Ethics of algorithms - from theoretical discussion to practical application
summary
Panel argues concrete steps: require informed consent and right to explanation when automated systems make decisions; give oversight bodies access to code and resources for audits; add ethical training, background checks, and pairing concepts into developer QA; treat hiring data and predictive policing as concrete risk zones (Clearview and predictive policing reinforce harmful correlations); build open source tools like KI Park’s ethics lab and the battlesman foundation’s algorithmic guidance as process standards rather than universal technical specs. These measures aim to prevent harm and make ethics auditable.
AI-written summary · about
chapters (14)
- 0:00Introduction to ethics in AI
- 2:07Defining ethics of intelligence
- 3:53Sources of ethical problems in AI
- 4:56Ethically excluded AI applications
- 6:00Bias in algorithmic decision-making
- 7:47Strong AI and ethical focus
- 9:53Weak AI and ethical limits
- 11:17Organizational responsibility for trustworthy AI
- 13:22Building trustworthy AI systems
- 14:46Legal and moral obligations in AI
- 16:53Predictive policing and ethical dilemmas
- 17:40Racism and predictive policing systems
- 20:45Need for new norms in the digital age
- 23:11Closing statements on AI ethics
related talks
56mEthics in Machine Learning Panel
16mExistential Engineering or How Ethics Precedes Error Handling
47mThe Ethics of Software - some practical considerations
01h 01mDoing Data Science with Ethics
47mAI Ethics, Strategic Decisioning and Game Theory
47mEthics in AI Panel
31mEthics in technology: History, practice, and mitigation
33mEthics in Machine Learning Panel Q&A