Languages Magazine

Neural Networks & the Virtuous Cycle of Innovation

By Expectlabs @ExpectLabs

By the time the mid-2000s had rolled along, neural networks had mostly been relegated to an interesting CS footnote, but in 2009, Geoffrey Hinton and several of his University of Toronto students published a report that took the world by storm, demonstrating the latent potential of neural nets and recapturing the attention of tech’s top innovators. Since then, progress has been astonishing, with yesterday’s science-fiction what-ifs fast becoming today’s business section headlines.


The graphic above portrays the virtuous cycle of innovation – how innovation, once kickstarted, self-propagates in a feedback loop. In the case of A.I., the costs of GPUs and data storage have fallen at the same time as mobile apps have collected unprecedented quantities of structured user data. This has led to an incredible burst of discovery in the areas of speech, image, and video processing.

Back in January, Expect Labs CEO Tim Tuttle, who has worked in A.I. for over twenty years, spoke with CNBC about the groundswell of voice innovation industry leaders predict. He expounds that all of this will be hastened by an increasing number of wearables – an estimated 3 billion by 2020, less of 5% of which will possess a traditional keyboard. Check out the video here.

Back to Featured Articles on Logo Paperblog

Paperblog Hot Topics