Machine Learning and Irresponsible Inference: Morally Assessing the Training Data for Image Recognition Systems

Research output: Chapter in Book/Report/Conference proceedingChapterAcademicpeer-review

33 Downloads (Pure)

Abstract

Just as humans can draw conclusions responsibly or irresponsibly, so too can computers. Machine learning systems that have been trained on data sets that include irresponsible judgments are likely to yield irresponsible predictions as outputs. In this paper I focus on a particular kind of inference a computer system might make: identification of the intentions with which a person acted on the basis of photographic evidence. Such inferences are liable to be morally objectionable, because of a way in which they are presumptuous. After elaborating this moral concern, I explore the possibility that carefully procuring the training data for image recognition systems could ensure that the systems avoid the problem. The lesson of this paper extends beyond just the particular case of image recognition systems and the challenge of responsibly identifying a person’s intentions. Reflection on this particular case demonstrates the importance (as well as the difficulty) of evaluating machine learning systems and their training data from the standpoint of moral considerations that are not encompassed by ordinary assessments of predictive accuracy.
Original languageEnglish
Title of host publicationOn the Cognitive, Ethical, and Scientific Dimensions of Artificial Intelligence
Subtitle of host publicationThemes from IACAP 2016
EditorsDon Berkich, Matteo Vincenzo d'Alfonso
PublisherSpringer
Pages265-282
ISBN (Electronic)978-3-030-01800-9
ISBN (Print)978-3-030-01799-6
DOIs
Publication statusPublished - 2019

Publication series

NamePhilosophical Studies Series
ISSN (Print)0921-8599
ISSN (Electronic)2542-8349

Fingerprint

Image recognition
Learning systems
Computer systems
Machine Learning
Inference
Intentions
Person
Learning Systems

Keywords

  • Machine Learning
  • Image Recognition
  • training data
  • data ethics

Cite this

King, O. C. (2019). Machine Learning and Irresponsible Inference: Morally Assessing the Training Data for Image Recognition Systems. In D. Berkich, & M. V. d'Alfonso (Eds.), On the Cognitive, Ethical, and Scientific Dimensions of Artificial Intelligence: Themes from IACAP 2016 (pp. 265-282). (Philosophical Studies Series). Springer. https://doi.org/10.1007/978-3-030-01800-9_14
King, Owen Christopher. / Machine Learning and Irresponsible Inference : Morally Assessing the Training Data for Image Recognition Systems. On the Cognitive, Ethical, and Scientific Dimensions of Artificial Intelligence: Themes from IACAP 2016. editor / Don Berkich ; Matteo Vincenzo d'Alfonso. Springer, 2019. pp. 265-282 (Philosophical Studies Series).
@inbook{41863e4448e3495885127638c1ea418f,
title = "Machine Learning and Irresponsible Inference: Morally Assessing the Training Data for Image Recognition Systems",
abstract = "Just as humans can draw conclusions responsibly or irresponsibly, so too can computers. Machine learning systems that have been trained on data sets that include irresponsible judgments are likely to yield irresponsible predictions as outputs. In this paper I focus on a particular kind of inference a computer system might make: identification of the intentions with which a person acted on the basis of photographic evidence. Such inferences are liable to be morally objectionable, because of a way in which they are presumptuous. After elaborating this moral concern, I explore the possibility that carefully procuring the training data for image recognition systems could ensure that the systems avoid the problem. The lesson of this paper extends beyond just the particular case of image recognition systems and the challenge of responsibly identifying a person’s intentions. Reflection on this particular case demonstrates the importance (as well as the difficulty) of evaluating machine learning systems and their training data from the standpoint of moral considerations that are not encompassed by ordinary assessments of predictive accuracy.",
keywords = "Machine Learning, Image Recognition, training data, data ethics",
author = "King, {Owen Christopher}",
year = "2019",
doi = "10.1007/978-3-030-01800-9_14",
language = "English",
isbn = "978-3-030-01799-6",
series = "Philosophical Studies Series",
publisher = "Springer",
pages = "265--282",
editor = "Don Berkich and d'Alfonso, {Matteo Vincenzo}",
booktitle = "On the Cognitive, Ethical, and Scientific Dimensions of Artificial Intelligence",

}

King, OC 2019, Machine Learning and Irresponsible Inference: Morally Assessing the Training Data for Image Recognition Systems. in D Berkich & MV d'Alfonso (eds), On the Cognitive, Ethical, and Scientific Dimensions of Artificial Intelligence: Themes from IACAP 2016. Philosophical Studies Series, Springer, pp. 265-282. https://doi.org/10.1007/978-3-030-01800-9_14

Machine Learning and Irresponsible Inference : Morally Assessing the Training Data for Image Recognition Systems. / King, Owen Christopher.

On the Cognitive, Ethical, and Scientific Dimensions of Artificial Intelligence: Themes from IACAP 2016. ed. / Don Berkich; Matteo Vincenzo d'Alfonso. Springer, 2019. p. 265-282 (Philosophical Studies Series).

Research output: Chapter in Book/Report/Conference proceedingChapterAcademicpeer-review

TY - CHAP

T1 - Machine Learning and Irresponsible Inference

T2 - Morally Assessing the Training Data for Image Recognition Systems

AU - King, Owen Christopher

PY - 2019

Y1 - 2019

N2 - Just as humans can draw conclusions responsibly or irresponsibly, so too can computers. Machine learning systems that have been trained on data sets that include irresponsible judgments are likely to yield irresponsible predictions as outputs. In this paper I focus on a particular kind of inference a computer system might make: identification of the intentions with which a person acted on the basis of photographic evidence. Such inferences are liable to be morally objectionable, because of a way in which they are presumptuous. After elaborating this moral concern, I explore the possibility that carefully procuring the training data for image recognition systems could ensure that the systems avoid the problem. The lesson of this paper extends beyond just the particular case of image recognition systems and the challenge of responsibly identifying a person’s intentions. Reflection on this particular case demonstrates the importance (as well as the difficulty) of evaluating machine learning systems and their training data from the standpoint of moral considerations that are not encompassed by ordinary assessments of predictive accuracy.

AB - Just as humans can draw conclusions responsibly or irresponsibly, so too can computers. Machine learning systems that have been trained on data sets that include irresponsible judgments are likely to yield irresponsible predictions as outputs. In this paper I focus on a particular kind of inference a computer system might make: identification of the intentions with which a person acted on the basis of photographic evidence. Such inferences are liable to be morally objectionable, because of a way in which they are presumptuous. After elaborating this moral concern, I explore the possibility that carefully procuring the training data for image recognition systems could ensure that the systems avoid the problem. The lesson of this paper extends beyond just the particular case of image recognition systems and the challenge of responsibly identifying a person’s intentions. Reflection on this particular case demonstrates the importance (as well as the difficulty) of evaluating machine learning systems and their training data from the standpoint of moral considerations that are not encompassed by ordinary assessments of predictive accuracy.

KW - Machine Learning

KW - Image Recognition

KW - training data

KW - data ethics

U2 - 10.1007/978-3-030-01800-9_14

DO - 10.1007/978-3-030-01800-9_14

M3 - Chapter

SN - 978-3-030-01799-6

T3 - Philosophical Studies Series

SP - 265

EP - 282

BT - On the Cognitive, Ethical, and Scientific Dimensions of Artificial Intelligence

A2 - Berkich, Don

A2 - d'Alfonso, Matteo Vincenzo

PB - Springer

ER -

King OC. Machine Learning and Irresponsible Inference: Morally Assessing the Training Data for Image Recognition Systems. In Berkich D, d'Alfonso MV, editors, On the Cognitive, Ethical, and Scientific Dimensions of Artificial Intelligence: Themes from IACAP 2016. Springer. 2019. p. 265-282. (Philosophical Studies Series). https://doi.org/10.1007/978-3-030-01800-9_14