AI in Scientific Imaging: Drawing on Astronomy and Nanotechnology to Illustrate Emerging Concerns About Generative Knowledge
AI in Scientific Imaging: Drawing on Astronomy and Nanotechnology to Illustrate Emerging Concerns About Generative Knowledge
Abstract
Recent advances in AI technology have enabled an unprecedented level of control over the processing of digital images. This breakthrough has sparked discussions about many potential issues, such as fake news, propaganda, the intellectual property of images, the protection of personal data, and possible threats to human creativity. Susan Sontag (2005 [1977]) recognized the strong causal relationship involved in the creation of photographs, upon which scientific images, rely to carry data (cf. cromEy 2012). First, this essay is going to present a brief overview of the AI image generative techniques and their status within the rest of computational methodologies employed in scientific imaging. Then it will outline their implementation in two specific examples: The Black Hole image (cf.EVENT horIZoN TELEscoPE coLLABorATIoN 2019a-f) and medical imagery (cf., e.g., orEN et al. 2020). Finally, conclusions will be drawn regarding the epistemic validity of AI images. Considering the exponential growth of available experimental data, scientists are expected to resort to AI methods to process it quickly. An overreliance on AI lacking proper ethics will not only result in academic fraud (cf. GU et al. 2022; wANG et al. 2022) but will also expose an uninitiated public to images where a lack of sufficient explanation can shape distorted opinions about science.
Citation
Copy Citation
BibTex
Except where otherwise noted, this item's license is described as:
Under Copyright
Under Copyright
Keywords:
AI technologydigital imagescomputational methodologiesscientific imagingvalidity