There's no way to keep up with all the "Canonical-isms" that show up weekly thanks to marketing trying to give things a marketing name. Actually that is quite feasible and uses the same learning methods as the current wave of AI tools. If a Canonicalism is introduced it is picked up. Every online document needs to be thrown into a giant mill and words (strings, tokens) pop out like a coffee bean grinder. I worked on that more than 20 years ago but today we have the modern Chat-GPT. If I have time (not right now) I might knock together a proof of concept experiment. Just needs scraping of Canonicalisms buried deep in URL's.
Last edited by dragonfly41; December 4th, 2023 at 08:13 PM.
I still see AI "hallucinating" far too much to be trusted.
Well I've kicked off my engine. Will explore findings.
@The Cog @TheFu That's really wonderful. Thanks for bearing with me while I learned.
View Tag Cloud
Ubuntu Forums Code of Conduct