Hints
So you want to be successful in the challenge? Would you like some insider information about how the entries will be evaluated? The following will be the focus of the evaluation rubric:
Transparency
How well is the 'artificiallity' concealed? How smooth is the interface in terms of creating a natural interface or interaction? For example, is it apparent if an entry is a bot or an AI process?
Ease of use
How easy is it for an end user to use, implement, or otherwise experience the AI? Is it cumbersome for the user? Do things have to be setup, turned on, rezzed, configured in a way that might be confusing or frustrating for an end user? Are they too complicated to use?
Depth and Breadth
How sophisticated is the problem that is being solved? What is the depth and range of the problem set or task? For example, is the task to provide greeter for a lobby or to recreate an adaptable emergency preparedness automated response system?
Implementation Complexity
How sophisticated is the AI implementation from the end user's (rather than a programmer's) perspective. What is the adaptability of the implementation? What is the scope and range of automation as it pertains to customized responses based on previous input? For example, in a conversation with a bot, natural language interaction with a broad range of responses and various topic directions would be considered very sophisticated.
Creativity
How clever or unusual is the implementation? How unique is the entry?