Artificial Intelligence and A.I. Ethics
What is A.I.?
Artificial intelligence is the replication of human intelligence and senses in a machine or software.1 The five human senses [touch, sight, taste, and smell] and cognition [the ability to process information pertaining to the aforesaid senses] are capabilities in which a machine can mimic and excel.
How does A.I. work?
Artificial intelligence utilizes ingested, labeled data on which to train and to make predictions.1 An image classifier [artificial intelligence that is utilized to identify or classify images] is used to classify cats and dogs. Simplistically, the image-classifier algorithm trains on images of cats and dogs and through training, it recognizes or predicts if a future image is a cat or a dog.2
Why is A.I. important?
Artificial intelligence is important because it has integrated into society.1 Healthcare, finance, law, and other domains utilize artificial intelligence to automate and quicken tasks. For example, healthcare may use an image-classifier algorithm to detect anomalies in X-Rays for disease screening.3
What are A.I. Ethics?
A.I. ethics is the morality and principles underpinning artificial intelligence implementation. Artificial intelligence, according to science fiction writer Isaac Asimov, must abide by The Three Laws of Robotics. The first law disallows robots from actively and passively [through negligence] hurting humans. The second law is that robots must obey humans, unless the task conflicts the first law. Lastly, the third law permits robots to protect themselves, as long as it is complementary to the aforesaid two laws. The Three Laws of Robotics are precepts upon which organizations enact policies. Organizations, such as Center for AI and Digital Policy, The Stanford Institute for Human-Centered Artificial Intelligence, and AI Now Institute, investigate algorithms and/or policies to guardrail the utilization of A.I.4
Why A.I. Ethics?
Ethical A.I. is necessary in all domains because without it, A.I. causes harmful consequences. For example, the COMPAS algorithm has been used to predict criminals’ recidivism, though it has been found to be biased to people of color. Amazon Inc., in 2018, penalized résumés that had words, like women’s, instead of scraping for pertinent skills. Lastly, A.I. has denied 80% of African-Americans, 40% of Latino-Americans, and 70% of Native Americans from mortgages, compared to White-Americans.5
Inequity is pervasive, and A.I. can deepen human biases; however, A.I. does not have to continue to be weaponized but be a shield to protect and prosper humanity. Therefore, ethical A.I. is not an option but a necessity.
Endnotes:
1 Nicole Laskowski and Linda Tucci, “artificial intelligence”, TechTarget, accessed Jan 9, 2024,
https://www.techtarget.com/searchenterpriseai/definition/AI-Artificial-Intelligence
2 “Image Classification", Hugging Face, accessed Jan 9, 2024,
https://huggingface.co/tasks/image-classification
3 “Medical Image Classification, Papers with Code, accessed Jan 9, 2024,
https://paperswithcode.com/task/medical-image-classification
4 George Lawton and Ivy Wigmore, “AI ethics (AI code of ethics)”, TechTarget, accessed Jan 9, 2024,
https://www.techtarget.com/whatis/definition/AI-code-of-ethics
5 Alexander S. Gillis and Mary K. Pratt, “machine learning bias (AI bias)”, TechTarget, accessed Jan 13, 2024,