{"id":115119,"date":"2024-04-17T08:30:08","date_gmt":"2024-04-17T06:30:08","guid":{"rendered":"https:\/\/prohoster.info\/blog\/novosti-interneta\/proekt-gentoo-zapretil-prinyatie-izmenenij-podgotovlennyh-pri-pomoshhi-ai-instrumentov"},"modified":"2024-04-17T08:30:08","modified_gmt":"2024-04-17T06:30:08","slug":"proekt-gentoo-zapretil-prinyatie-izmenenij-podgotovlennyh-pri-pomoshhi-ai-instrumentov","status":"publish","type":"post","link":"https:\/\/prohoster.info\/en\/blog\/news\/proekt-gentoo-zapretil-prinyatie-izmenenij-podgotovlennyh-pri-pomoshhi-ai-instrumentov","title":{"rendered":"The Gentoo project has banned the acceptance of changes made with the help of AI tools.","gt_translate_keys":[{"key":"rendered","format":"text"}]},"content":{"rendered":"
The governing board of the Gentoo Linux distribution has approved rules that prohibit the acceptance of any content created using AI tools that process natural language requests, such as ChatGPT, Bard, and GitHub Copilot. Such tools must not be used in writing Gentoo component code, creating ebuilds, preparing documentation, or submitting bug reports. <\/p>\n
The main concerns that led to the prohibition of AI tools in Gentoo are: <\/p>\n
\n
Uncertainty regarding potential copyright infringement in content created with models trained on large datasets that include copyrighted works. It is also mentioned that there is no way to guarantee compliance with licensing requirements in code generated through AI tools. AI-generated code may be considered a derivative work of the code used to train the model, which is distributed under specific licenses.\n
For instance, if the model is trained on code with a license that requires attribution, the requirements in the code provided by AI tools are not fulfilled, which could be seen as a violation of most open licenses, such as GPL, MIT, and Apache. There may also be issues with license compatibility when inserting AI-generated code, trained on copyleft licensed code, into projects under permissive licenses. <\/p>\n
Potential quality issues. Concerns arise that code or text generated by AI tools may appear correct but contain implicit problems and discrepancies with facts. Using such content without verification may lead to a decline in project quality. For example, synthesized code may replicate the errors of the code used to train the model, ultimately resulting in vulnerabilities and a lack of necessary checks when processing external data.\n
Verification requires significant labor for fact-checking and code review. By analyzing automatically generated error reports, developers waste a lot of time sorting through useless reports and must double-check the information mentioned multiple times, as the external quality of presentation builds trust in the information and creates the impression that the reviewer misunderstood something. <\/p>\n
Ethical questions arise regarding copyright violations in model training, the negative environmental impact due to high energy consumption in model creation, job losses resulting from the replacement of personnel by AI services, reduced service quality following the substitution of support teams with bots, and increased opportunities for spam and fraud. <\/ul>\n