This article “Whatever happened to te Semantic Web?” about the failure of the Semantic Web is interesting. I may note that my personal failure to realize the semantic desktop – an end user application to use data from the semantic web for personal information management – contributed to the overall fail.
I can confirm the observation of the article that our focus at W3C and in academic research was too much on standardizing something that does not exist yet. It would have been good to start with a simpler standard, establish a live and profitable ecosystem of applications, then standardize more. The group I worked within on the Semantic Desktop focussed too much on standards and not enough on usability, finishing the applications, maintaining them, fixing performance issues.
We had about 4.5 Million end-user devices (KDE desktops and maemo Nokias) enabled with a nepomuk rdf semantic web backend including sparql index database of all metadata of all data for personal information management. We had tagging and annotating in many applications (browser, files, email,…) Based on RDF. Users were able to query “what movies have I downloaded featuring female actors born in 1960?” as IMDB/dbpedia data were integrated with the local database.
The ability to search all public knowledge of humanity in combination with all your personal knowledge and all your organization’s knowledge was – cool! Useful!
Having such an end-user application available as part of the operating system of the user would have raised demand for websites to publish more semantic web data – people benefited from it. People would have demanded more of semantic web data. If we had kept the Semantic Desktop running from 2004 until today (2020), more people would today demand publicly linked open data for their own private and business uses. You would demand that product datasheets or price information is available as RDF because you would need it for your daily work when researching products. You would demand that open jobs are published by companies as linked data because when looking for a job and comparing open positions, you would want to download the offers into your personal semantic web note taking tool and compare them and add notes. The list goes on. Not only Google would consume schema.org but individual users.
But this collapsed when a few key people left and funding for further maintenance ran out. Also I should have focused my time in 2009-2013 on the PIM semantic wiki/desktop and not on group collaboration. And only a handful of people understood what we really did, so this vision died.
Which makes my daily work less convenient, as I really miss the ease of working with data I had with the Semantic Desktop. Knowledge work was more open, more easy, more collaborative.