In this episode I speak with Dr. Luciano Siebert about how humans remain in control of Artificial Intelligence. This episode is in English (the next one will be in Dutch again).

We talk about killer drones, self-driving cars and AI-based hiring. All to explain why it is so important that humans remain in control (#responsibility) and what it means to do this in a meaningful way. For example, can we really expect someone to pay attention in a self-driving car for hours on end? Luciano uses his recently written paper to explain what MEANINGFUL human control means.

We talk about:

1. The perspective of engineers and if they are fit to design our future.
2. How the question of 'meaningful' human control originated around killer drones and autonomous weapons.
3. The difference between technology like airplanes and artificial intelligence.
4. The responsibility gap and why humans should always be resonsible (and not machines).
5. What it means to track and trace technology in order to meaningfully control it
6. Four aspects of meaningful human control Luciano defined in his paper:(1) We need to define in which moral situations a given AI should or should not be used. (2) Humans and AI should have some form of basic “understanding” of each other.(3) If a given person cannot do anything about what the AI system is doing, maybe it is not such a good idea to make this person responsible for whatever the system is doing.(4) AI systems need to be more transparent so whenever something bad happens (and it will happen!) we can understand what happened (and learn from the mistakes)

The transcript has been automatically generated and does not completely match the content but can still be helpful to look up a specific topic.

Addendum:

• L. Cavalcante Siebert et al., “Meaningful human control: actionable properties for AI system development,” AI Ethics, no. 0123456789, 2022, doi: 10.1007/s43681-022-00167-3.

Special Guest: Luciano Siebert.

Hosted on Acast. See acast.com/privacy for more information.