fightingAD 0.0.1

Automatic Differentiation Library

cs107-FinalProject

CS107 final project

Group Number: 10

Members:

• Manana Hakobyan
• Tale Lokvenec
• Hugo Fernandez-Montenegro
• Golo Feige (since Milestone 1)

Broader Impact and Inclusivity Statement

Broader Impact

This automatic differentiation package can be used in machine learning, numerical analysis, or economic optimization applications. Machine learning is a highly controversial field of science. Alan Turing already discussed consequences of machine learning in 1950, and Norbert Wiener draw an analogy between learning machines and intelligent slaves to warn against it ten years later. Wiener (1960) asked how "often in ancient times the clever Greek philosopher slave of less intelligent Roman slaveholder must have dominated the actions of his master rather than obeyed his wishes" to illustrate how powerful learning machines might become. Differentiation is also used in fields like ballistics or economic optimization. The latter often aims to reduce input costs such as human labor. These cost reductions can disrupt the careers of people. Nevertheless, automatic differentiation is a well-researched field and this package is unlikely to have a broader impact.

Inclusivity

GitFighters shares the commitment to "Diversity Inclusion & Belonging" defined by Harvard University as embracing "individuals from varied backgrounds, cultures, races, identities, life experiences, perspectives, beliefs, and values". As a group uniting different nationalities, backgrounds, and genders, GitFighters reflects inclusive diversity. GitFighters also make an effort to limit implicit exclusion in the development of this package. Contributing to this package requires programming knowledge and experience. This requirement excludes a large group of people. Therefore, GitFighters agreed on considering input from a broader audience and to implement changes together. Users, for instance, might have good ideas on how to improve the string-parsing or on how to import the package, but will not necessarily be able to implement these improvements. GitFighters are happy to consider suggestions and to implement them, if possible.

Sources:

Harvard University. About Diversity Inclusion & Belonging. https://dib.harvard.edu/about. Accessed 30 November 2020.

Turing, A. M. (1950). Computing Machinery and Intelligence. Mind, 59(236), 433-460.

Wiener, N. (1960). Some Moral and Technical Consequences of Automation. Science (American Association for the Advancement of Science), 131(3410), 1355-1358.