I had coffee last week with Alar Kolk, one of the smartest guys I know. Among other things, Alar is with the the H-STAR project at Stanford doing research on the future of interaction between people and technology. 

Alar asked me if I thought we had reached “peak mobile”,  and now the mobile app business would be going into decline. Huh? I was confused by his question because people are using their mobiles more than ever, and all the research I’ve seen suggests that mobile use will continue to grow. But Alar was actually saying something quite different – he was suggesting that while people will continue to use their mobile devices to do more and more things, they will be less and less likely to do so via standalone downloadable apps.

The statistics certainly back up the notion that people are downloading fewer mobile apps. The most recent reports indicate that most smartphone users now download approximately zero new apps per month, as consumers increasingly concentrate their use on a few multi-use platforms that they already have (and people are less interested in having their phones cluttered with ad hoc single-use apps).

So if people will continue to use mobile devices more for more kinds of tasks, but won’t be using mobile apps to do it, what will they be using?

“Chatbots”, said Alar.

The fact is, today a mobile app is usually just simply a front-end interface for the actual software application which is running in the cloud. With many of the complex mobile app projects that Tivix does, for example, a minority of the lines of code we write are running on the phone, while the majority are in the cloud. Think of the fact that when you use Facebook via mobile, tablet, or computer, you’re just touching a thin interface on the device which is interacting with the big Facebook application in the cloud where all the actual logic and data reside.

So if everything is in the cloud why are we still using our clumsy fingers to use touch tiny screens on our tiny little devices?

I’ve written before about how speech/text may well become the next powerful software interface. We’ve had the Graphical User Interface as a way to operate software for 30 years now, and we’ve tried to condense that paradigm down to smaller and smaller screens, making good user interface (UI) design more and more difficult. Perhaps it’s time for a new paradigm, where we use software in a more natural way. As Alar pointed out to me, recent advances in artificial intelligence and speech recognition have been enormous and these are coming together in powerful ways. 

So rather than needing to open an Uber app on your phone, you could just say “get me a car” and the Uber platform in the cloud has the intelligence to know your location, your customer history, car availability, your preferences, etc. No app required. 

Or, instead of opening up the OpenTable app to book a restaurant for dinner you could just simply text the restaurant saying “need a table for 4 at 6pm” and have the cloud-based chatbot text you back to confirm or offer alternative reservation times (Hanahouse in Palo Alto is already doing this as a way to book tables).

Intelligent chatbots as a way to interface with software. Wow. As Accenture says in their 2017 tech vision paper, “AI is the new UI”. 

The use of mobile devices to perform complex tasks will continue to increase, but we may start doing that in new ways. Gartner Research, in fact, says that by 2020 more than 30% of our web browsing will be without a screen.  Think about that for a moment. 

As the current trend lines come together, it seems clear that mobile smartphone use and increasingly complex cloud applications are definitely the future. But downloadable mobile apps may be history.