CRPGs, SIS, and SOlo Play: Is it Role Playing?

[M. J. Young]

M.J....

Until computers can come up with the algorithms independent of specific human input,
Until computers can create concepts that they do not have an algorithm or program pre-existing for, and
Until computers can retain and manipulate a memory indepent of human control or input,

they will not be capable of active participation in an RPG.  They will not be capable of thought or imagination.

I started another thread specifically on computers as players in CRPG play, related to negotiation of the shared imagined space; but this is here, and I thought I should answer it here.

My problem is that the third of these three conditions are a thing computers can do, as I understand it, and the first two are things for which it has not been demonstrated that people can do them.

Of course, this would be a debate far beyond The Forge. I am not persuaded that people create concepts that are not built on data collected. We've had a lifetime to collect data and to develop algorithms which permit us to combine such data in realistic ways. There are computers that collect data and develop algorithms based on the data. The CRPGs we play in most cases are not these--they are programmed with the algorithms and data that the programmers believed would be most useful in running the games. However, the question isn't where we got the data or the algorithms, but whether we are capable of using them in play. Computers patently are able to use their data and their algorithms; they are just considerably more limited in what they know than we are.

I'll mention the experiments in computer intelligence and gender identification. A computer equipped with camera input and software to analyze images was given hundreds of photos of men and women, in which nationalities, cultures, hair styles, races, and dress were all varied. The computer was informed which images were male and which were female, and set to the task of analyzing those images. It was then given hundreds of photos of other men and women, and asked to identify the genders of those in the new photos. It succeeded in doing so more accurately than humans shown the same photos. No one knows how the computer succeeds in parsing males from females, but it does--it has developed its own algorithms by which to do so.

But I do not see the difference between algorithms I developed and algorithms I was taught; in either case, they are knowledge I use. It doesn't matter whether when I was very young I observed that if I had one of something and added to it another of that thing I then had two, or if when I got to school I was taught that one plus one is two--once I have the knowledge, it is my knowledge. In the same way, there is no distinction between the computer game that develops its own algorithms and the one that has been programmed with algorithms sufficient to support intended play.

I'm contending that the computer is the player. It is not a particularly bright player in some ways, but has some very sophisticated abilities in others. A CRPG would then be a role playing game with a shared imaginary space in which one of the players doing the imagining is a machine.

It may be most profitable to take any answers to this to that other thread;  unfortunately, I don't have the link yet (and if I try to get it, I'm likely to erase all the markers telling me what threads I missed this past week), so you'll have to look for it.

--M. J. Young