Commentary: Different Users' Needs Require Different Devices

One thing I have come to learn is you must understand a problem before you can provide a solution.

Working in government for more than 10 years now, I have come to recognize that many services provided by government and private industry are very different. While this is true and can guide us to an understanding of a best solution, being a government employee, I have to stay with the concept of trying to provide as many possible solutions to accommodate as many citizens as possible. As we all believe in our freedoms to choose, the government should always have a goal to provide a path to as many options as possible.

This is why I believe chatbots should be platform-agnostic. Which flavor is better depends on many factors: correctly identifying the problem, the user’s access to the solution, and which company is providing the solution. Amazon and Google have been in a device war, between the Alexa Echo Dot and the Google Home.

As a consumer device, Alexa wins, mainly because Amazon has linked so many consumer services and purchasing power to Alexa. However, I believe that Google wins the battle when it comes to government, or informational resources, mainly in the GUI-based Assistant.

Leveraging response cards with a format that provides links, buttons, suggestions and more adds power to the tool and better guidance and direction to users; they’re able to gain information quicker and route to calls, maps, websites and even laws, governance policies and ordinances. This is where the AI begins to kick in, and concepts such as context can provide more refined answers to users. Outside of context, this is where Alexa fails to come to the plate, not providing any button, link or suggestion options to users. The Alexa application on my Android also seems to be a bit resource-heavy to run.

There is no reason to pick one or the other; you can actually install both on your phone. I have a few Internet of Things devices using Google Home to operate, but these can be operated from either solution.

Google also has some Betas in its Dialogflow product, specifically the Knowledge Beta and Phone Gateway Beta. Knowledge Beta has the ability to scan FAQ pages and uploaded documents, and it extracts Q&A-type data dynamically and can be refreshed manually. The Phone Gateway opens yet another path into a chatbot deployment with no coding at all. Simply turn on the Phone Gateway and setup a dial-in number. The rest is mainly done already by your prior Dialogflow agent and intents. Being able to provide a response to connect someone to the appropriate desk, through the use of dialing in on their cellphone, is immediate power and direct guidance after the AI is able to determine the right destination.

Sure, it’s a computer, and you might get to the wrong person, but that happens now anyway. Why not at least make it happen faster? You might even get your question answered more quickly by asking twice!