Google has open-sourced its TensorFlow machine-learning software in an attempt to become a leader in AI development (Photo : Facebook)
Google hopes to become a world-leading artificial intelligence (AI) developer, causing the Alphabet subsidiary to open-source its TensorFlow machine-learning software. The tech giant announced that it is the same in-house system that it has used to support its search engine's AI, and other mathematically complex programs.
IBM's Watson supercomputer defeated the all-time top "Jeopardy!" champions about four years ago, according to The Christian Science Monitor. However, no computer system has truly matched humans' problem-solving skills.
AI is a critical tech tool with several applications. They include Facebook's facial recognition software that automatically identifies and tags uploaded photos, self-driving cars from Google and Tesla, Facebook's Oculus virtual reality headset, and personal assistants including Apple's Siri, according to The Salt Lake Tribune.
TensorFlow is a library of several algorithms that allows Google to train computer systems known as "neural networks" to think like humans do. The neural networks use complex math operations on arrays of data known as tensors, to find patterns and connections by processing data.
Google's public release of TensorFlow now lets anyone download and change the software. It shared in a blog post that the program supports RankBrain, which supplies AI power to Google's search engine.
Several tech companies including Google, Apple, Facebook, and Amazon are racing to develop leading-edge AI technology. This has involved support of research labs, academic publications, and conference talks.
Facebook has also released free source code for its Torch AI development tool. Intel and several other companies use it.
Google spokesman Jason Freidenfelds explained that the company hopes that open-sourcing TensorFlow will make it researchers' go-to toolset. It could also help the company to find tech talent.
TensorFlow's users have described the software as being powerful but time-consuming to program. Google's support of the open-source community will be key to how widely researchers use it.
