Google is giving back — sort of — by sharing a core piece of its technology with the rest of the world.

TensorFlow is the second-generation machine learning system that underlies many of Google's business and software. By open sourcing part of the code that makes it tick, the rest of us can put it to work, too.

There's a catch though. Google, of course, can't give away all of its secrets. That's why TensorFlow's release into the wild only includes part of the code that allows it to run on a single machine. Despite being shared under what's called an Apache 2 license (meaning anyone is free to use it), the initial open source version will be limited to run on only one desktop, laptop or smartphone.

What we won't have access to is how Google uses TensorFlow across the entire Alphabet empire. To do so at scale at an enterprise level would require an exceptionally unique hardware infrastructure that Google rightfully keeps to itself. Nonetheless, Google could have just simply shared a research paper describing TensorFlow, as its done in the past with older technology it had already moved on from, but it didn't.

By making TensorFlow public, Google has allowed the rest of the machine learning community to catch up to the five to seven years it had on the rest of the world. Regardless of its advantage, Google says they can't do it alone.

"Machine learning is still in its infancy — computers today still can't do what a 4-year-old can do effortlessly, like knowing the name of a dinosaur after seeing only a couple examples.... We have a lot of work ahead of us. But with TensorFlow we've got a good start, and we can all be in it together," says the company on its blog.

Without doubt, Google sharing such an integral piece of its technology will win the company some goodwill from the general public and the developer's community. At the same time, should TensorFlow become the adopted standard in artificial intelligence, Google would be getting help from non-Googlers building up the software it already uses. For Google, it's win-win.

ⓒ 2024 TECHTIMES.com All rights reserved. Do not reproduce without permission.
Join the Discussion