Types of AI Ethics Paper
Political philosopher Dr Annette Zimmermann published a very piercing cartoon on Twitter. It is a pastiche of the XKCD cartoon Types of Scientific Paper, and creatively reimagines titles of papers that expose flawed thinking around AI Ethics that we come across in the industry on day-to-day basis.
I have provided an image description for the image below to make the content available for more people.
Image description
Twelve imagined research paper headlines placed in a 3x4 grid, sketched in the form of symbolic cover pages.
The image title is: Types of AI Ethics Paper
In order from left to right, row by row from the top, the headlines are:
- My algorithm is ~fair~
- We're a corporate research group and our product is very, very ethical
- We used facial recognition in a cool new way and we have pure hearts and good intentions
- If it's legal it's ethical: Case closed
- Cars are kind of like trolleys, so all ethical problems involving self-driving cars are structured like the trolley problem
- My critics are losers who hate innovation
- Why ethics (and philosophy more broadly) can't possibly adress questions of power
- Theory is boring so we came up with a very practical ethics ~framework~ and it has 39238 parameters
- It's not unethical if it's transparent, d'uh
- We are 500 scientists and we'd like you to know that we too work on AI ethics now
- We are philosophers and we refuse to sully our reputation with empirical research
- We are once again requesting that you ~regulate~ tech (but we won't tell you how)
I think Dr Zimmermann did a brilliant job of covering many of the misconceptions that exist when addressing ethics in a digital context, and I will surely be challenging my students and clients to come up with more. These provide a valuable artefact for discussions in an educational setting.
Update: Types of AI For The Developing World Paper
I also came across this spot-on version of the meme by cognitive science PhD researcher Abeba Birhane. Image description below.
Image description
Twelve imagined research paper headlines placed in a 3x4 grid, sketched in the form of symbolic cover pages.
The image title is: Types of AI For The Developing World Paper
In order from left to right, row by row from the top, the headlines are:
- We're coming to save you but this time we have AI
- Leapfrogging Africa into Utopia using state-of-the-art ML
- We used satellite image data to map poverty
- The answer to all Africa's problems lies in more data
- We don't know how to pronounce Lagos but we'll build you a cool geo-map of the city.
- Fintech
- More fintech
- We, in Silicon Valley, keep awake at night worrying about the unbanked women of Africa
- Food scarcity? We have an AI solution
- The refugee problem: We used facial recognition to solve it
- Based on a conversation with my Ghanaian friend, here is what I think Africa needs
- Of course Facebook is the Internet
More versions of this meme:
