Understanding artificial intelligence sometimes isn’t a matter of technology so much as terminology. There’s plenty of it under the big AI umbrella – such as machine learning, natural language processing, computer vision, and more.
Compounding this issue, some AI terms overlap. Being able to define key concepts clearly – and subsequently understand the relationships and differences between them – is foundational to your crafting a solid AI strategy. Plus, if the IT leaders in your organization can’t articulate terms like deep learning, how can they be expected to explain it (and other concepts) to the rest of the company?
What is deep learning?
There’s a Russian doll analogy here: Deep learning sits inside of machine learning, which sits inside of artificial intelligence.
Deep learning is a particularly good example in this regard: It’s related to – but not interchangeable with – the broader category of machine learning. This exacerbates the possibility for misnomers and misunderstandings. In fact, there’s a Russian doll analogy here: Deep learning sits inside of machine learning, which sits inside of artificial intelligence.
[ Read also: AI vs. machine learning: What’s the difference? ]
“Artificial intelligence is essentially when machines do tasks that typically require human intelligence. The field of artificial intelligence includes machine learning, where machines can learn by experience and acquire skills without human involvement,” explains Bill Brock, VP of engineering at Very. “Deep learning is a branch of machine learning where neural networks – algorithms inspired by the human brain – learn from large amounts of data.”
Deep learning vs. machine learning
Let’s mitigate potential confusion by offering a clear-cut definition of deep learning and how it differs from machine learning.
“In deep learning, the algorithm is given raw data and decides for itself what features are relevant.”
“Deep learning is a branch of machine learning that uses neural networks with many layers. A deep neural network analyzes data with learned representations similarly to the way a person would look at a problem,” Brock says. “In traditional machine learning, the algorithm is given a set of relevant features to analyze. However, in deep learning, the algorithm is given raw data and decides for itself what features are relevant. Deep learning networks will often improve as you increase the amount of data being used to train them.”
Deep learning is essentially a branch of AI that closely tries to mimic how the human brain works. This is worth keeping in mind – no pun intended – when explaining or evangelizing deep learning to others, especially if they do not possess technical backgrounds.
“Just like humans learn from experience, a deep learning algorithm can perform a task repeatedly, each time tweaking it to improve the outcome,” Brock says. “The term ‘deep learning’ refers to the neural networks having many layers that enable learning. Deep learning can solve just about any problem that requires ‘thought’ to figure out.”
[ How does RPA fit in with AI, ML, and deep learning? Read also: How to explain Robotic Process Automation (RPA) in plain English. ]
How to best explain deep learning: An analogy
It’s also helpful to have a little historical context to set the stage for why deep learning matters – not just to IT pros but to a vast array of people.
“For decades, in order to get computers to respond to our requests for information, we had to learn to speak to them in a way they would understand,” says Tom Wilde, CEO at Indico Data Solutions. “This meant having to learn things like boolean query language, or how to write complex rules that carefully instructed the computer what actions to take. This forced a major imposition on the user and meant that only a relatively few skilled people could successfully retrieve information from computer-based information systems.”
Wilde notes that he regularly fields questions about how to best explain deep learning from customers and others. This is particularly important because non-technical folks may benefit the most from the paradigm shift deep learning promises from traditional computing.
“Deep learning’s arrival flips that [historical context] on its head,” Wilde says. “Now the computer says to us, you don’t need to worry about carefully constructing your request ahead of time – also known as programming – but rather provide a definition of the desired outcome and an example set of inputs, and the deep learning algorithm will backward solve the answer to your question. Now non-technical people can create complex requests without knowing any programming.”
Want another analogy? Think of a young child learning language.