Facebook AI: Open frameworks and responsible development

The bit at the end about social responsibility seems bolted-on and gratuitous.

How does AI even factor into the blood donations example they cite? It sounds like a simple database query, not AI.

Also the talk about what they are doing internally to "ethically use AI" seem pretty weak. A lot of "we can do this and that" without any "we are doing the following and it is making a difference". And saying that one of their key strategies to ensure ethical AI is to hire a diverse workforce seems again to be a bit thin on details and supporting evidence.