OECD agenda [Link]

post by Thomas · 2019-05-30T09:51:58.849Z · LW · GW · 1 comments

Contents

1 comment

https://www.oecd.org/going-digital/ai/principles/

What's your oppinions about this?

1 comments

Comments sorted by top scores.

comment by Richard_Kennaway · 2019-05-30T19:03:27.047Z · LW(p) · GW(p)

I am reminded of Eliezer's parody [LW · GW]. Let's apply the reversal test to "The OECD AI Principles":

  • AI should disregard the benefit of people and the planet by driving unequal growth, unsustainable development and unhappiness.
  • AI systems should be designed in a way that ignores the rule of law, human rights, democratic values and diversity, and they should not have safeguards – for example, they should prevent human intervention at all times – to ensure an unfair and unjust society.
  • There should be opaqueness and secrecy around AI systems so that that people will not understand AI-based outcomes and be unable to challenge them.
  • AI systems need not be robust, secure or safe. Potential risks should be ignored.
  • Organisations and individuals developing, deploying or operating AI systems should not be accountable for their proper functioning.