Text: Articles published on this blog are written by a human, not dreamt up by an LLMs/chatbot or the output of an ‘agent’ digesting and summarising.
The use of AI tools in the wider ‘information ecosystem’ is out of my control.
Reporting on projects, companies and individuals means working from information within and, often, quoting from might could be AI-generated press releases, changelogs or website copy. In an ideal world, it would be labelled as such…
Search results and social media now default to providing AI-generated summaries to queries and search terms and – disturbingly – even use AI to rewrite headlines, post excerpts and on-page SERP/answers.
I practice good “hygiene” as a reporter: vetting sources, ensuring quotes are in quote marks, attributing paraphrasing and, of course, exercising my own editorial judgement.
Heck, I’ve taken to snarking in-article if I suspect the project info, press release or statements I’m having to work is (often quite obviously) AI generated/machine written.
Industry-wide usage of machine learning and AI tools means I may be reading or working from AI-generated audio transcripts of interviews or subtitles, or content provided in English via machine learning translation technologies, e.g., Google Translate.
I have made infrequent use of Pangram, an AI-powered AI detection tool, on text I’m suspicious of. Pangram is not foolproof, however.
Images: All custom graphics on this site (article thumbnails, hero images, banners, etc) are created by hand without the use of generative AI tools in any part of the process.
I use third-party assets and elements, e.g., app logos, fonts, textures and stock images, in my design work. There is a risk of third-party elements being the result of AI generation and not labelled as such. I actively avoid AI assets, but note the residual risk.
Should any purportedly “real” photos be sent to me showing improbable or unique situations related to topics I cover, I would run them through AI-powered AI image detection tools to check for signs of generation AI usage or watermarks. If flagged, they won’t be used.
Topics: As a reporter, I sometimes need to cover relevant AI developments, tools and integration efforts as they pertain to Ubuntu and the wider open-source ecosystem. Covering AI/ML related news does not count as an endorsement.
My role is to share information and report on updates (e.g., Firefox adding AI features) and platform changes (e.g., Ubuntu adding AI features) as well as wider community and industry policy and debate around AI (e.g., Flathub rejecting AI-assisted submissions).
This AI disclosure/policy will be amended as/if needed with a note explaining what’s changed.