Recently unexpurgated legal documents and depositions New York TimesThe lawsuit against Microsoft and OpenAI revealed something extraordinary: executives knew full well the destructive powers of AI when they rushed to release products incorporating the technology, following the explosive success of OpenAI’s ChatGPT.
Internal Microsoft documents cited by the paper’s lawyers warn of a sad reality now playing out across the economy: “doom loops,” in which AI monetizes the work of human creatives while undermining the economic systems that funded their work in the first place.
“Our AI content strategy has triggered a ‘doom loop’ that will harm the performance of our models and the entire web at the same time,” reads an internal Microsoft document. “It is very unusual for an end product to threaten the economic foundations of its critical suppliers, but this is the situation we have created for our LLM business as it relates to its ‘content supply chain’.
The closer you look, the more you see doom loops everywhere. Journalists, for example, published their work on the assumption that online platforms would attract readers to them. But now tools like ChatGPT and Google’s Gemini simply summarize their articles without directing readers to their sites, triggering a crisis that appears poised to wipe out independent media.
Similar stories are playing out in other industries, from music to film. Generally speaking, many fear that AI will ultimately be powerful enough to eliminate large numbers of jobs, creating a situation where unemployed consumers will not have the funds to support the economy.
The irony, as Microsoft’s paper points out, is that the AI industry is a snake eating its own tail: the very products it offers users rely on novel cultural content, even as they destroy the economic incentives for its creation.
It’s also a dire situation for the open web, which has recently celebrated the grisly turn of being dominated by more AI garbage than human-created content. As companies, including Google, fundamentally change the way users interact with the web, more and more newspapers, media outlets and local news channels are having serious difficulty making ends meet.
Worse still, like the legal documents in the New York TimesAs the copyright lawsuit showed, insiders admitted that the entire operation is based on theft.
“Millions of people around the world will soon view big models’ ‘hoovering up’ all their work as an astonishing theft of unprecedented proportions,” admits an internal Microsoft document.
Many other creative industries are also feeling the pressure as AI-related doom loops continue to undermine livelihoods. According to the latest data from the US Bureau of Labor, it’s a bloodbath: more than 200,000 jobs in “creative” industries have been lost in the last four years alone, roughly the time since OpenAI launched ChatGPT.
AI-generated music is quickly supplanting human-produced tracks on streaming services, forcing platforms to figure out where to draw the line. Some have implemented strict rules, while others continue to allow users to exploit the algorithms with slop.
Hollywood studios are also doubling down on AI while protesting actors struggle to make their voices heard. Many of them fear that their image or voice will soon become worthless, posing an existential threat to their livelihood.
Even the job market is being dragged into its own catastrophic spiral. More and more desperate job seekers are turning to AI to update their resumes and manipulate AI-powered job portals, while recruiters are drowning in a wall of nearly identical applicants.
“We’re faced with this tragic situation where every side has a problem,” said Daniel Chait, CEO of applicant tracking systems company Greenhouse. Wired. “They are using AI to solve their own problem, but in a way that makes the problem worse.”
“So, more use of AI begets more use of AI, to no one’s benefit,” he added. “The more this happens, the worse it gets.”
Likewise, the use of AI in the classroom has exploded, with students using the technology to complete their assignments, thereby transferring their cognitive faculties to the machine. Meanwhile, their teachers use AI to provide homework, gradually erasing the human in the loop. Educators who chafe at the idea of using this technology find themselves at odds with school administrators.
The effects of AI on education have been nothing short of a disaster. A recent study from the Center for Universal Education at the Brookings Institution concluded that the risks of using generative AI in education far outweigh the benefits. The associated dangers have never been more evident, with teachers warning that students are rapidly losing the ability to read.
AI supporters including OpenAI and Microsoft lawyers in ongoing project New York Times lawsuits, argue that AI brings something new to the table. They argue that the output of AI models is sufficiently differentiated from the original content on which it is based, thereby falling within the ambit of “fair use,” a long-standing doctrine of U.S. copyright law.
It remains to be seen whether this argument will hold up in court. Additionally, we have already seen companies, like Anthropic, settle for billions of dollars after being found guilty of violating copyright laws.
But considering that technology is simultaneously undermining the economic systems that produced the data they need to exist, this strategy may not be a winning one in the long run.
By pulling the rug out from under many parts of the economy – and reaping all the financial gains – the AI industry is locking itself into a doom loop of its own creation, a cynical strategy that will benefit no one but those at the top who brought the ladder down behind them.
Learn more about doom loops: Microsoft chief admitted in internal document that AI creates a ‘doom loop’ that destroys news media by stealing their content without providing anything in return, rendering AI useless
Gn bussni
