AWS, Microsoft work together for open AI ecosystem
AWS and Microsoft don't often appear in the same sentence unless its about competition and the cloud, but the two companies turned heads Thursday with the announcement of a deep learning partnership. The two companies launched Gluon, an open source interface to simplify the construction of neural networks, according to a Microsoft announcement.
Gluon will open up machine learning to more developers with simpler code and pre-built components without a sacrifice of speed and performance, according to AWS. Integration of neural network with training models will facilitate development, quality assurance and reuse.
The companies hope an "open AI ecosystem" and tools widely available for developers across platforms and languages will grow AI accessibility and applications, said Eric Boyd, corporate VP of Microsoft AI and Research.
The enterprise has high hopes for deep learning. The technology is expected to be an integral part of developers' tool kits by next year and a driver of performance, fraud and failure predictions by 2019, according to Gartner. Early investors in the technology are expected to reap short and long term benefits, and AWS' and Microsoft's partnership will challenge competitors looking to stake their claim in the market, including Intel, NVIDIA and IBM.
Microsoft launched a real-time deep learning acceleration platform for the cloud in August, an important step for the tech because in the past quickly processing AI was not an option. With most companies on or planning to move to the cloud, the ability for AI and deep learning to have real-time capabilities is growing because cloud infrastructure processes live data streams.
This partnership is not a first for the two companies. AWS and Microsoft partnered with Google, Facebook and IBM in 2016 to foster public understanding of AI and research and recommend best practices.
The two companies are also not the first to try opening up the AI and deep learning space for more IT specialists. Andrew Ng launched an online deep learning course with the goal of training millions of new AI experts in August. With a well document and widespread shortage of tech talent, ground up initiatives and friendlier platforms for development will be key to AI's expansion.
Follow Alex Hickey on Twitter