I was wondering if you guys could recommend anyone’s research in particular in regards to human studies and behavior. There are a lot of references out there that have to do on the subject and I don’t really have time to go though them all. My end objective is to design a system that can make choices based on its surroundings as well as past experiences. Ultimately giving the AI the option to go against what the AI core objective, under a believable set of circumstances. At present I have a design that uses a series of filters and a feedback loop, however these filter nodes are empty right now because I’m still researching the topic. The basic loop looks like this Input : Emotion Engine : Logic Engine : Emotion Engine : Output : <Feedback> Logic Engine evaluates final output The emotion engine is used to alter the logics perspective of the input. Output being altered by the emotion engine is an accurate example of how ones emotions might get in the way of ones judgment and actions. Having the AI evaluate the final product will allow the AI to make minor changes to its Emotion Engine and Logic Engine to get more of a desired reaction, as long as these values can never be precisely calculated the AI will never be able to fully cancel out (or accurately predict) the effects of the emotion layer. Well that’s my basic idea, I don’t need a lot of them and CPU being as cheap as it is I can get away with costs of this loop, I think. Added: At the moment I'm reading this. http://en.wikipedia.org/wiki/Emotion [Edited by - MrRage on March 9, 2007 9:53:05 AM]

Here's a big old book that may give you an insight to people...

Chapter one may be of most use to you.
http://www.mises.org/humanaction.asp

It's online in pdf format.

Oh, and BTW, don't expect any code from that site. It's all economics (trust me on this one, it's cool)

No problem about the code, I wasn’t really expecting any research to be intended for an AI. Well I’m going to go though chapter one now, thanks a lot.

Quote:Original post by MrRage
anyone’s research in particular in regards to human studies and behavior.

very vague description, any psychology literature fits that. maybe "cognitive psychology" may be one good next step.

Well I completely skipped psychology in college and even though I dated a psychology major she wasn’t very interested in her major and never wanted to talk about it. So I’m just trying to figure out where to start is probably the best way to put it.

I can recommend, as a starting point, Emma Norling's research. She did her PhD on incorporating human behaviours into simulation environments. Her work was done in conjunction with Australia's defence research agency (DSTO), but since she couldn't write her thesis on classified materials (the training simulators she was working on), I believe she developed a lot of her material in a FPS environment (Quake I think it was). Her papers will provide you with some starting references for the field. I'm sure that if you needed more information, she'd be only to happy to point you in the right direction.

You might also want to look into the rest of the research done in the Agent Lab at Melbourne Uni.

Here's the link to Emma's publications: http://www.cs.mu.oz.au/agentlab/publications/Author/NORLING-E.html

...and here is the Agent Lab link: http://www.cs.mu.oz.au/agentlab/