The following question arises from the dialogue and research of the Working Group (WG) on “Trust Research” in AdmEthics/Udesc. The WG proposes to research how and in what way the term “Trust” is being adopted and or used in the field of study on Artificial Intelligence (AI). In this sense, we will present the origin of the term and its implications. Next, we will discuss trust and its different uses in AI studies. Finally, we will point out some emerging trends around thematic trust.
Ethnology of the term “Trust”
The term “Trust” is derived from the Latin “Confidentia and Confidens”; the main ideas that these words express are of skill, quality, and ideas in action (OXFORD DICTIONARY AND THESAURUS, 1996). In this sense, it presents the concept of a state and not the affirmation of something.
Different uses of the term trust
The word “Trust” in the Portuguese language has several synonyms, such as certainty, confidence, credit, faith, and firmness, to name a few. However, when linked to the AI area, the term “Trust” is most often used to express the idea that we can trust in this process.
We quickly explored the AI literature to understand which words equivalent to “Trust” are being used in the English language. We found the following words: Trust/trustworthy, reliability/usability, attachment, and intention to trust. If, on the one hand, the term ‘Trust’ in the Portuguese language can present a set of ideas that expresses the entire dimension that studies in the field of AI require, on the other hand, it can represent some limits in its use, as it does not present a given precision. In the English language, although it offers several words to express the same idea, which can provide a more significant margin of interpretation of a given situation, however, the lack of terminological precision can represent a non-standardization in the epistemic community, which can create confusion in the production of knowledge in this field of work.
Trends in the study of the theme “Trust”
The trust theme has been studied since human beings began to live in society. Hence, we have the famous phrase according to whom, “Trust is the basis of every human relationship,” this phrase, in a broader sense, expresses precisely what is relevant in trust as a state and at the same time as an affirmation of a thing.
In AI studies, the issue of trust is necessary, especially in the treatment of personal user data. An example of one of the AI trends would be “Browsing Cookies,” which, in a broader sense, is not known how and in what way they are instrumentalized, which triggers distrust about the security of people’s data. Anyway, there is a World to be explored regarding this them; our goals were to try to provoke readers to think about the possibilities that the topic of “Trust” in IA studies can raise.
We would end this reflection by saying that the subject of trust in AI deserves to be better explored to understand the advantages and limits inherent in using this device, whether as a state or affirmation of some condition.
References:
Oxford Dictionary and Thesaurus.Biblioteca do Callipolis/Esag. Florianópolis, 1996.