talking to and with computers

Chatterbot programs fascinate me. Which is in a way unfortunate because as far as can tell its a form of software that isn’t exactly thriving with masses of people innovating and dedicating their time to developing. Which is understandable. They have absolutely no practical use. Nobody is going to pass the Turing test any time soon. Although there are ones that are more technically advanced I’m not sure I’ve ever even had a conversation with a chatterbot that was any more convincing then talking to ELIZA. I don’t expect chatterbots could possibly be able to effectively replace humans in any significant way in really any field any time soon and even if they could hypothetically be expected to reach a point of competence in the foreseeable future for that to be possible they would still be often hindered by the fact that speech recognition tends to be nowhere near what may be required. There is little reason for anyone to invest in developing them in any significant way. They are at best toys.

One problem worth noting with chatterbots is that they tend to not know anything and they don’t want anything. They are often blank slates. Which means that everything they say has to be reactionary. They cant lead a conversation and they cant communicate anything meaningful because they don’t know anything meaningful.

Now I want to talk about interactive fiction games for a second and why nobody plays them. See the problem with text adventures as I see them is that they are just simply anachronistic. Their fall to obscurity as a niche/dead genre was the logical outcome because they don’t make sense in a modern context. Even if you get past the other problems of text adventures like the confusion and frustration of the often limited vocabulary allowed as input they still just don’t make sense anymore. The idea of typing instructions for your character to travel north, to travel north again, north again, then move west then pickup the key is absurd when computers have evolved to become capable of visually rendering characters and environments and allowing you direct control over them and the ability to move in real time and interact with objects in a context sensitive way.

Using text as an input mechanism for physical control over a character is awkward and inefficient. But there is something that text is still ideal for communicating and emulating and that’s verbal communication. There is no better way to speak with someone then in text. There will never be enough buttons on any controller to allow you to give every necessary word a dedicated button. So just have people type what they want their character to say. You could go with voice recognition but as I said before I don’t think that’s really at a point yet where its usable when not being restricted to only a small handful of accepted words and your aren’t always going to be playing a game in a environment where your comfortable and capable of making noise by speaking aloud. You could also alternatively limit the players input to a few pre scripted responses and display them on a menu or map them each to a button. That’s what most modern games do. But by comparison to the freedom of just typing what you want to say having pre written options will always feel more linear. Even if in both cases the actual available options and narrative are identical it will seem more freeing if the player can phrase what they want in their own words. Just like how giving the player direct control over their character and how they navigate rooms will make them feel like they are less restricted even if they are in effect still just moving north through three rooms then to the area to the west to pick up a key.

So what I’m saying is that I want chatterbot to become a game genre. Get rid of the physical actions you normally make in a traditional text adventure and just have a game focused around a conversation and use that conversation to convey a plot and a narrative.

That’s really my vision of the ideal game. Fuck time control and unconventional gravity and stylistic graphics and all the shit that indie developers usually do. I just want a simple story communicated with a dynamic conversation. That would be innovative.

But I’m not making that game.

I am doing something kinda interesting that involves a chatterbot though. I’ll talk about that tomorrow.

2 thoughts on “talking to and with computers”

  1. ROFL I love your stab at indies for always making games with stylistic graphics and world changing game mechanics… indies ! They suck! ……. wait..

    Seriously though I actually LOVE the idea of making a game where you could some how convey a story through just talking to a bot. That would actually be freakin awesome if done right! A ton of ideas run through my head for that idea.

    1. Yeah. I hate indie developers. Especially the Canadian ones.

      by the way I’m thinking of remaking/completing Time Frack 2D and integrating some of the stuff I did in Gravity Glide but I cant decide if the graphics should have a cell shaded look or if I should have some really heavy motion blur and screen distortion.

Comments are closed.