Opinion: AI Firms Are Engaging in Widespread Theft, Masking It as "Training"

Ursula K. LeGuin once wrote "There isn’t a correct response to an incorrect query." Even though AI may find this difficult to grasp, as human readers, we comprehend it instantly. In order to address an issue effectively, one must concentrate on what truly matters regarding it.

That's a significant issue within the current discussion, which has now been escalated by Professor Nicholas Creel , over whether artificial intelligence models “learn” the same way as humans.

The straightforward response, clearly outlined in Erik J. Larson’s book, is as follows: The Illusion of Artificial Intelligence, That’s because they fail to do so. These systems struggle with initiating or producing original ideas; they get stuck when required to function under ambiguity or limited details, and they fall short on empathy and genuine comprehension. While they excel at replicating and leveraging vast amounts of data, they cannot derive novel insights or grasp hidden truths from fragmented pieces of info like human beings can.

I posed the query, "Is AI reasoning similar to human reasoning?" to Microsoft's AI tool called Copilot. Here was my response:

Artificial Intelligence depends on extensive datasets and algorithms for making decisions and forecasts. The processing of information hinges upon recognized patterns and statistical examination. Adhering to established guidelines and models, AI reaches various outcomes. Intuition and emotional influence do not play roles in these determinations. Although AI has the capability to learn from data using methods such as machine learning, this educational process stems from mathematical frameworks rather than individual encounters.

Like Moiya McTier from the Human Artistry Campaign has explained , real creativity flows from far more than crunching big data sets to pull out patterns and connections. It “is the product of lived experience and grows organically from the culture, geography, family, and moments that shape us as individuals.”

Therefore, it is evident that AI acquires knowledge and generates outcomes through methods that are inherently distinct from those used by humans.

However, for individuals like myself who encounter AI in our daily lives—musicians whose creations have been extracted from the web and utilized within AI systems without permission—the philosophical debate seems somewhat irrelevant. As Le Guin suggests, the crucial issue at hand is understanding what AI actually accomplishes—and determining if these outcomes justify the associated costs.

To create their models and deploy their pattern-recognition systems, AI firms inevitably require machines to replicate copyrighted materials or generate derivative works based on them. Additionally, they disseminate these creations over an extensive network. Such activities fall under three specific privileges granted exclusively to creators according to federal legislation.

Typically, a business aiming to participate in such activities would straightforwardly obtain licenses from creators. However, AI firms, who could be seen as rivals, have largely chosen not to acquire licenses but still utilize the works, essentially devaluing these copyrights to nothing.

That represents a significant depreciation of the world's creative heritage—a substantial loss in potential opportunities and employment for actual individuals. Furthermore, this trend could lead to a shallow and imitative cultural landscape. If not addressed promptly, it might leave behind an immense void where innovative and original contributions from future creators ought to emerge. While AI systems can generate various renditions based on their analysis and synthesis of numerous source materials (thus avoiding direct accusations of copying), these efforts fall short when it comes to transcending established boundaries and delivering genuinely groundbreaking work.

Unfortunately, proponents of artificial intelligence insist on portraying commercial AI systems as adorable robots that acquire knowledge similarly to how humans do. By adopting certain terminology, they've recast creativity as "data" and widespread replication as "training," making it sound like these AI entities are simply pets. In reality, this narrative serves to mask the truth: a cabal of multitrillion-dollar corporations along with well-funded tech investors are either engaging in or turning a blind eye to extensive copyright violations.

Perhaps that game could deceive an AI, but as for us humans, we easily see through it.

David Lowery serves as both a mathematician and an artist, being a writer, musician, producer, and vocalist for the groups Cracker and Camper Van Beethoven.

To stay updated with the most recent news, weather reports, sporting events, and live streams, visit The Hill.