers. The added value of such a project must be directly recognisable. In AI projects, however, there is no guarantee of success. Whether a neural network The success of training depends on many factors, but above all on the data. Even the exploration of this data and the training of the networks requires effort and costs, even if the process is not successful. Thus, every user assumes a certain financial risk. Accordingly, expectations should not be too high - because there is a possibility that the project will fail.
4. the interpretation of the results
As common as scenarios of a machine taking over the world are, it is unlikely in the foreseeable future that decisions will be made solely by an AI. Instead, the technology provides decision-makers with a data basis on which they can make optimally informed decisions. To do this, however, the results of the AI must be interpreted correctly. These are rarely black and white, but always have to be seen with a certain accuracy of prediction. This accuracy must be put in relation to the prediction of analyses by humans. Because humans also make mistakes, for example when diagnosing diseases on the basis of X-ray images. Therefore, it is important to compare the prediction quality of AI with that of humans.
5. AI is often not explainable
In most cases, AI is a black box. Not even a data scientist can clearly explain why the AI has made a certain decision. That is why "Explainable AI" is a broad field of research with the goal of making the technology more transparent for humans - which will reduce any reservations and concerns about AI in the future. So far, the basis for decision-making can be explained, at least to a certain extent. But until there is full transparency, one has to accept that AI will make decisions whose solution path may not be comprehensible.
As with all new technologies, AI is also fraught with a certain amount of scepticism. In addition to the factors already mentioned, the effort required to collect and label the training data often discourages those responsible from introducing and using it - even if the use of AI means an increase in the effectiveness of certain processes in the long run. As a result, potentials in companies are not exploited and the companies risk losing their competitiveness. In our Basic article on AI 2.0. we explain why even non-tech companies should now intensively engage with AI and their own use cases so that Germany does not fall behind as an AI location.
Before an AI project starts, the question of who owns the trained neural network and thus the intellectual property should be clarified in any case. In any case, the data is the property of the creator, i.e. the respective department, but often the service provider would like to use the insights from the project elsewhere. When using AI functionalities of the large providers via services or APIs (such as Amazon's Alexa), the customers often have to agree to the use of the data to improve the services. This aspect should therefore be explicitly regulated in the contract to avoid disagreements later on.