By Rebecca Dasen, Midston Daily Press, reporting for the Associated Civic News Bureau
Researchers at an international digital communications consortium announced Thursday that the global information ecosystem now appears to operate almost entirely on screenshots of other screenshots, a development they described as "unexpected but no longer surprising."
The findings come after a two-year effort to map the origins of news, rumors, memes, political claims, and public outrage as they moved across social media platforms.
"We attempted to trace information back to original sources," said Dr. Markus Eberle, one of the lead researchers. "In most cases, there weren't any."
According to the report, many widely shared claims originated from screenshots of posts referencing screenshots of articles citing screenshots of conversations that no longer existed.
Researchers said they repeatedly encountered image files that had been compressed, cropped, reposted, translated, reposted again, and then cited as primary evidence by users who had never seen the original material.
"The average information object now exists approximately seven generations away from whatever actually happened," Eberle said.
The study itself nearly failed after researchers discovered the project's data infrastructure consisted of three Raspberry Pi computers connected by what one auditor described as "optimistic networking." Raspberry Pi boards have long been popular for low-cost computing projects, with entry-level models historically selling for around $50, though prices vary by configuration.
In an effort to reduce operating costs, the team attempted to migrate the system to lower-cost single-board computers manufactured by Banana Pi, an open-source hardware platform developed as an alternative to Raspberry Pi.
That transition was abandoned after technicians received what appeared to be reconditioned hardware accompanied by documentation that did not match the product packaging.
"The installation guide referred to a board called the 'BPI-M7 Professional AI Development Platform,'" said systems administrator Elena Kovacs. "The box said something else entirely. The photographs in the manual showed a third device. At one point we thought we were reading instructions for a rice cooker."
Project logs show the migration effort stalled for nearly six weeks while engineers attempted to determine whether the hardware was incorrectly labeled or whether they had downloaded the wrong manual.
Researchers said the incident ultimately reinforced the study's central conclusion.
"The documentation was screenshots of forum posts linking to screenshots of older forum posts," Kovacs said. "By the end we couldn't tell whether we were troubleshooting a computer or researching our own report."
Despite the setbacks, the team completed its analysis and concluded that original documents now account for less than 2 percent of information circulating online. The remainder consists largely of screenshots, reposts, summaries, reactions, corrections to reactions, and screenshots of those corrections.
The report recommends that future historians be trained in image enhancement software, meme archaeology, and advanced cropping analysis.
At press time, researchers were attempting to verify a widely shared screenshot claiming the report itself was based on a screenshot of another report. Initial findings suggested that may be correct.
