In recent months, a number of novelists, artists, and newspaper publishers have sued generative artificial intelligence (AI) companies for free riding on their content. These lawsuits allege that companies using that content to train machine learning models may be violating copyright laws.
From a tech industry perspective, this content mining is necessary for tech companies to build the AI ​​tools they claim will benefit us all. In a recent statement to legislative bodies, OpenAI asserted that “it is impossible to train today's leading AI models without using copyrighted material.” It remains to be seen whether the courts will agree, but it's a bad situation for content creators. In February, a California court dismissed most of Sarah Silverman's lawsuit.
Some of these cases could shed light on ongoing negotiations as some companies look for ways to pressure other companies to share a piece of the AI ​​pie. For example, publisher Axel Springer and social media platform Reddit recently struck a deal to make money by licensing their content to AI companies. Meanwhile, in the UK, legislative attempts to protect content produced by the creative industries have been abandoned.
But there is a larger social dilemma at play here that is not easy to detect. our Content — content that is not normally associated with copyright law, such as emails, photos, and videos uploaded to various platforms? There have been no high-profile court cases regarding it. Nevertheless, the appropriation of this content by generative AI reveals itself as a commemoration of social and cultural transformation.
After all, this type of content is considered a type of common property that no one owns, so this change is easy to miss. But this appropriation of the commons involves a kind of injustice and exploitation that is not captured in copyright litigation and that we still struggle to name. This is the kind of injustice we've seen throughout history whenever someone claims ownership of a resource. just there To take.
In the early stages of colonialism, colonizers such as Britain considered Australia, their recently “discovered” continent, legally a “no man's land”, even though it had been inhabited for thousands of years. In other words, he claimed that it was no one's land. This was known as the principle of discovery, the colonial version of “discoverer, controller.”
These claims have been repeated recently by companies who want to treat our digital content and even our biometric data as mere emissions that exist only to be exploited. The principle of discovery survives today in the seamless transition from cheap land to cheap labor to cheap data. This is a phenomenon we call “data colonialism.” The word “colonialism” is not used here figuratively, but rather as a very real term based on the continued expropriation of human life through data rather than the extraction of natural resources or labor. It is used to describe an emerging social order. Data colonialism helps us understand today's changes in social life as an extension of a long historical dispossession. All of human culture is the raw material that feeds into commercial AI machines that promise huge profits.Earlier this year, OpenAI launched a $7 funding round Trillion“It's more than the gross domestic products of Britain and France combined.” financial times put it.
What really matters is not whether the output of the generative AI plagiarizes content from famous authors owned by powerful media groups. The real problem is an entirely new revenue model that treats our lives as free input in the form of data. Generative AI is just one egregious example, but this lucrative data acquisition is actually part of a larger power struggle with a long history.
To challenge this, we need to move beyond the narrow lens of copyright law and regain a broader perspective on why extractivism masquerading as discovery is wrong. Today's new, and so far largely uncontroversial, transformation of our lives and cultures into a colonized data realm has been linked to decades, if not centuries, of Big Tech and the rest. will define our relationship. Once a resource is appropriated, it is almost impossible to get it back, as evidenced by the fact that discovery doctrine is still cited in modern government decisions denying indigenous peoples rights to their land.
As with land, so with data. If we do nothing, we will be counting the costs of Big Tech heuristics for a long time to come.
Applying history lessons in the age of AI
Unfortunately, a unidirectional approach such as quitting certain social media platforms is not enough to address these issues. Colonialism is a multifaceted problem with a centuries-old history, so combating its new manifestations will also require multifaceted solutions that borrow from the rich anti-colonial tradition .
The most important tool in this fight is our imagination. Decolonizing data needs to become a creative and cultural movement. Indeed, no colonized society has ever succeeded in definitively and permanently undoing colonialism. However, even if they could not resist the colonial powers with their bodies, they could resist them with their hearts. Collective ingenuity is our most valuable asset.
in our recent book Data exfiltration: Big Tech’s new colonialism and how to fight backHere we outline some practical ways to begin applying this type of creative energy. We borrow models from Latin America and Latin American activists, who encourage us to act simultaneously at three different levels: within the system, against the system, and beyond the system. It is not enough to limit yourself to one of these levels.
What does this look like in practice? Works At the inner This system puts pressure on governments to do what few have been able to do so far: regulate Big Tech by passing antitrust laws, consumer protection laws, and laws that protect cultural works and heritage. It may mean continuing to apply. It may be tempting to want to abandon mainstream politics, but doing so is counterproductive in the long run.
But you can't wait for the system to repair itself.this means we need to work against This system embraces the politics and aesthetics of resistance, just as decolonial movements have done for centuries. Inspiring examples abound, including movements against unionization, labor rights, indigenous data sovereignty, environmental organizing, and the use of data technologies to carry out war, surveillance, apartheid, and immigrant persecution.
Finally, we need to think beyond It is a system that builds ways to limit the misuse of data and redirect its use towards more social and democratic goals. This is probably the most difficult, but most important task. It will require not only new technologies, but also new ways of rejecting technology. Resisting the new injustices of data colonialism will require a massive collective and imaginative effort. This work is an important step in a long journey to confront and reverse colonialism itself.
We're not going to stand for that. you?
Don't bury your head in the sand. You know what we are facing as a country, as a people, and as an international community. Here at Truthout, we're gearing up to meet these threats head-on, but we need your help. We still need to raise $4,000 to continue publishing independent journalism that doesn't shy away from difficult and often dangerous topics.
We're able to do this important work because, unlike most media, our journalism is free from government or corporate influence or censorship. However, this is only possible with your support. If you like what you're reading, or simply value what we do, would you consider taking a few seconds to contribute to our work?