Conversational agents

Language is tough to synthesise on AI models. There are claims that it does “sentiment analysis” on scale, but hard to pull off.

Voice assistants are not doing it for big tech • The Register

But behind the scenes, the industry is also seeing the demise of another popular craze Amazon was trying so hard to make a thing. With its voice-interface device, Alexa, the global ecommerce and tech behemoth in 2014 set in train a fashion for speech interfaces in the home, which had keen households around the world demanding their AI house guests to tell them jokes and answer trivia questions.

But why? That’s just what Amazon may be wondering. According to reports, its Worldwide Digital unit, which Echo smart speakers and Alexa voice technology fall under, hit an operating loss of over $3 billion. Business Insider also claimed the vast majority of the losses were tied to Amazon’s Alexa and other devices.

I earlier read “reports” on voice skills based on Alexa (much like the app store) and realised there could be an opportunity in the healthcare space. Amazon often runs its divisions to sniff out the next best thing (including strategic investments in “start-ups” to leverage its cash cow-Amazon Web Services. However, in these times of layoffs and recessions, it is best to sacrifice something that wasn’t giving them any returns on investment.

Is conversational AI “dead”? Not really. Siri is still around the corner without much ado (only if you consent to send your valuable data to “third-party contractors”) to transcribe and possibly make sense of the verbal confusion. Conversational AI will evolve (slowly) with companies launching their “home-automation” and slowly (and gradually) step into healthcare (at least in consumer facing devices).