Can AI creation be considered art?
From drawing to AI-generated podcast dialogues to screenwriting, people are working together to replace human creativity with the automation of computers while discarding the concept of art as we know it. Jacobin author Luke Savage uses the 2013 film Tim's Vermeer as an example to discuss the range of issues behind AI-generated art and the ideas behind it.
In the film Thiem's Vermeer, actor Penn Gillette chronicles how his friend Tim Jenison recreates the techniques of 17th-century Dutch painter Johannes Vermeer. To that end, Jennison, a software executive and visual engineer, developed a complex set of methods that used mirrors and light to replicate Vermeer's technique, recreating its signature depth of field and chromatic aberration.
A still from "Tim's Vermeer"
Janison made an impressive effort to replicate Vermeer's 1660s work The Music Lesson. However, Jennison and Gillette seem to have misunderstood what they did. Gillette talks endlessly about the "photographic" and "cinematic" qualities of Vermeer's work, but fails to capture its interesting and abstract nature, as he excitedly says: "My friend Tim painted a picture of Vermeer!" He painted a picture of Vermeer! "But this reproduction is nothing more than an extremely elaborate experiment in digital painting, a derivative mimery of beauty."
The two actors saw Vermeer's work as a technique, a method, and wanted to represent reality as realistically as possible. In this understanding, Vermeer's work does not involve social or cultural processes, has no inspiration other than mechanical production acts, and has no purpose at all other than the characteristics of photographic realism. And this approach seems to have similarities with the artistic creation of artificial intelligence.
Luke Savage noted that like any technology-driven industrial process, AI may end up having far-reaching social and material impacts. But ultimately, AI has what has driven capitalism since the 19th century, which is the constant pursuit of more efficient production at a lower cost. This development is a threat to artists and cultural workers. As artist Molly Crabapple observed, existing apps like Stable Diffusion and Midjourney can already generate detailed images based on text prompts at little to no cost. "They're faster and cheaper," she writes, "though the images are still problematic and somehow lack soul." AI sometimes draws extra fingers and tumors in ears, but that's good enough. These images can be used for book covers and editorial illustrations, which is a source of income for many illustrators. ”
In the realm of culture, cultural goods will become extremely crude: computer-made fake paintings may be sold on artificially created scarcity markets such as cryptocurrencies or non-fungible tokens (NFTs), and virtual pop stars will be algorithmically asked to record formulaic music. Eventually, the author will be replaced by a generative algorithm. These algorithms continuously reduce the differences in dialogue and plot structure, while reducing the degree of author's involvement. In Luke Savage's view, the promoters of AI culture mistakenly view replication as creation, equating realism with artistic expression. In this concept, creativity is ultimately a mechanical endeavor, and every art: painting, film, music, poetry, is nothing more than a collection of granular data points; It is no exaggeration to say that "art" is the sum of its components.
Accelerated by the monopoly of tech companies, mass entertainment is increasingly becoming a wasteland for derivatives and algorithmically generated "content" — with little new meaningful content. With the help of technology, companies have carved out a zombified model of cultural production in which existing intellectual property (IP) is endlessly recycled and mass-produced in the form of sequels, prequels, remakes, and botched imitations. In terms of the revolution represented by artificial intelligence, it will perfect this process, and this is not a real revolution at all.
Although determining whether a certain work of art is good or bad, it is tortuous and complex. But it's safe to say that making a creative process more "efficient" doesn't make it better.
Art, music, almost the entire human life and mind transcends basic matters such as sleeping and eating, exuding an essence or spirit that cannot be reduced to a mechanical process, whatever we decide to call it (wisdom, humanism, creativity, soul). By definition, it produces something that cannot be quantified or classified. Once a painting or piece of music is created, it can be broken down into its constituent elements, which in turn can be rearranged or reconfigured to produce something else. However, unless some new creative element is introduced, the result of "innovation" will always be a fake copy.
Quote Article:
https://jacobin.com/2023/05/ai-artificial-intelligence-art-creativity-reproduction-capitalism
Surveillance controversy in Western workplaces
In September 2020, a reporter from Vice magazine discovered that Amazon was hiring two "intelligence analysts" for its global security operations (GSO). Analysts will use data analytics and other tools to detect and counter "labor organizing threats" and other political opposition against Amazon. This ubiquitous employee monitoring sparked an upsurge and opposition from employees. In 2022, Amazon warehouse employees in Staten Island formed a union, and employees publicly expressed their dissatisfaction with the constant monitoring of work.
For more than a decade, scholars, journalists, and technology leaders have been watching how digital technologies will transform jobs. Brishen Rogers, an associate professor at Temple University's Beasley School of Law, wrote in the Boston Review of Books about this phenomenon. Researchers believe that digital technology has two different applications. One is by automating tasks, thereby replacing specific workers; The other discriminates against workers on the basis of race, gender, nationality or disability. But in today's vast service economy, some companies are using digital technologies as tools of domination, using them to limit wage growth for employees, prevent workers from unionizing, and reinforce the exploitation of labor. Workers' resistance to digital surveillance represents their call for transparency and democratization of digital technology in the workplace.
On May 31, 2023, local time, the United States, Amazon and its subsidiaries will pay more than $30 million for user privacy violations.
Conflicts between corporate surveillance and workers are not new. For more than a century, companies have tried to generate, acquire, and quantify information about workers and processes to drive down wages. After a long struggle, employers took control of production from workers, processed and informatized production skills, and enacted so-called "laws" that tied output rates to wages.
With the advent of the telegraph, telephone, fax machine, and modern information technology, companies realized remote supervision of the workforce. In recent decades, as intranets, mobile computing, location tracking, image and natural language recognition, and other forms of advanced data analytics have matured, companies' monitoring capabilities have expanded dramatically. Today's company aspires to continuous monitoring of all aspects of work and production, and this monitoring is also asymmetrical: companies can monitor employees without their knowledge, while preventing employees from monitoring management.
Today, the largest employers in many countries are the retail, food service, logistics, hospitality, and healthcare industries. These companies employ large numbers of workers, but productivity growth is slow because producing their products requires human labor or attention and is difficult to improve through technology. As a result, these companies are very concerned about how to curb wage growth. Many businesses adopt high-employment, low-skilled and high-turnover business models and use new technologies to prevent workers from forming collective power.
According to Brishen Rogers, companies use data to constrain employees in three different ways; He called the first "digital Taylorism", which established management control over the labor process with a scientific management system. Digital Taylorism includes various forms of automation and enhanced surveillance. In Amazon's warehouses, for example, algorithmic monitoring systems report employees who aren't performing fast enough, or who go to the bathroom without permission from their managers, and sometimes even recommend firing them.
In addition to digital Taylorism, companies use surveillance and data aggregation techniques to prevent union organizing and other collective action. For example, companies could use new hiring algorithms to aggregate job applicants' work history with social media posts or political behavior data to screen out employees who might challenge management's authority. Workers' organizing is a process that requires trust and solidarity, and when workers are engaged in a common identity, they can take collective action to protect each other. But modern surveillance can stop this mobilization. First, workers who are often supervised and separated from each other have little opportunity to achieve a common goal. In addition, the development of speech processing and natural language processing software allows companies to "hear" almost everything being said in the workplace and see when workers are talking to each other.
Finally, many companies are using new technologies to change their business scope. They choose to buy labor, treating workers not as legally hired employees, but as independent contractors. Amazon, for example, outsources its delivery operations to various outside companies, but as one reporter found, Amazon's contract requires service providers to "provide Amazon with physical access to its premises, as well as all sorts of data that retailers want, such as geographic location, the speed at which drivers move their vehicles, etc." Amazon says it has the right to use the information as it pleases. "This oversight can give Amazon its traditional hiring authority without liability and cost."
Given the technological capabilities of today's employers, policymakers may need to consider banning long-standing and seemingly uncontroversial forms of workplace surveillance, such as monitoring of workers on the shop floor. In fact, advocates have begun discussing ways to abolish workplace data to some extent. Researchers at the University of California Berkeley Labor Center, for example, after extensive consultation with academics, unions and others, have recommended banning the use of facial recognition and algorithms in the workplace to identify workers' emotions and restricting employers' collection of worker data "necessary and essential to a worker's work." The researchers also recommend that employers should only be able to use electronic monitoring "as necessary to fulfill core business tasks, protect worker safety, or meet legal obligations."
Brishen Rogers said such reforms would require more fundamental changes in labor laws: weakening employers' power to select and implement workplace surveillance techniques and giving workers a real say in production planning and execution. Three types of reforms to data practices can advance these goals: prohibiting data collection and use in various contexts, deliberating on data practices in others, and placing data sources or technologies under public or social control.
In addition to the aforementioned ban on data collection, Brishen Rogers suggested that Congress could consider giving workers some collective rights to consult on technological change, regardless of whether they are unionized or not. In addition, Congress should pass reforms that give workers and the public greater control over data and related technologies: socializing data as a public resource. For example, Congress could require companies to share the data they collect about workers and workflows. Regulators or workers' rights groups can then analyze this data to spot violations of basic labor laws, such as non-compliance with wages and hours. Under such reforms, workers will have more bargaining power. However, given employers' wariness of workers in their grip on power, such reforms will not be easy. Many technology companies and service giants will circumvent restrictions in their own ways and continue to closely supervise employees.
Quote Article:
https://www.bostonreview.net/articles/workplace-data-is-a-tool-of-class-warfare/