What Is Deep Learning?

Deep Learning describes the mathematical calculations that occur within the hidden layers (hence, “deep”) of an Artificial Neural Network. ANNs are modeled after the human brain and nervous system: a group of neurons makes up a layer, and calculations traverse layers as neurological signals across synapses.

Input data for immediate correspondence.

Deep Learning is the catalyst propelling human progress exponentially forward—fostering major advances in medicine, commerce, biotech, aerospace, energy, and even human consciousness.
“The fourth industrial revolution, and the future of AI, is not going to be defined solely by our technological advancements, but by our ability to meaningfully apply this technology to solve real problems.”
- Ben Lamm


Deep Learning takes advantage of the physiological efficiency of a neural network- like the human brain-to rapidly and simultaneously calculate large amounts of data.


    Data is analyzed in an encoder/decoder model in which input is encoded, compressed, decoded, and then output as a copy of the input—e.g. making copies of flyers or handwritten notes


    Each subsequent hidden layer calculates inputs then cycles the new data through the previous layer—forming a time-based information as a sort of memory to be predictive of future time-based information—e.g. using the first words said in a sentence to predict the rest of the sentence.


    The network identifies and classifies a dataset by hierarchical layers, or frames, that increase in complexity—e.g. identifying letters in a handwritten note

The difference between Machine Learning and Deep Learning is that Machine Learning is a field that applies algorithms to big data to teach technology how to learn from data then make decisions based on what it learns, while Deep Learning is a series of mathematical calculations that occur in the hidden layers of a neural network within an ML algorithm. Artificial Intelligence describes any technology that simulates human behavior. Taken together, Deep Learning informs Machine Learning which informs Artificial Intelligence.



In December 2016,

Amazon opened Amazon Go, an almost fully automated grocery store (customer service reps are available for general questions), and in January 2018, it opened in Seattle. The store does not have cashiers or checkout stations. Instead, it operates through surveillance that intakes visual and sensory data, applies deep learning algorithms to recognize Amazon Go subscribers and products, and applies purchases to subscriber accounts.


In October 2017,

Google launched Pixel Buds—earbuds that perform real-time translation in the wearer’s preferred language. Google Assistant acts as the liaison to help the user select which language to translate. Alternatively, the user can speak to Google Assistant to ask for information, directions, or to dictate and send text messages. Natural Language Processing is progressing exponentially and honing its ability to recognize human language—with all its nuances and imperfections.


Bosch partnered with Hypergiant to experiment with applying Deep Learning to create art.

They incorporated device data, data from Mapbox and Foursquare, and municipal data. The resulting images are enchanting, eerie renderings of an urban landscape that represents human connectivity to one another, humans’ connection to the environment around them, and the true reality of the Internet of Things (IoT).

Machine Intelligence Techniques