Reference no: EM133679124

Introduction

While we live in a world that has rapidly deteriorated our trust in what's online and on TV, we as a society now have to contend with "AI" images, created from machine learning, to discern what is and isn't misinformation. Alongside this, social media has blown up with these generators creating art that anyone can prompt and have claimed to create, and businesses lay off graphic designers, on the mindset that a handful of people using these tools can replicate the work of artists. These generators have come under fire heavily, from both artists and the general public, for effectively stealing "more than two billion images culled from websites all over the internet, including Pinterest, Flickr, and Getty Image." (Utkum Ikiz, 2022) Both artists and those who deal with ethics in engineering have called for a more calculated approach to the gathering of base images. While unlikely to disappear overnight, the engineers behind the machine learning generators need to generate an ethical source of images for the tools to pull from, in order to prevent the theft of art, the avoidance of using copyrighted images, and finding a medium between the two sides' livelihoods.

Body Paragraph One

These programs were originally nothing more than tools being created by programmers and engineers to attempt to demonstrate that machine learning can outperform human capacities. While in a private setting, these tools used public datasets to begin keying in how words and images correlate, and the engineers affected the datasets to limit certain topics or keywords. These public datasets are made apparent when combined with the algorithmical bias created by limiting, or prompting niche, topics because despite being told not to recreate images, these programs sometimes get stuck on particular images in the training sets. The limited pool caused by a niche subject, affects the overall quantity of base images to learn from, resulting in a generator that, "with some text prompts, the system will produce an image that is essentially a low-resolution duplicate of an image in the training set." (Goetze, 2024) The struggle present is that these forgeries are largely produced due to specific or niche prompts, the obvious solution would be to feed the generator a greater number of prompts, however, this would continue to likely increase these cases, as more artists have their work used for training sets. Without an ethical source, claims of theft will never cease with these public tools.

Body Paragraph Two

In this regard, while the artists struggle with their works being stolen and reproduced, the individuals and companies who use these tools are coming under fire for using copyrighted images that were pulled from the generator.

One such attempt to begin trimming the training sets, "curated DeepfakeArt Challenge, a high-quality benchmark dataset of over 32,000 art records, encompassing a broad range of techniques for generative forgery," as they set out with the goal of "detecting nuanced instances of content replication." Aboutalebi, et al., 2023) These attempts to create new benchmarks and training sets to cause the algorithms to avoid copyrighted imagery are the first step towards ethical generators.Companies will continue to face increasing legal pressure as tools to detect copyright within generated art continue to grow, reinforcing the current need to remove them from the training sets, and establish an ethical database for these tools to learn from.

Body Paragraph Three

As companies will increasingly grow to utilize these programs, artists will find their livelihoods threatened, due to the increase in wages and time needed to produce proper digital art when compared to the machine-learned systems.

A report from Goldman Sachs notes that "two-thirds of current jobs are exposed to some degree of AI automation", estimating up to 300 million jobs being replaced. (Biggs & Kodnani, 2023) Given the prevalence of "AI Art" already present within technologic industries such as video games and social media, artists are some of the most at-risk occupations towards being replaced. While the AI companies cannot just close their doors and cost their livelihoods, the ethical pool needs to also keep artists in mind, in order to preserve everyone's way of life.

How does their thesis statement make a claim? Is their claim arguable and researchable?

restate what you believe your peer's thesis to be, and identify the key points that support the thesis. Do the key points directly support the thesis or move in another direction? Offer suggestions for how the key points could tie more tightly to the thesis.

What are the objections to your peer's claim that they will need to address in their persuasive essay? If there are no objections, what is a claim that would have valid objections?

Does the evidence for each body paragraph clearly support the key point of the paragraph? How might your peer indicate a stronger connection between the key point of the paragraph and the evidence?