It’s difficult to measure the effects that the current global crisis has on the future of technology. Rather than trying to take in an overwhelming flood of information or to predict broad-scale outcomes, we’ll take a look at a single field and how covid-19 is affecting it, and vice versa: conversational AI.
When we think “conversational AI” our first thought is probably of voice assistants like Alexa and Siri. And for good reason; according to this article, “Conversation is an incredibly hard problem for algorithms to master.” Which means we don’t see it very often in our daily lives, apart from our smartphones’ voice assistants. But that’s rapidly changing.
In 2019 alone, smart speaker sales hit a record high, with 146.9 million smart speakers sold in the US. 70% of those sold were Amazon speakers. It’s no surprise then that coronavirus-related skills were swiftly added to Alexa to meet unique demands—but not by Amazon.
In an effort to combat misinformation, both Amazon and Google wiped all coronavirus-related skills and actions from their voice assistants in March (which denotes an important, if unsurprising, mindset), with zero tolerance; however, new skills are currently available. One such skill answers questions about covid-19.
The Mayo Clinic even created the ‘Mayo Clinic Answers on COVID-19’ Alexa skill as a trustworthy source of reliable answers to coronavirus-related questions.
Other uses of conversational AI include education, entertainment, hobbying, and even combating loneliness in the midst of self-isolation. The first episode of VOICE Talks discusses more about the use of voice assistants during the pandemic.
Chatbots for Hospitals
Voice First Health recently held a webinar on how voice first technology has been used during the pandemic; a number of those videos also discuss the potential applications of voice tech in the health sector, such as diagnosing and screening patients.
Right now, conversational AI is being used to triage or screen patients. Many hospitals are even using chatbots and helplines to reduce incoming patient overload (Boston, France).
The nature of this sector means that chatbots are handling private data or handing out information to possible patients. Unfortunately, not every company offering chatbot solutions is consistent about their sources, or usage of private data.
One recently published paper describes a “question-answering dataset that...could help gauge the accuracy of chatbots and search engines that answer topics about the novel coronavirus.”
Conversational AI materialized promptly in the health sector because of the staggering requirements, but it has a ways to go before it can adequately meet both safety and privacy demands.
We are seeing gaps and pain points that either didn’t exist before, or didn’t exist on the kind of urgent scale that drives mass adoption and enterprise implementation. Conversational AI by itself is proving to be the scalable solution to a lot of issues, even without taking other AI tools into account.
The question is, where will this technology take us as consumers and businesses on the other side of the coronavirus crisis?