What is Open Ethics Vector

Why do we need an Open Ethics Vector

#AIFORGOOD… But who decides what’s “good” and what’s not?

What we know is that different communities can have different cultural codes. Indeed, culture is defined by our set of values, guiding behaviors and attitudes towards religion, gender, relationships, money, food, or health. These sets differ from one society to another. Can these values be aligned between AI systems and the users that belong to a specific culture? Instead of debating about what’s good, in Open Ethics initiative we start from the basic principles. A set of such principles defines an Open Ethics Vector (OEV) which reflects the values about how data-driven decisions are made. Disclosing these value sets, making them open, can help users to learn which apps are best for them.

Content

  1. OEV Use-cases
    1. Product-owners
    2. End-users
    3. Integrators
  2. OEV Vision
  3. OEV Implementation

How does Open Ethics Vector work: use-cases

AI product owners get their products listed on the platform and are able to showcase their algorithmic decision-making approaches.

AI product-owners can list their product on the Open Ethics platform (we are building it) and receive product feedback. To do that, product owners can register on the platform and describe how their product works by uploading Open Ethics Vector. This Vector specifies which ethical choices were made in training data that was used for development of the machine learning model and in boundaries that restricting the decision space.

End-users can choose the AI solution based on their personal “vector of preferences”.

The users can filter products based on personal ethical preferences, and to provide feedback to products they use.

Solution providers/integrators can seamlessly access information about their vendor’s AI products.

Integrators of AI-powered technology could assess how their vendors are handling AI’s ethical choices thanks to Open Ethics Vector (OEV). They can use machine-readable OEV schema to inform more complex decisions of their solutions. They can also seamlessly incorporate information about AI vendors into their solution’s terms and privacy policy if needed.

The vision for the Open Ethics Vector

Decisions reflect our point of view (POV). If we can make sure that POVs are explicit, and our tools do not lead to ethically conflicting decisions, we can achieve better personal outcomes, higher societal alignment, and adhesion.
Open Ethics Vector animated

Open Ethics Vector Implementation

Open Ethics API in OpenAPI v3.0 format will be published soon