And as of Monday morning, Microsoft (MSFT) is offering a free, downloadable deep learning toolkit from Github, the popular software-repository-slash-lending library. Developers can use this toolkit to develop “smarter” applications that can, in theory, learn from their interactions with humans and other computers.

Microsoft’s Computational Network toolkit or CNTK was put together by Microsoft Research’s brain trust to help computers learn to understand human speech faster and better. In a statement Microsoft’s chief speech scientist Xuedong Huang, said this toolkit is “just insanely more efficient than anything we have ever seen.”

The toolkit was already available to academic researchers via the company’s own Codeplex hosting site, but the Github submission, plus the toolkit’s availability under a broader open-source license, could bring it to a much wider audience.

Subscribe to Data Sheet, Fortune’s daily newsletter on the business of technology

What does better deep learning mean to big companies? Well, software that trains itself could be a pretty big labor time saver, and deep learning makes such things possible.

What if your customer service application could understand speech fully? Or if your shopping application could recognize images that customers scan. Or recognize user or employee faces for security purposes?

For more on artificial intelligence, watch the following Fortune video:

Some of Microsoft’s internal projects illustrate what can be done with deep learning. Project Oxford can tell a person’s mood from facial expressions and Skype Translator offers near-real time translation on Skype calls, for example.

IBM Launches Watson Unit to Try to Capitalize on AI

Indeed, big learning or AI, is now a fiercely competitive battle ground where tech vendors are trying to show off their prowess.

Google (GOOG) last year offered up its Tensorflow deep learning technology for free to developers and last week it said it was partnering with Udacity, on free online courses on deep learning.