Source: Google. 

Most of the time, when tech companies have something to announce in the artificial intelligence (AI) world, it's about a new company they've acquired. Apple's (AAPL -0.81%) recent purchases come to mind. But Alphabet's (GOOG 0.56%) (GOOGL 0.69%) Google made a different AI announcement earlier this week when it said that it's opening up its TensorFlow deep learning system for others to use -- for free. 

Don't feel bad if you've never heard of TensorFlow; most people haven't. But you've likely used its technology if you've ever used Google's app to make a voice search, used Google Photos to search though your pictures, or (more recently) allowed Gmail to send a computer-generated response with Smart Reply.

Deep learning is a part of AI's machine learning, and allows computers to add to their knowledge based on new inputs of information. So when Google Photos starts searching through photos, it learns more about what it's looking at and what it's looking for. 

And now Google wants others to use it, too. 

Give and take
With Google opening up TensorFlow to other companies, developers, etc., the company is allowing the deep learning to expand, as well as make certain aspects of artificial intelligence a bit smarter. Of course, Google isn't giving away all of its AI secrets, but it's likely enough to help other companies create better systems themselves. 


Source: Google.

But why give it away for free? Just as with many other Google offerings, the company does receive something in return. In this case, Google's deep learning system will have access to more information, and thus will be able to learn more, faster. 

For Google, giving away some of its technology in order to gain something greater is ultimately worth it (Android comes to mind). By opening up TensorFlow, Google may be able to stay ahead of Facebook, Apple, and others in the AI space.

The competition
The AI sector has been heating up lately, and Apple has made some notable acquisitions to expand its own artificial intelligence experience. Recently, the company purchased VocalIQ, a deep-learning company based in the U.K. which develops voice recognition software. With each new conversation, VocalIQ's system better understands how humans talk, and then applies that knowledge to voice-recognition commands. 

In doing so, voice recognition software on mobile devices (and perhaps even Apple's rumored electric car) will be able to understand commands without the user having to choose from pre-selected phrases.

In addition to VocalIQ, Apple also purchased Perceptio back in September. The company specializes in deep learning that can run on mobile devices, but without having to collect lots of user data. Similar to how TensorFlow looks for images in Google Photos, one of Perceptio's main focus was image classification. 

Unsurprisingly, Apple's pretty tight-lipped about how it may or may not apply its new acquisitions to new technologies, but it's clear that both Apple and Google are making bigger and bigger AI plays to better compete with each other. 

Lead by example
Google's not going to directly increase its revenue by giving away TensorFlow, but it could gain valuable knowledge by allowing the system to be  used by others. And if it's able to use that to build better artificial intelligence systems than Apple and other rivals, then it'll have been huge success. Google's already shown how well TensorFlow works with its own apps, and now it's time to see how others will apply its algorithms. If all goes to plan, Google's opening up TensorFlow could help push the company into the AI lead -- if it's not there already.