An Intelligence in Our Image: The Risks of Bias and Errors in Artificial Intelligence

Machine learning algorithms and artificial intelligence influence many aspects of life today and have gained an aura of objectivity and infallibility. The use of these tools introduces a new level of risk and complexity in policy. This report illustrates some of the shortcomings of algorithmic decisionmaking, identifies key themes around the problem of algorithmic errors and bias, and examines some approaches for combating these problems.

[1]  A. Turing.  On Computable Numbers, with an Application to the Entscheidungsproblem. , 1937 .

[2]  Giuseppe Nuti,et al.  Algorithmic Trading , 2011, Computer.

[3]  L. S. Shapley,et al.  College Admissions and the Stability of Marriage , 2013, Am. Math. Mon..

[4]  Carlos Eduardo Scheidegger,et al.  Certifying and Removing Disparate Impact , 2014, KDD.

[5]  J. Manyika,et al.  Are you ready for the era of ‘big data’? , 2010 .

[6]  George G. Woodworth,et al.  Identifying Comparatively Excessive Sentences of Death: A Quantitative Approach , 1980 .

[7]  Tom L. Beauchamp,et al.  An enquiry concerning human understanding : a critical edition , 2000 .

[8]  O. Gandy.  Engaging rational discrimination: exploring reasons for placing regulatory constraints on decision support systems , 2010, Ethics and Information Technology.

[9]  Latanya Sweeney,et al.  Discrimination in online ad delivery , 2013, CACM.

[10]  B. Russell,et al.  Problems Of Philosophy , 2004, Synthese.

[11]  Frank A. Pasquale.  The Black Box Society: The Secret Algorithms That Control Money and Information , 2015 .

[12]  Alessandro Vespignani,et al.  Google Flu Trends Still Appears Sick: An Evaluation of the 2013-2014 Flu Season , 2014 .

[13]  Frank A. Pasquale,et al.  [89WashLRev0001] The Scored Society: Due Process for Automated Predictions , 2014 .

[14]  Chris Arney.  Probably Approximately Correct: Nature's Algorithms for Learning and Prospering in a Complex World , 2014 .

[15]  John Schulman,et al.  Concrete Problems in AI Safety , 2016, ArXiv.

[16]  Cynthia Dwork.  International Conference on Theory and Applications of Models of Computation , 2008 .

[17]  Abdul V. Roudsari,et al.  Automation bias: a systematic review of frequency, effect mediators, and mitigators , 2012, J. Am. Medical Informatics Assoc..

[18]  Danah Boyd.  The politics of "real names" , 2012, CACM.

[19]  Toniann Pitassi,et al.  Fairness through awareness , 2011, ITCS '12.

[20]  Helen Nissenbaum,et al.  Bias in computer systems , 1996, TOIS.

[21]  Marvin Minsky,et al.  Steps toward Artificial Intelligence , 1995, Proceedings of the IRE.

[22]  Joaquin Quiñonero Candela,et al.  Counterfactual reasoning and learning systems: the example of computational advertising , 2013, J. Mach. Learn. Res..

[23]  Helen Nissenbaum,et al.  Big data's end run around procedural privacy protections , 2014, Commun. ACM.

[24]  Cynthia Dwork,et al.  An Ad Omnia Approach to Defining and Achieving Private Data Analysis , 2007, PinKDD.

[25]  Vitaly Shmatikov,et al.  Myths and fallacies of "Personally Identifiable Information" , 2010, Commun. ACM.

[26]  Paul Ohm.  Broken Promises of Privacy: Responding to the Surprising Failure of Anonymization , 2009 .

[27]  Christopher T. Lowenkamp,et al.  False Positives, False Negatives, and False Analyses: A Rejoinder to "Machine Bias: There's Software Used across the Country to Predict Future Criminals. and It's Biased against Blacks" , 2016 .

[28]  Malte Ziewitz.  Governing Algorithms , 2016 .

[29]  Simon DeDeo,et al.  "Wrong side of the tracks": Big Data and Protected Categories , 2014, ArXiv.

[30]  K. Crawford.  Think Again: Big Data , 2013 .

[31]  Andrew D. Selbst,et al.  Big Data's Disparate Impact , 2016 .

[32]  Nicholas Diakopoulos,et al.  Accountability in algorithmic decision making , 2016, Commun. ACM.

[33]  Karrie Karahalios,et al.  Auditing Algorithms : Research Methods for Detecting Discrimination on Internet Platforms , 2014 .

[34]  Felix Salmon.  The formula that killed Wall Street , 2012 .

[35]  George G. Woodworth,et al.  Comparative Review of Death Sentences: An Empirical Study of the Georgia Experience , 1983 .

[36]  David H. Autor,et al.  The polarization of job opportunities in the U.S. labor market: implications for employment and earnings , 2011 .

[37]  D. Citron.  Technological Due Process , 2007 .

[38]  Susan Athey,et al.  Machine Learning and Causal Inference for Policy Evaluation , 2015, KDD.

[39]  Joanna Bryson,et al.  Semantics derived automatically from language corpora contain human-like biases , 2016, Science.

[40]  Alan M. Turing,et al.  Computability and λ-definability , 1937, Journal of Symbolic Logic.

[41]  J. Wilson,et al.  BROKEN WINDOWS: THE POLICE AND NEIGHBOURHOOD SAFETY , 1982 .

[42]  Demis Hassabis,et al.  Mastering the game of Go with deep neural networks and tree search , 2016, Nature.