Conversation Bytes: Ever thought of a great conversation with AI?
What people should know when they use AI to interact with humans. Since communication is a two-way art, we want to make sure if it is understood by others. Trust me, it’s not easy! The diverse experiences leading to several interpretations decide the way the information is received by others. Knowing your audience, determining what information would be well received contributes a lot to great conversation. Grice, an eminent philosopher, and linguist outlined four maxims of conversation, which made me introspect my conversation bytes.
Sharing those 4 bytes below!
Maxim of quantity
Adding unnecessary information
- can hurt your business
- can overload the person you are conversing with.
You certainly don’t want the person to not get the essential information you mean to communicate.
Being as informative as possible and avoiding extra information.
Maxim of quality
Trying to be honest in not providing a false message to one’s knowledge or information not supported by evidence.
However, being diplomatic & a kind human tends to help with business and customers (dealing with humans overall). It does not mean we gotta lie. The former is mastery of people’s emotions, while the latter is just a cheat sheet.
Being authentic
Maxim of relation
Saying relevant things
Have an interesting example of conversational agents being a pro at it
When it doesn’t recognize my accent or misinterprets it and providing wrong information instead of confirming what I said -
How automated calls overcome this — ‘is this what you asked?’ Press 1 to confirm or wait to speak to a human who can understand what I might not understand. Teaching AI to know its limitations.
Well-crafted and thought-through information is well received. The sole purpose we communicate is to make others understand what we say.
Maxim of manner
If we are confused with what we want to deliver. The receiver would be super confused. Presenting it to the receiver in a logical manner and briefly.
When a person says, I think I’m not putting it right. This means they are trying to grab their vocab to express themselves. With this limitation of us —
Being more empathetic and compassionate can lead us to significant conversations.
What should conversational agents take into account?
We use AI to automate calls in customer care; our homes are smart with assistants, appliances. So what makes a great AI conversation? In continuation of responsible AI in the previous article. I’m tempted to say, in humanizing design —
- How can we apply these 4 principles?
- How should we design conversational agents to be more human-like?
- What language constructs should be used?
- How should AI get to know your audience and tweak conversations
- When do we see an AI thinking through all these constructs when our ancient nature of communicating still runs into such issues every day. Can those masters teach AI to master? So, we can be taught in turn.
I have my friend’s 2-year-old trying to say ‘Ok Google and Hey Siri’ all the time. I believe the best time to instill the seeds of such a revolution. Fingers crossed to learn more. Please inbox me if you have exciting reads on the same page!
Semantics (Trust, being accurate) play a vital role in humanizing.
Syntax (Communication) helps them to do it better.
Thanks to my mentor for sharing Grice’s Maxim and filling my knowledge gap. Blessed to have people like you in the UX community.
References
- Grice’s Maxims of Conversation: The Principles of Effective Communication, https://effectiviology.com/principles-of-effective-communication/#Using_the_maxims_of_conversation_as_guiding_principles
- 7 Benefits of Conversational AI for Customer Service, https://www.replicant.ai/blog/7-benefits-of-conversational-ai-for-customer-service/
- What Makes a Great Conversation? https://www.replicant.ai/blog/what-makes-a-great-conversation/
