PJ Hagerty: AI and Ethics - What We SHOULD Be Pondering
summary
Defines foundation models and LLMs, warns of inherent bias and hallucination, and cites a study where LLM time‑series forecasts are wrong about 60–70% of the time and an industry survey claiming 77% of businesses report using “AI.” Recommends concrete tools and practices: AI Fairness 360, ARX anonymization, IBM Watsonx governance and Granite Guardian for prompt safety, plus smaller local models (Llama 3) to cut compute. Emphasizes guardrails, auditability, accountability, and choosing AI only where it outperforms simpler methods.
AI-written summary · about
chapters (26)
- 0:00Introduction
- 1:04Defining artificial intelligence and related terms
- 4:15Machine learning and deep learning
- 7:23Foundation models explained
- 9:09Ethics in AI
- 11:16Bias in AI models
- 12:41AI replacing jobs
- 15:30AI failures and misuse
- 17:13Marketing vs real AI
- 20:01Consumer AI limits
- 22:27AI manipulation risks
- 24:13AI resource consumption
- 25:57AI learning and emergence
- 26:58Facial recognition bias
- 28:00Ethical AI frameworks
- 29:48Human rights in AI
- 31:33Accountability in AI
- 34:00Tools for ethical AI
- 35:47AI hallucinations
- 37:32Ethical challenges and standards
- 38:57Responsibility in AI
- 40:02Q&A: AI hype
- 40:44Q&A: Ethical media concerns
- 41:26Q&A: AI guardrails
- 43:09Q&A: Legal AI gaps
- 45:20Q&A: Fighting unethical AI
related talks
29mResponsible AI Challenges and Recommendations
44mAI, Corporate Responsibility & Democratic Legitimacy
28mEthical AI?!
01h 01mPanel discussion: Practical and responsible applications of AI in businesses
32mSponsored Session: Beyond the Buzz: Navigating AIs Myths and Realities
49mKeeping Up with the Algorithms: A Journey in Artificial Intelligence
17mAI Law & Ethics: developing responsible AI systems
47mEthics in AI Panel