Posted in: Archive
Some of the most interesting tech startups popping up a the moment, are for me, anything to do with chat bots. This year, of all years seems to have been something of a bumper year for them.
We’ve seen Ben Parr co-founding Octane AI , Trello launching GoMix as early as last week, and a whole host of others. Microsoft is making big bets on chatbots with its Bot Developer Platform, as are companies like Facebook (the Messenger Platform), Apple (Siri) , Google, and Slack. There is a new wave of startups trying to change how consumers interact with services by building consumer apps like Operator or x.ai, and fully featured bot platforms that allow you create your own bots with out coding like Chatfuel, Motion.AI, Howdy and Botsify.
There’s no shortage of a whole host of other supporting technologies, with the land grab for the bot platform space continuing amongst the bigger platform players. Since Facebook’s chat bot platform launched back in April, more than 11,000 bots have been added to Messenger.In May, David Marcus Facebook’s VP of product said that “tens of thousands” of developers were working on Messenger bots.
That’s phenomenal growth, and is a trend, not an anomoly. To put it mildly. Bots are hot.
Chatbots are finding their feet because they are now the connecting bridge between humans and APIs, essentially helping turn every user into a human client consumer app.
More and more companies are now investing in building out bot infrastructure to automate tasks and solve problems for their customers. After all if you can code a bot to find out how much stock is left on a product, or trade shares and stock that’s where the major customer service wins are to be had. Code the bot to recognise particular input parameters, and perform the action for you via APIs. If the amount of chatbot services popping up over the last year are anything to go by, that’s where the money is flowing too.
Major companies are putting efforts into building out infrastructure solutions for one simple reason. Chatbots are also the missing link between real time voice control and backend systems. Building an automated solution that interacts directly with your business via someone typing at a keyboard today will undoubtably mean that someone will be able to get the same information via speech technologies tomorrow. Facebook and Google out of all the companies mentioned understand and recognise the importance of this.
If you own the chatbot infrastructure and have a number of systems on your platform, then it stands true that talking with your phone to perform physical actions on the services you choose to link is a logical next step. In fact we are already there to a degree, with Amazon’s Alexa and Google Home both already interact with IFTTT, and I wouldn’t be surprised to see an acquisition on that front soon. Silicon Valley based bot platform Gupshup yesterday partnered with Google to enable enterprises to quickly build ‘Actions’ on the artificial intelligence–powered Google Assistant. Microsoft have launched their own competitor (Flow) this year as well, a further sign of the dot joining process. When Facebook tell us that their voice recognition department have now shipped a component for the Messenger Platform – that’s when we’ll really see a flood of conversation chat bots on the web.
As long as everyday services are building out automated chat bots that interact with their own systems that allow us to get stuff done, then voice interaction with multiple services across the web is a distinct possibility. Rather than writing an IFTTT integration that just solves a single interaction problem, everyday companies will be writing chat bots that a third party will be able to interact with via voice commands. The heavy lifting of the machine learning technology that recognises your voice abstracted away by Google or Facebook APIs and platforms.
Whereas IFTTT performs a single API action at a time. e.g. “Turn on the lights” > action, “Play my music Alexa” > action, Chatbot infrastructure is going to facilitate much more granular commands on everyday services that as of yet haven’t written a full IFTTT integration. It stands to reason, that these will be the conversational triggers that will help power future API interactions with the services we know and love via voice commands.