FAccT
In plain wordsA yearly academic conference on the fairness, accountability and transparency of computer systems, where much influential research on AI bias and ethics has been published.
The ACM Conference on Fairness, Accountability, and Transparency, the main venue for research on how algorithmic systems affect people and institutions, whose authors mix computer science with law, philosophy and the social sciences. Its proceedings include Gender ShadesGender Shades, prop. n.: A 2018 study showing that commercial face-analysis software guessed the gender of darker-skinned women far less accurately than that of lighter-skinned men. (2018), model cardsmodel card, n.: A short document published with an AI model explaining what it is for, how it was trained and tested, how well it works for different groups of people, and its known limits. (2019) and 'On the Dangers of Stochastic Parrots' (2021), the paper behind the phrase stochastic parrotstochastic parrot, n.: A critical label for AI language systems, likening them to parrots that echo patterns in the text they were trained on, fluently but with no grasp of what the words mean., whose internal review at Google preceded the departures of its co-authors Timnit Gebru and Margaret Mitchell from the company. Topics range from algorithmic fairnessalgorithmic fairness, n.: The effort to make automated decisions fair: to say in numbers what fair treatment means, measure whether a system achieves it, and correct the system if it does not. metrics and audits to labor, surveillance and the political economy of AI. A sister conference, AIES (AI, Ethics, and Society), is run by AAAI and the ACM.
Origin
Grew out of the FAT/ML workshops held alongside machine learning conferences from 2014; the first stand-alone conference, FAT* 2018, met in New York, and it took the name ACM FAccT in 2020.
Related
- algorithmic fairnessalgorithmic fairness, n.: The effort to make automated decisions fair: to say in numbers what fair treatment means, measure whether a system achieves it, and correct the system if it does not.
- algorithmic biasalgorithmic bias, n.: Unfairness in software that judges or serves people: it consistently produces worse results for some groups, such as women, older people or ethnic minorities, often because it learned from skewed or incomplete records.
- Gender ShadesGender Shades, prop. n.: A 2018 study showing that commercial face-analysis software guessed the gender of darker-skinned women far less accurately than that of lighter-skinned men.
- model cardmodel card, n.: A short document published with an AI model explaining what it is for, how it was trained and tested, how well it works for different groups of people, and its known limits.
- stochastic parrotstochastic parrot, n.: A critical label for AI language systems, likening them to parrots that echo patterns in the text they were trained on, fluently but with no grasp of what the words mean.
- AI ethicsAI ethics, n.: The study of what is right and wrong in how AI is built and used: who it may harm, who is responsible when it fails, and which values it should respect.