measures regarding AI ethics

[suchona.kani.z]

However, in both cases there are measures to establish correct AI ethics or to allow AI to make better ethical decisions.

In his book "Human Compatible: AI and the Problem of Control," author Stuart J. Russell, professor of computer science at the University of California, Berkeley, gives a fitting example of how far-reaching the ethical aspect of AI is. He mentions the selection and recommendation algorithms in social media: These are not particularly intelligent, and yet they influence billions of people. Another new example is deepfakes. These are media content (videos, audio files, or photos) that appear authentic, but have been altered with the help of AI or deep learning. These are particularly unethical because they deliberately manipulate people.

There are various measures for creating an ethical AI application that is correct for your company. It starts with the data that is used for the machine learning algorithm. The data is always divided into two data sets: The training data set is used to create the AI ​​model that is able to make an (automated) decision using the algorithm and newly collected data. To italy consumer email list stay with the simple example given above, the decision would be made here that the fruit is classified as an apple or a pear based on its color. The validation data set is then used to test this AI model for its "accuracy". (Did the result for red fruit really come out as an apple?) This is where the human or data scientist comes in. The data scientist checks the result and checks whether the AI ​​model needs to be adjusted - for example, there was fruit in the validation data set that could not be assigned to either of the two properties red or green.

Here the data scientist can consider whether to add further properties - for example, an apple is round and a pear is oval, to increase the accuracy of the result. In the AI ​​genre, this process is called feature selection or feature engineering and also indirectly belongs to the area of ​​AI ethics. There will be no unethical decisions with fruit, but if we convert the fruit back into the creditworthiness of person XY, then we have to be careful that the combined data and properties comply with ethical standards.