- Apple's new Machine Learning Journal makes academic research slightly more accessible than research papers.
- The first post on the site concerns a paper published in December.
In contrast to academic research papers -- the conventional format for publishing research -- the blog posts on the new site appear in Apple's signature San Francisco system font with plenty of white space, and blocks of text are divided up with animated GIFs and whimsical subheads.
The launch of the site is another sign of Apple's interest in being seen as an active player in artificial intelligence research. An institutional "journal" does signify some type of ongoing commitment to an area, even if it doesn't maintain the standards of a more formal academic journal, such as the Journal of Machine Learning Research.
The site doubles as a recruiting tool -- certain machine learning engineers could be inspired to apply for jobs after reading one of the blog posts.
It comes nine months after Apple appointed Carnegie Mellon University professor Ruslan Salakhutdinov as its director of AI research. And Apple made a big splash in AI at its developer conference in June with the introduction of Core ML software for incorporating machine learning models into applications, to do things like recognize objects in images.
Oddly, the inaugural post on the Machine Learning Journal -- which is based on a paper that was published seven months ago -- does not provide credit to specific Apple employees who did the work, although it does provide references to other work.
It's not clear how frequently Apple will post updates on the site. (Apple employees rarely publish research papers, so anything more frequent would be a step up.) And Apple has not provided an RSS feed for it yet.