The Wikimedia Foundation has paused an experiment which showed summaries generated by AI in the upper part of the articles after an avalanche of criticism of their own editors.
Why is it important. Wikipedia remains one of the last great bastions of human content on the Internet, in front of the survey wave that has degraded other platforms. His model, which is committed to democratic governance, has just stopped an important technological advance.
What has happened. He “Simple Summaries” experiment He was born with the intention of making complex articles more accessible through automatic summaries marked as “not verified.” These summaries were made by an aya model of COPE.
The editors responded with comments such as “very bad idea”, “my strongest rejection” or simply “Puaj”.
The background. OpenAi continues to advance in Your plan to become the next GoogleGoogle herself He has embraced the generative AI even in his search engine. In this environment, Wikipedia has maintained the quality of its articles for its human commitment.
In fact, its editors actively filter the content generated by AI, and that makes the platform a reliable information refuge. You know knowing that there will be no Slop.
Marked in red, an example of Wikipedia’s summaries. Image: 404 average.
Between the lines. These protests speak of something deeper than the simple acceptance of synthetic content:
- Wikipedia must evolve to attract new generations …
- … but its editors fear that AI destroys decades of collaborative work.
“No other community has dominated collaboration to such a wonderful point, and this would throw it down,” said an editor quoted by 404 average.
Yes, but. The Foundation has not ruled out the AI completely, at least for the moment. He has promised that any future function will require “participation of editors” and “human moderation workflows.” It sounds like tactical pause.
In addition, the experiment was born precisely from discussions in Wikimania in 2024, when some editors did see this format potential.
In summary. The question now is if Wikipedia will be able to maintain its enormous historical relevance, already eroded since Chatgpt reached our lifewithout sacrificing part of the human criterion that distinguishes it.
The answer to this question, which will not arrive tomorrow, will be what determines whether Wikipedia remains a reasonably reliable knowledge … or another space in automated internet noise.
Outstanding image | Oberon Copeland @seeyinformed.com in Unspash
GIPHY App Key not set. Please check settings