Week 7 - Automata

2) Write a one paragraph describing the Turing test and another paragraph describing an argument against the Turing Test, known as the about the Chinese room.

Turing test aims to determine the a machine's ability for intelligent behavior. For the test a human judge will engage in a normal conversation with another human and a machine in natural language. The judge does not know which one is machine or which one is human. If the judge cannot reliably distinguish who is the real human and which is the machine, then the machine would succeeded in passing the Turing test. The test only allows the participants to communicate using text based channel.

However, the Turing test is not without its weakness. One such problem is that to pass Turing Test the Machine to pretend to be human, however not all human behavior is intelligent or certain intelligent behaviors will be perceived as un-human like. So for the machine to pass the test it needs to make the same mistake as human would and intentionally add mistakes such as typing errors. As such the Turing Test is not testing real intelligence behavior, but rather human behavior which at times can be very different things.

Source: http://en.wikipedia.org/wiki/Turing_test#Human_intelligence_vs_intelligence_in_general

3) Can virtual agents succeed in delivering high-quality customer service over the Web?

Virtual agents has been used by many companies to provide customer services in place of real human. How successful it has been really is dependent on the nature and the complexity of customer service provided. For example the many banks has virtual agents to answer banking inquiries from customers. For simple inquiry such as how do I reset my password or where is the nearest bank branch to a certain location virtual agents has no problem in handling them. In fact it probably offer a better experience than real human as the information is returned instantaneously with links and instructions/solution to your inquiry. The advancement in technology such as voice recognition, better indexing and search strategies and a consistent answer has really increased the usefulness of virtual agents over the years. However the capability of the virtual agents is still pre-determined by its creators, it can only provide the customer service that it is programmed to provide. If it need to provide a customer service that was not programmed into its strategy or if it feels to understand the needs of the customer, then it will be at loss at what solution they can provide the customer with or in some cases providing the incorrect solution. For example if

If your inquiry is very personal in nature and are complex then virtual agents will not be able to help you. For example recently I used virtual agent to help me with my inquiries for a hotel booking. It was a great experience at until the very end where I had to put in a certain special request that the virtual agents keep on giving me the wrong information. After couple of times of back a forth it suggested I call the hotel for clarification. I got my request answered by the hotel when I called them the very next day. However in this case it wasn't a high quality experience from the virtual agents due to its inability to deal with variables that it wasn't programmed to solve. Instead of helping me, it wasted my time and got nowhere.

In conclusion, I believe virtual agents can provide a high quality customer experience in certain situation if it is programmed to handle the particular service. However at its current state it is incapable to fully replace real human in dealing with complex and difficult customer services.