Artificial Intelligence   

Practice of Ethical AI

Ethics are one of the concern of people around the world and people are very sensitive about their ethical values. We think about our ethical practices in our life and ethical practices of our work place. Ethical practices are involved in all the decisions we made and all the things we do and chose not to do. Artificial intelligence is part of our lives now and helping us in making better decisions. They are making our lives better and improving our decisions and work environment. For example Amazon and Google helping us in identifying what we want with the help of artificial intelligence. AI learn from us and about us every time we interact with them in order to improve ethical decision making.

According to study it is shown that if we get our ethics right then machines can never be able to take over the world. We should focus practicing ethical values and avoid hurting other people or making inhuman decisions. Soon we will be able to fully automate the healthcare sector and be able to make ethical decisions. AI will be able to help doctors in making right ethical decision according to the situations. On the other hand if we implement or automate our system without any ethical practices then it would result it brutal and opaque decisions.

Working on ethical AI

According to the recent interview from House of Lords Select Committee it was shown that UK is working continuously on the ethical practices for AI and developing a specific role as pioneer for ethical AI. Ethical issue of artificial intelligence are now being moved toward mainstream consciousness. Recent ethical issue of Facebook data sharing ring the alarms of ethical norms and values of consumers and artificial intelligence. Now Facebook internal ethics unit focus on the ethical practices and issues of artificial intelligence. Microsoft also used artificial intelligence in order to increase sales above ethical concerns. Sometimes organizations run and implement AI in order to be the first in the market and to capture the advantage while ignoring the ethical concerns of AI. After some time those errors and mistakes start surfacing just like the case of NHS breast cancer screening error.

IEEE and AI ethics

Institute of Electrical and Electronics Engineers have been developing technology standards for decades. During 2016 IEEE developed the first version of ethical standards for autonomous and artificial intelligent technology. These ethical standards can help a lot of businesses to solve their problems and enable them to deal with issue in future precisely. According to ethical standard P7000 there is a specific approach in the AI system which helps in identification and analyses of potential ethical issues in the software or system. After identification values based system designed methods address those ethical considerations on different stages of development. This helps in avoiding the negative consequences of the AI implementation and improve innovations and effectiveness.

Transparency of autonomous systems

According to the P7001 we follow the standard of autonomous systems and their transparency. Artificial intelligence helps in assessing the actions of system and take proper action accordingly. It also makes it easier for the user to understand why AI is better at decision making in different situations. AI also provides accountability and transparency for the systems in order to improve their performance and provide guidance. For example incorporating data recorder in autonomous car or access data through sensors of devices.   

Data privacy

As part of the artificial intelligence systems we follow the standard of P7002 which employs that how we should manage the privacy concerns for all the software and systems that collect personal data. Collecting personal data without the consent of users is unethical practice that’s why P7002 was introduced in order to make artificial intelligence more ethical. This will help designers through multiple ways by identifying and measuring the privacy control in the system that utilize privacy impact assessment.

Algorithmic bias

We use different types of algorithms in systems and those algorithms can be developed to be biased. According to P7003 code of the ethical AI it uses algorithms for the artificial intelligent and autonomous systems with specific protocols in order to avoid the negative biases in the codes. These biases can include multiple usage of subjective or incorrect interpretations of the data. This protocol allows artificial intelligence to follow specific steps in order to eliminate negative biases from the system. This helps in making artificial intelligence ethical and follow ethical practices.

Personal data AI agent working group

We follow another standard known as P7006 which is related to the personal data and how artificial intelligence made it ethical and easier for us. There was a concern that machines will be able to make decisions without any help or human input. This standard will provide individuals with tool to create their own terms and conditions in order to protect their data. AI agent will provide the technological advancement and tool to people in order to manage their privacy and control their identity in the virtual world.

Ethically driven robotics and automation systems

As we have seen before each standard is making artificial intelligence more autonomous and ethical in order to make better ethical decision in different situations. According to the standard of P7007 there are number of ontologies which have multiple abstraction levels and involve many concepts, axioms and definitions which are essential to make ethical methodology for the robotics and autonomous systems. It makes it possible for the artificial intelligence to make better and ethical decisions.

#Artificial Intelligence 

By Anne Leckie Technology Expert, Quantega
7 months ago


Jobs Related to Artificial Intelligence Domain

  • Creative Director - Hyderabad
  • AI Scientist - Hyderabad
  • See Jobs