As we interact more and more with generative artificial intelligence, it’s essential not only to understand the basics of this technology but also to be aware that it exposes us to certain cognitive phenomena that can influence the way we perceive and use it. Here are five of these phenomena to keep in mind during your next interactions with AI.
1. Algorithm bias
One might think that because generative AI is a “machine” that processes information using statistical approaches, its results are necessarily objective… However, the design of the algorithm (the procedure that guides the result) can reflect or even amplify biases (sexist, racist, etc.) present in the original data, which are produced by human beings and in which certain population groups may be under-represented. Users’ perceptions can thus be influenced by falsely logical, misleading or discriminatory results from a machine they believe to be objective, which can lead them to make potentially prejudicial or damaging decisions.
2. The black box effect
Generative AI can produce results of surprising complexity and coherence… Between their query and the result generated by the AI, users (whether novices or experts) are faced with a system whose operation remains opaque and which can leave them with a feeling of incomprehension or mystery. This phenomenon, known as the “black box effect,” can affect their confidence in the results.
3. Interpreting results
Depending on factors specific to the user, such as their knowledge, previous experiences and expectations of such a technology, the user may subjectively interpret the result of their AI query.
4. Overestimating or underestimating AI
Users may fall into the trap of overestimating the capabilities of AI, believing it to be capable of solving problems and answering queries without making mistakes or requiring human supervision, when, in fact, this is not the case. This cognitive bias, which leads us to believe that machines are infallible, is known as the “automation bias.” Note that the opposite can also occur if the user underestimates the real capabilities of AI to carry out critical tasks.
5. The social acceptability of AI
Generative AI is a veritable technological revolution whose impact on our daily lives, and even on our humanity, promises to be major. In particular, it raises several ethical issues (security, autonomy, privacy, democracy, etc.) that are far from settled. Yet its social acceptability may affect the way users approach it.
Related articles:
- Simplifying AI…and Making Cakes
- [VIDEO] Cognitive biases
- Cognitive biases in education: the Pygmalion effect
- 3 cognitive biases to know in education
- 10 categories of generative AI tools
- Digital Learning from the Perspective of Affective and Social Neuroscience
- Deciphering ChatGPT
- Deep learning 101
- Mini glossary of artificial intelligence
- Initiatives for a Responsible and Human-Centered Artificial Intelligence
- Mini glossary of technology in learning
- The Web from 1.0 to 4.0
- Artificial Intelligence: Test Your Knowledge!
Author:
Catherine Meilleur
Communication Strategist and Senior Editor @KnowledgeOne. Questioner of questions. Hyperflexible stubborn. Contemplative yogi
Catherine Meilleur has over 15 years of experience in research and writing. Having worked as a journalist and educational designer, she is interested in everything related to learning: from educational psychology to neuroscience, and the latest innovations that can serve learners, such as virtual and augmented reality. She is also passionate about issues related to the future of education at a time when a real revolution is taking place, propelled by digital technology and artificial intelligence.
