That’s a fair question. Compass Vermont does use AI-assisted research tools, primarily for tasks such as document analysis, summarization, and identifying relevant information within large sets of public records.
Many news organizations are now experimenting with similar tools to help reporters review large document collections more efficiently. At Compass, editorial responsibility, sourcing, and final judgment remain human and rest with me as publisher.
The goal is not automation, but enabling a very small newsroom to analyze complex policy documents and datasets that would otherwise require a much larger reporting staff. As with any newsroom, the reporting should ultimately be judged by the accuracy of the facts, the strength of the sources, and the transparency of the analysis.
There are genuine concerns, voiced by many, about how reliable AI is as a research tool. I would be curious to see you respond to what CWA is actually saying about it, but it kind of seems like your mind is made up here.
On the plus side, this is the best example of what “begging the question” actually means in rhetoric (as opposed to how it’s commonly used).
That’s a fair observation. The focus of the article was primarily on the leadership instability at VTDigger and the operational risks that can create for a newsroom, rather than a full analysis of the union’s position on AI.
More broadly, the reliability question around AI tools is often less about the technology itself and more about how it is used. Like any research tool, the quality of the output depends on the rigor of the reporting, sourcing, and editorial verification behind it.
If the CWA’s specific concerns or proposals regarding AI become clearer in public reporting, they would certainly be worth examining on their own terms.
Would you be willing to write about the extent to which Compass Vermont uses AI?
I was thinking the same thing! Transparency on AI use for thee but not for me?
That’s a fair question. Compass Vermont does use AI-assisted research tools, primarily for tasks such as document analysis, summarization, and identifying relevant information within large sets of public records.
Many news organizations are now experimenting with similar tools to help reporters review large document collections more efficiently. At Compass, editorial responsibility, sourcing, and final judgment remain human and rest with me as publisher.
The goal is not automation, but enabling a very small newsroom to analyze complex policy documents and datasets that would otherwise require a much larger reporting staff. As with any newsroom, the reporting should ultimately be judged by the accuracy of the facts, the strength of the sources, and the transparency of the analysis.
There are genuine concerns, voiced by many, about how reliable AI is as a research tool. I would be curious to see you respond to what CWA is actually saying about it, but it kind of seems like your mind is made up here.
On the plus side, this is the best example of what “begging the question” actually means in rhetoric (as opposed to how it’s commonly used).
Hi Paul,
That’s a fair observation. The focus of the article was primarily on the leadership instability at VTDigger and the operational risks that can create for a newsroom, rather than a full analysis of the union’s position on AI.
More broadly, the reliability question around AI tools is often less about the technology itself and more about how it is used. Like any research tool, the quality of the output depends on the rigor of the reporting, sourcing, and editorial verification behind it.
If the CWA’s specific concerns or proposals regarding AI become clearer in public reporting, they would certainly be worth examining on their own terms.
Tom Davis, Publisher
AI is everywhere, and it's going to get worse, and in more ways than one......