Skip to Content

Step 4: Claims

Version 3.2 by Michaël Grauwde on 2023/03/27 03:49
TopicQuestionAnswer

12.png

Measurements

For each positive and negative effect listed in step 3, describe how you could evaluate (measure) whether they actually occur.

1. The evaluation of the conversational agent's effect can come from individuals having deliberations without the conversational agent versus with the conversational agent. This could allow us to see if more ideas were generated or if deliberation went smoothly.

2. The reflection could be analysed by doing interviews or conducting surveys with participants and asking them how they felt before and after they engaged with the conversational agent. In this manner, we can see if the user was able to reflect on their opinions and values than before they engaged with the agent.

The impact of the negative effects

1.  We can analyse how the humans feel about themselves and if they feel put in a box by examining how they engage in discussion and seeing how open they are to new ideas. miscommunication between the conversational agent and the citizen could be analysed by the lagging time in which emergency services were dispatched to the citizen requiring help. Furthermore, this could also be seen by the lack of understanding in the text communicated by the citizen on the end of the conversational agent. 

2. In disaster times, it is required that the authorities communicate well with the public. A lack of communication could lead to distrust in the system as well as other AI systems. This distrust could lead to citizens being unwilling to use the conversational agent in the future. This could lead to lagging response times from the emergency services as human respondents may be overwhelmed by the amount of calls that they may receive. This could lead to an increase in deaths and injuries in crisis times. 

14.png

Benchmark

For each measurement, what are the benchmarks (criteria)? (i.e., what are desired values?) 

 

13.png

Demonstration of AI-functionality

Can you describe how you could demonstrate that your AI function (s) achieve(s) the effects that you listed in the previous question?I could describe achieving previous effects by the response times in dispatching of emergency services and in the lives saves. Furthermore, a positive response to the conversational agent could also be seen as a demonstration of the usefulness of the conversational agent.