The everyday blog of Richard Bartle.

RSS feeds: v0.91; v1.0 (RDF); v2.0; Atom.

Previous entry. Next entry.

4:37pm on Friday, 11th May, 2018:



Catching up on the MMORPG news I missed while giving myself blisters walking around Poland, I noticed an article on MassivelyOP about Occupy Whitewalls.

OWW is a virtual world for sure, although whether it is (or will become) an MMORPG remains to be seen. I've been involved with it for several years now, as I'm friends with its CEO (Yarden Yaroshevsky) and meet up with him occasionally to tell him why whatever it is he wants to do next with it is crazy.

It's basically a virtual world in which you build your own art gallery which other people visit. As you can deduce from that brief description, there's not a lot of gameplay to it, which is why it doesn't qualify as an MMORPG. That doesn't mean there isn't a lot to do in it, though. The thing is, if you want to build an art gallery then yes, you can do some Sims-like room-building and so on, but you need actual artwork to display. Which artwork, though?

This is where it gets interesting. OWW has an AI system behind it called D.A.I.S.Y (which if I knew what it was an acronym for then I could perhaps argue for its having an extra dot at the end of its name). Anyway, the purpose of D.A.I.S.Y is to help people find works of art that they like. If it thinks you like a work of art (if you bought it, say, or if you zoomed in on it or whatever), it compares it with other works of art it thinks you like and tries to abstract common features. It has no idea what these features mean, just that you may like them. If you look at a bunch of portraits, it'll pick up on that; if you look at works by a particular artist, or from a particular country, or of a particular size, or of cats, it'll pick up on that, too. Then, when you ask for new suggestions for which art you might like, it'll present you with some ideas based on what it thinks you'll like, plus a similar number of pieces chosen entirely at random.

I think the AI it uses is some kind of automated neural network, but there could be an evolutionary algorithm component to it, too. I myself would also like to see some kind of comparison with what other people like (so if you happen to have identical tastes to a truck driver from Milwaukee, and said truck driver likes Da Vinci's Lady with an Ermine and Weird Hand, then you might like it too); I don't think it does that yet, though. Selecting works of art is surprisingly absorbing for some people, who do it to relax. That may sound odd, but then so does relaxing by killing orcs...

Although OWW is 3D, all its artworks are currently 2D (paintings and photos; I think there may be a video there too to show it works). It's not going to be the break-out application for virtual reality, then... As for where the artwork comes from, well it's currently all from public domain collections in national art galleries. There are plans to allow people to upload their own artwork, but it'll probably cost money to do so in order to deter griefers from spamming it with pornography. Unfortunately, a large number of artists use Apple computers, which have high-end displays and low-end graphical capabilities; they have dreadful frame rates for 3D environments, but at least the frames are crystal clear while they're loading. This means that artists might upload their own works, but they won't be spending much time showing people round their virtual galleries unless they invest in a proper computer...

I knew that AI and virtual worlds would come together when I did my PhD back in the 1980s. I wasn't expecting it to work out quite like this, though..!

Latest entries.

Archived entries.

About this blog.

Copyright © 2018 Richard Bartle (richard@mud.co.uk).