Page 331 - Artificial Intellegence_v2.0_Class_11
P. 331

Establish a Mechanism of Global Governance
                 Ethical  AI  governance  bodies  on  global  and  regional  levels  need  to  be  established.  These  bodies  must  include  all
                 stakeholders—AI designers, manufacturers, owners, developers, researchers, employers, lawyers, CSOs, and trade unions.


                 Prohibition of Assigning Responsibilities to Robots
                 The design and operation of robots should comply with existing laws and fundamental rights and freedoms, including
                 privacy, as much as possible.

                 Prohibit AI Arms Race
                 Lethal autonomous weapons, as well as cyber warfare, should be prohibited.



                                Reboot


                      Give any two examples where AI is helping people.
                      ____________________________________________________________________________________________________________________
                      ____________________________________________________________________________________________________________________






                         What is Bias?


                 Bias is any prejudice against individuals or groups, especially in ways that are considered unfair. "Bias in AI" has long been
                 a key area of research and attention in the machine learning community. It is a term used to describe situations where
                 ML-based data analysis systems are biased against certain groups of people. These prejudices usually reflect the prevailing
                 social prejudices about race, gender, biological sex, age, and culture. AI systems learn to make decisions based on training
                 data, which may include biased human decisions or reflect historical or social inequities.


                 Types of Bias and How It Influences AI Based Decisions
                 The most popular classification of bias in artificial intelligence divides it into three categories: algorithmic, data, and human.
                 This classification uses the source of prejudice as the primary criterion. However, as human bias is the root of and exceeds
                 the other two, researchers and practitioners of AI urge to watch out for the latter. The most typical forms of AI bias that
                 seep into the algorithms are listed below:




                                                                Reporting






                                                                 Types
                                               Implicit                            Selection
                                                                 of Bias





                                                                 Group
                                                               Attribution



                                                                                   AI Values (Bias Awareness)      329
   326   327   328   329   330   331   332   333   334   335   336