Microsoft has made another foray into the world of AI chatbots, unveiling its latest creation Zo just eight months after it was forced to shut down its millennial bot, Tay, after users turned it into a racist.
Zo is currently being tested in messaging app Kik, with users who have opted to try the service saying the bot is effectively an English language version of Microsoft’s existing Chinese bot, Xiaoice.
First spotted by Twitter user Tom Hounsell, Zo is powered by AI and isn’t available on any other platforms as of yet – which isn't surprising given how quickly its predecessor Tay was corrupted Twitter users.
Tay was taken offline to "make adjustments" in March after living on the platform for just one day and falling foul to trolls who took advantage of its conversational learning abilities.
The bot had been modeled to speak like a "teen girl" but instead used its platform to praise Hilter and Donald Trump and share 9/11 conspiracy theories after it was taught how to say a plethora of unsavoury statements by Twitter users.
However, with Zo Microsoft has been cautious to ensure it avoids topics like politics, and following a “super abbreviated personality test”, Zo is able to chat conversationally with select Kik subscribers.
Whereas Tay was trained to speak like a millennial, Zo appears to have been designed to appeal to a broader audience and according to Engadget is “uncannily knowledgeable about Microsoft products, and will profess to being a Windows phone fan.”
Since the start of 2016 chatbots have been heralded by some marketers as the next frontier for brand and consumer interaction, so its no shock that Microsoft would turn to a platform like Kik to host Zo.
Earlier this year Kik said its army of bots, which includes services from the likes of CNN and HP, was gaining traction among users and businesses alike with 1.8bn messages being sent to date between Kik subscribers and businesses bots since it launched the offering.