What Are the Key Terms in AI Ethics?

  • Thread starter Thread starter Tom.G
  • Start date Start date
  • Tags Tags
    Ai Ethics
Click For Summary
SUMMARY

The discussion centers on key terms in AI ethics, highlighting definitions such as "audit," "diversity, equity, and inclusion," and "impact assessment." An audit is defined as an external review of an AI system for transparency without necessitating actual changes. Diversity, equity, and inclusion refer to the superficial hiring of marginalized groups, often without genuine commitment to their input. Impact assessment is an internal review aimed at acknowledging potential downsides of AI systems while avoiding substantive modifications.

PREREQUISITES
  • Understanding of AI system audits
  • Familiarity with diversity, equity, and inclusion principles in tech
  • Knowledge of impact assessments in organizational contexts
  • Awareness of ethical implications in AI development
NEXT STEPS
  • Research the process of conducting AI system audits
  • Explore best practices for implementing genuine diversity, equity, and inclusion in tech teams
  • Learn about comprehensive impact assessment methodologies for AI
  • Investigate the ethical frameworks guiding AI development and deployment
USEFUL FOR

AI developers, ethics researchers, organizational leaders, and anyone involved in the governance and ethical considerations of AI technologies.

Tom.G
Science Advisor
Gold Member
Messages
5,661
Reaction score
4,506
Here are a few excerpts from an article I stumbled across on
https://www.technologyreview.com/2021/04/13/1022568/big-tech-ai-ethics-guide/

The complete list of definitions runs about 50 items. It is sadly true to some extent, and a bit of a change-of-pace from this site.

audit (n) - A review that you pay someone else to do of your company or AI system so that you appear more transparent without needing to change anything. See impact assessment.

diversity, equity, and inclusion (ph) - The act of hiring engineers and researchers from marginalized groups so you can parade them around to the public. If they challenge the status quo, fire them.

impact assessment (ph) - A review that you do yourself of your company or AI system to show your willingness to consider its downsides without changing anything. See audit.
 
Reply
  • Like
Likes Keith_McClary and 256bits
Computer science news on Phys.org
I notice more often now that when something 'screws up' , there is usually a 'contractor' around who wasn't following 'protocol'. The solution has been that the business agreement has been 'terminated', thus making everyone feel more secure and welcome using the service.
 
256bits said:
there is usually a 'contractor'
"Uberization"?

You just need to train a classifier on ethical and unethical ... .
ai_methodology_2x.png


2451
 
Reply
  • Like
  • Haha
Likes 256bits, Tom.G and collinsmark
The bitter biting rapier-like wit of the authors obscures fundamental problems. Can one expect ethical behavior without morality? How do we replace morals based on superstition and tradition with reason?
 

Similar threads

Current and Future AI Threats to Humanity and the Human Response
  • · Replies 10 ·
Replies
10
Views
5K