19 May 2016 | Team Tamar

Conversational AI is the future of search

google io june 2016 gif

Last night marked the first day of Google’s I/O 2016 conference, a two-day conference where Google shows off their latest ideas, products and developments. Amongst the many announcements this year – which included VR, wearables, new messaging apps (because why let Facebook have all the fun, right?) and video conferencing – came the announcement of Google Home, the widely anticipated rival to Amazon’s Echo.

This, for us, brought us again to the topic of personal assistants, which I have previously looked at. We previously looked at how they will change the future of search – but now we’re recapping on what all the tech giants are doing. Read on to find out.

Google

google home

One of the key takeaway points from I/O 16 was the announcement of Google Home. A Google-powered, voice-controlled product to rival Amazon’s Echo was reported back in December of 2015 in The Wall Street Journal and, more recently, in days leading up to the Google I/O event itself certain articles started to appear, teasing the product.

Google is smart to integrate their AI & chatbot with a speaker. The common problem with chatbots and their ilk is that, while it’s definitely annoying having to swap between a hundred different apps, it’s not that hard. If chatbots mean that you have to type – or they make mistakes – then that’s no incentive for people to swap away from apps.

google home ai search conversation

Plus, Google’s previous experience utilising natural language in search makes them very qualified to understand conversational language as well as conversational context and subtexts. Smart.

Facebook

Facebook led the way with personal assistants – or, at least, appeared to – reports about Facebook M started to appear towards the end of 2015, with select members being invited to trial it in the San Francisco area.

facebook m

Facebook M differs from the other AIs in that it is more of a chatbot – rather than a proper AI. Part of it is powered by Artificial Intelligence, but then the other half is run by humans.

M sits nicely in your Facebook Messenger app (which we’ve seen a growing trend towards over the course of early-2016) and can respond to pretty much any request. M has been able to send GIFs, draw personalised pictures, write songs and even order a small army of parrots.

facebook m assistant

It seems only natural that Facebook would want to do this. After all, they can’t hope to beat Google in search engine dominance, so what do they do? Easy – they cut Google out of the picture. If you’ve got a chatbot or AI that can do everything you need within the Facebook ecosystem without even looking at a search box, who’s the real winner?

Facebook’s motives became crystal clear this year during their F8 conference – they want Messenger to replace everything. Why have a whole heap of apps when you could just have one function that caters to your every need?

Apple

Apple’s current personal assistant – Siri – still lags behind a lot of the competition. Apart from its often-corny jokes, Siri has no real use. Anyone with an iPhone can attest to that.

siri game of thrones joke jon snow

Perhaps Apple are just holding their cards close to their chest? We don’t know. Apple did acquire an AI startup earlier in the year, so maybe they’re planning to announce something at the next Apple event? Only time will tell.

Amazon

Amazon’s Echo device was previously seen as the best personal assistant in the industry. Part speaker and part personal assistant, it’s able to play music, retrieve information, tell the weather, create shopping or to-do lists, set alarms and timers, convert units or even define words. The Echo is able to hear you from up to 20 feet away, and through several rooms of the home.

amazon echo speaker

However, if we’re getting technical, Echo is the speaker that houses the true AI – Alexa. While a lot of Alexa’s functions are quite basic, no more advanced than Siri could muster up, what sets Echo apart is the fact that it seems the most able to appeal to everyone. It is the closest that a current personal assistant has come to being a solid Internet of Things product.

Microsoft

Microsoft seems to have done what the others tech giants have lacked in their ability to scale. Their Build2016 conference was mostly aimed at an army of personal assistants and chatbots, all centered around the main Microsoft AI offering РCortana. 

It looks like Microsoft are well aware that language will become the new typing – with one of the slides from their Build2016 conference boldly stating that;

Human language is the new UI

Very bold. But for now, Cortana is limited. She’s still miles ahead of Siri in terms of ability, but still far behind say, Echo or Google Home, in terms of usefulness. Cortana is mainly limited to either your laptop or a Windows Phone (and who has one of those..?)

microsoft cortana ai

However, Microsoft have recognised this. Part of their main announcement was that they were going to open up the capabilities for developers to build chatbots that interact with Cortana. They have essentially developed the framework and are giving it away for free, which is a smart move. By allowing developers to do the work for them, they can scale a lot faster than if they kept it all in-house.

However, their initial foray into AI was plagued with mishaps. Days before Build2016, Microsoft unveilved¬†Tay – an experimental chatbot that was designed to learn from millennials on Twitter. Anyone with half a brain could’ve seen how that would go.

tay tweets microsoft anti semitic comment

And they were right. Within mere hours, Tay had become a racist, anti-semitic Trump supporter. But what can you expect from a platform that’s notoriously full of trolls? C’est la vie.

 

Team Tamar