5,645 karma · joined September 25, 2018
https://martinanderson.ai
Email: martin@martinanderson.ai
The VirusTotal behavior analysis linked to says 'No security vendors flagged this file as malicious'
They could be turned off at the time but only on a case-by-case basis. In the end, I got rid of Disqus.
I presume that this falls under the same consideration as direct links to science papers in articles that are covering those releases. Far as I can tell, the central tactic for lowering bounce rate and increasing 'engagement' is to link out sparsely, and, ideally, not at all.
I write articles on new research papers, and always provide a direct link to the PDF,; but nearly all major sites fail to do this, even when the paper turns out to be at Arxiv, or otherwise directly available (instead of having been an exclusive preview offered to the publication by the researchers, as often happens at more prominent publications such as Ars and The Register).
In regard to the few publishers that do provide legal PDFs in articles, the solution I see most often is that the publication hosts the PDF itself, keeping the reader in their ecosystem. However, since external PDFs can get revised and taken down, this could also be a countermeasure against that.
I’m a technical writer focused on internal documentation and outreach content for AI startups. I’ve produced in-depth material for tools and workflows involving ML, often incorporating video and multimedia when appropriate. I also curate and train datasets in a variety of model categories.
Past clients include Metaphysic.ai, Unite.ai, and Bucharest-based startups. Comfortable working with developers and adapting quickly to evolving systems.
Portfolio: https://martinanderson.ai LinkedIn: https://www.linkedin.com/in/martin-anderson-ai-writer/ Email: martin@martinanderson.ai
https://www.theguardian.com/books/2015/aug/01/writers-recall...
(In case not obvious, current title is 'I use my terminal')