25.09.2018
 Let's say an AI fails at a certain task. What's the worst that could happen? Adversarial effects could lead to people getting seriously injured.Prior to the misinterpretation of singals or patterns the AI must be deceived during the machine learning process - purposefully or not- that's another story. AI's may need a technical certificate in the future. That's why our project "EmmA" relies on independent ethical support from the ELSI. Project EmmA: https://www.interaktive-technologien.de/projekte/emma |