The AI-first world requires new products

It's not enough to simply add an Alexa skill to an existing product, like it wasn't enough to make your website responsive or launch a mobile app without changing anything else. AI-first means reimagining your product for a world driven by AI and voice interfaces.

Last year, Google announced a shift from mobile-first to AI-first. Others, like Microsoft and, to a lesser degree, even Apple followed suit. This new rallying cry indicates a huge shift for the tech industry.

Of today’s tech giants, only Apple and Microsoft are children of the PC age, the first cycle of personal computing (1975-1995) they helped to create themselves. Amazon, Google and Facebook came along with the web age, the second cycle of personal computing (1995-2010). To survive, all five of them had to adapt to the third cycle, the mobile age (2010-2020). With iOS and Android, Apple and Google won the mobile platform battle, while Microsoft, Amazon and even Facebook struggled, before they finally figured out how to thrive under the new conditions.

Now we are about to enter the fourth cycle, the AI age (2020-2025), and that’s a new ball game. For the new age, AI-first is what mobile-first has been in the early years of the mobile age. Each cycle is dominated by a certain combination of hardware and software. The desktop PC with Windows and Intel inside was superseded by laptop machines and the browser, only to be replaced themselves by smartphones and apps. And now we’ll probably see voice-driven hardware taking over, powered by AI systems.

Of course it is true that no new generation of hardware and software completely erases the older ones. Some people still use desktop machines today, but more laptops have been sold than desktop PCs for years now. And tablet sales overtook laptop sales years ago. Not to mention mobile devices in total – 2010 was the first year with more smartphones being sold than laptops or desktops, marking the dawn of the mobile age. Compared to a whopping 1,495.36 million smartphones sold in 2016 alone, the voice-first device market is still tiny, with 24.5 million units predicted for 2017.

But needless to say, AI-powered assistants are already on smartphones, tablets, laptops and even desktops today. This means usage by now is way ahead of hardware sales, and with more people getting used to the new interface paradigm, devices might take off soon. Less than ten years from now, we could probably live in a world where we talk to HomePod while at home and use AirPods and Apple Watches on the go. Of course, the same scenario plays out with Google Home, Amazon Echo or other devices.

AI-first is a huge challenge for product engineering, product design and product management. Even Google estimates that only 10 percent of their 25,000 engineers are proficient in machine learning (ML) at this point. How many of your engineers can master ML/AI? For product designers, the shift is even greater. No more Photoshop or Sketch – the design of voice interfaces is a new frontier for UX design. And product managers need to rethink their products from the ground up.

It’s not enough to simply add an Alexa skill to an existing product, like it wasn’t enough to make your website responsive or launch a mobile app without changing anything else. AI-first means reimagining your product for a world driven by AI and voice interfaces.

And first and foremost this means developing entirely new products – the next generation of transformational products.

Photo by Alice Donovan Rouse on Unsplash