Simon Chan, the product director from Salesforce Einstein wrote a very interesting post about Artificial Intelligence on TechCrunch. It is called Crossing the AI chasm with a nod to Geoffrey Moore’s famous book Crossing the chasm.
In line with crossing the chasm, for ChatBots I read the post Survival of the fittest: Battle of the bots ensues. Below my thoughts about how to cross the chasm for ChatBots.
How to Cross the AI chasm?
Artificial intelligence is improving our lives and businesses. AI is already analyzing x-rays, powering the Internet of Things and recommending best next actions for sales and marketing teams. The possibilities seem endless.
But in practice there are lot of stories and not so much is working yet. As Simon Chang describes it:
Countless projects never make it out of the lab. That’s because putting machine learning research into production and using it to offer real value to customers is often harder than developing a scientifically sound algorithm. Many companies I’ve encountered over the last several years have faced this challenge, which I refer to as crossing the AI chasm.
So to come back to the Survival of the fittest: Battle of the bots ensues.
So what does a very good ChatBot need?
Conversational artificial intelligence
It is capable of holding an intelligent, two-way conversation.
Cognitive artificial intelligence
It understands what people mean and want to do within the domain
Human assisted artificial intelligence
Supervised AI. Using humans as bots “partners” to accelerate machine learning and ensuring that bots are learning “the right things” from humans
A central cognitive brain, deployed across a number of channels – messaging applications, mobile applications, web searches, chat applications and social media.
Intelligent authentication and security
An very easy to use and secure authentication is needed. Probably pins or passwords already have negative effect on the UX.
Simon mentions 4 lessons to cross the AI chasm:
The technical AI chasm
Data is key to AI. For example, if you want a chatbot to learn, you have to feed its algorithm with examples of customer requests and the corresponding correct responses. Such data are often presented in a well-structured, but static, format such, as CSV files.
So if you use only structured and static data, it is way easier to show a fantastic demo. But in practice data is often not structured and needs to be used real-time. By introducing human assisted artificial intelligence in ChatBots the technical AI chasm will be much easier to solve.
In general for AI: Data is the holy grail
So take care about how to collect the enough but the right data.
Quality assurance on collected training data.
Companies should be thinking about data quality from Day One — especially for user-generated data. It’s exciting when machine learning is automated, but it can also backfire.
Also here human assisted artificial intelligence in ChatBots will solve this problem. This way crossing the AI chasm doesn’t have to be intimidating.
The product AI chasm
Optimize for the right goal.
AI success hinges on defining your prediction problem correctly. From the beginning, you need to clearly identify the input query, the prediction output and what qualifies as a good or bad prediction. Data scientists will use these evaluation metrics to determine the accuracy of the AI model.
So important to define your goal. In case of chatbots, the first goal is ofter User eXperience (UX). As long as the UX doesn’t feel like chatting with real person OR maybe chatting with an excellent ChatBot that fulfill all users requirements. Conversation will be the new interface but there should be a conversation possible.
In other words conversational- and cognitive artificial intelligence with the right metrics is the way to cross the product AI chasm.
Humans are complicated. So when they interact with AI, it presents new challenges that don’t arise when dealing with data sets in a lab. Be aware that customers won’t use an AI-powered product if they don’t trust it. And while you can try to build trust by showing how accurate a predictive model is, most consumers can’t really relate to robust scientific metrics.
Logically human assisted artificial intelligence may solve this problem. Also an intelligent authentication and security will gain trust of the user. Besides of that as long as you can give the user a try without the risk to loose anything (like money), the trust can easier be built. Besides an omnichannel integration maybe be a benefit above human interaction. Although you can also realize this in manual chatting systems. In both cases an intelligent authentication is required.
And then I would like to finish with the conclusion of Simon:
The truth is, crossing the AI chasm doesn’t have to be intimidating. Just be sure you approach it with a well-formed plan that keeps you looking ahead instead of down. And remember that to be AI-first, your company also has to be customer-first.
If you would like to know more about crossing the AI chasm for chatbots or share experiences, please contact me via the contact form.