Lecture 14. Ethics in ML & Outlook - clairedavid/ml_in_hep GitHub Wiki

Applications to question

Aurelien Barrau commente les utilisations :

  • ciblage publicitaire
  • ciblage de contenu
  • standardisation du recrutement via CV lus par des machines
  • finance : personne ne sait comment ils fonctionnent

Source: Univershifte YouTube at t = 2047

chatGPT

Talk to dead people

Canada imitation of fiancee (!)

Articles/Websites to study

People/thinkers on Ethics

  • Kate Crawford and Ryan Calo: Crawford and Calo are co-founders of the AI Now Institute at New York University, which focuses on the social and ethical implications of AI. They have written extensively on issues such as bias in machine learning and the need for accountability in AI systems.

  • Timnit Gebru: Gebru is a research scientist at Google and co-founder of the Black in AI organization. She has done pioneering work on issues of bias and fairness in machine learning and has advocated for more diverse and inclusive approaches to AI development.

  • Safiya Noble: Noble is a professor at the University of California, Los Angeles, and the author of "Algorithms of Oppression," a book about the ways in which search engines and other digital platforms reproduce and reinforce social inequalities. She has written extensively about the social and political implications of AI and machine learning.

  • John Danaher: Danaher is a philosopher and legal scholar who has written extensively about the ethical implications of emerging technologies, including AI and machine learning. He has focused on issues such as the impact of automation on employment, the ethics of autonomous weapons, and the role of ethical frameworks in guiding the development of AI.

  • Barbara Grosz: Grosz is a computer scientist and philosopher who has done pioneering work on the development of computational models of conversation and collaboration. She has also written extensively about the ethical implications of AI and machine learning, including issues of bias and fairness, and the need for transparency and accountability in these systems.

  • Alexeï Grinbaum. "Le phénomène Chat GPT" on France Inter 13 Dec 2022 They won't take control of us as they do not define their own goals. Finality/end goal is always defined by a human being. There is no consciousness. No 'thoughts', period. Europe is working on a new law on AI. Replace humans? No but help yes/assist. Judge+laywer on jurudiction texts, awesome help. App for student support (remise a niveau personalise). CEA works on med school students. For autistic kids: these machines are never getting tired. Medical: prediction of heart attack. 4th version of ChatGPT next year to be "500 times more powerful". Are you worried? AG: no, at all. I am fascinated. What worries me is: are we going to advance as fast with the control.

Outlook:

Working with Real Data : p.38 Geron's book! Lots of links.

data kind!