All In for AI First

If your businesses’ website has been redesigned sometime in the past decade, you’ve probably heard the phrase “mobile first”.  The idea is simple – design websites and applications with mobile phones in mind first and then scale from there. It’s a concept that started to take hold as consumers began accessing content on their mobile devices more than their desktop computers.  You can start to see a rise in searches for “mobile first” in 2010, soon after Google began pushing the design philosophy:

Seven years later, mobile first is the accepted doctrine, but we’re ready for a dramatic shift.  Over time, smartphones have only increased their speed and processing power, leading us to a new philosophy: AI first.

This past week, Google announced a new product that highlights their shift in philosophy from mobile first to AI first.  Up until now, artificial intelligence applications have been so resource expensive that their algorithms are required to be processed in the cloud.  This comes with a few drawbacks for consumers, though:

  1. Privacy:  Consumers must share their data to see results
  2. Connectivity:  To use AI applications, you must have an internet connection
  3. Speed:  Running on the cloud can take time

As smartphones have become more powerful, it is now possible to perform machine learning tasks locally, on a smart phone. Google’s announcement of TensorFlowLite builds on the work that they have already done with their open source artificial intelligence platform, TensorFlow.  This platform aims to optimize machine learning algorithms so that they may be run on any modern device as opposed to requiring developers to perform the algorithms in the cloud. While processing power on modern smartphones makes this possible, it will still be important for more powerful hardware to be developed to fully realize the benefits of locally run AI platforms.

The development of TensorFlowLite inherently solves some of the issues with AI platforms:  data never leaves a consumer’s phone, an internet connection will no longer be required, and the algorithms will be optimized for speed on mobile.  This will allow your phone to complete some amazing tasks without the need for cloud accessibility.  For example, your phone could help you to automatically tag photos or identify objects in the real world.

At CompassRed, we believe that companies that push to be AI first will have similar success to those that pushed to be mobile first less than a decade ago.

Ultimately, companies that adopt an AI first strategy will be able to offer the most personalized experiences for their customers, draw the deepest insights, and guide their companies with the most powerful tools.  Similar to how mobile first allowed for companies to provide customers with the highest quality experiences, an AI first strategy will do the same.

The push towards an AI first philosophy does not diminish the importance of being mobile first.  Rather, it provides companies with the opportunity to provide an even better mobile experience for their customers with the addition of AI tools to their applications.  Right now, utilizing artificial intelligence can be difficult for all but some of the largest companies.  Creating open source artificial intelligence platforms will help to democratize AI for all companies – even the smallest ones.

With a new paradigm comes new strategies for businesses to employ.  In the coming weeks, we’ll explore ways that companies – large and small – can utilize AI first strategies across their platforms.

This post originally appeared on the CompassRed Data Labs blog.

Leave a Reply

Your email address will not be published. Required fields are marked *