The term was in use by 2008, credited to Marisa Olson, and it was always a description of a condition rather than a medium. Net art of the 1990s made work that lived in the browser; post-internet artists assumed the browser had already changed how every object and image was seen, and made work for galleries, print and shops that behaved accordingly. Artie Vierkant's Image Objects from 2011 are the cleanest statement of the method: printed sculptures whose documentation was altered in software, so that the photograph and the object never agreed and neither was the original.
The visual language came from the platforms and from the stock and retail imagery that fed them. DIS magazine ran editorials shot in the register of catalogue photography and product listings. Jon Rafman collected images from Google Street View and presented them as found photography. Hito Steyerl's essay In Defense of the Poor Image, from 2009, gave the compressed, copied and degraded JPEG a politics: the low-resolution image was the one that actually circulated. Lo-fi 3D renders with default shading, acrylic display stands, shrink-wrap, watermarks and browser chrome all became legitimate material because they were the material of everyday looking.
By the mid-2010s the tendency had a name, a market and a backlash, and the artists mostly stopped using the word. The language stayed. It is the native register of a generation of designers who photograph products to be reposted, print screenshots on vinyl, and treat a lo-fi render as a finished surface. Its defining quality is a refusal to say which version of a thing is the real one: the object, the listing, the screenshot or the print. Done well it is intelligent about circulation; done lazily it collapses into Y2K nostalgia, cyberpunk or a mockup of an app.