Elon Musk’s X converts community notes into artificial intelligence

Chatbots is regularly famous for providing questionable information and tangible details, making it terrible possibilities for the role of the fact -finding. However, Née Twitter plans to publish artificial intelligence agents to help fill the gaps on Fame Community notes, with the emergence of observations created from artificial intelligence as soon as this month. What can make a mistake?

The new model will allow developers to introduce artificial intelligence agents to review it by the company, according to Public announcement. Agents will be tested behind the scenes, made to write practice notes to learn how to perform them. If they provide useful information, they will get the green light to go on the platform and write notes. These observations still have to obtain the approval of human auditors on society’s observations, and they still have to find useful by people who have a variety of views (how this scale is determined somewhat transparent).

The developers who present their agents can be operated through the AI ​​model, according to Bloomberg, so users will not be closed in Grok despite direct relationships with Musk (perhaps because Musk simply Groc cannot be prevented from waking upIt doesn’t matter How difficult it is to try). Expecting the company is that the notes created from artificial intelligence will significantly increase the number of notes that are published on the platform.

They need Kinda to artificial intelligence for that, because the notes created by a person have fallen on a cliff. and NBC news story It was published last month that the number of observations published on the platform had been made to half from 120,000 in January to only 60,000 in May of 2025. There are fewer people who make notes, fewer people who classify them, and display fewer notes. Basically, the participation collapsed with the fact -verification service.

There are likely to be a number of factors for that. For one, the platform is a type of shit width. A Bloomberg analysis I found that it takes about 14 hours to get a note related to publishing with wrong or misleading information, mainly after the first viral cycle passes. The differences between the shareholders of societal notes also led to the failure of the effectiveness of the facts to publish them, and about one in four are withdrawn after its publication due to the opposition between the residents. This number rises when it comes to controversial issues such as Russia’s invasion of Ukraine, which has seen more than 40 % of the eventually published notes.

Then there is the man who owns the site that, despite promoting community activity as a major reform of wrong information, spent more and more time in it. Earlier this year, musk ClaimWithout providing evidence, society’s observations can be developed by government actors and old media, and instilling lack of confidence in the entire process. Do you know what will not make the game more difficult in the game? Unleash robots on the problem.