The Difference Between AI and Deep Learning—And Why You Need Both

on Monday, 01 May 2017. Posted in Blog

The Difference Between AI and Deep Learning—And Why You Need Both

(Also posted on Linkedin https://www.linkedin.com/pulse/difference-between-ai-deep-learningand-why-you-need-both-antonatos)

If you were to name the buzzword of 2017 in the tech space, artificial intelligence would definitely be at the top of the list. AI has already made its way into our smartphones, homes, and gadgets, and it continues to transform other sectors of society like ecommerce, healthcare, and more. It’s the next “big thing” in tech, just like mobile platforms have been in the past decade, and internet-based platforms in the decade before that. AI is such a big deal, in fact, that just recently, eBay’s CEO Devin Wenig said “If you don’t have an AI strategy you are going to die in the world that’s coming.” 

AI may get all the buzz in the industry, but what most people don’t recognize is what’s really behind the rise in AI—and that’s Deep Learning. 

What’s Deep Learning, Anyway?

Deep Learning is what goes on behind the scenes to make the AI we use on our devices, gadgets, and in our homes possible. In technical terms, it’s defined as a class of machine learning algorithms based on learning multiple levels of features or representations of the data. Experts refer to this “artificial neural networks” with huge amounts of data in hierarchal representation. 

It’s called “deep” learning because it’s structured in many different layers, and these layers build on each other to interpret sensory data for a given output. In other words, it takes data like images, sounds, text, etc. and processes it through series of these hierarchal layers to produce the output that we ask for. If you’re interested the in-depth explanation, here’s a great resource

Examples of Deep Learning you might already be familiar with are things like facial recognition in photos or sentiment analysis in text (like you might see on Facebook). But the potential applications for Deep Learning are practically endless and are applicable in practically every sector—from manufacturing and automotive, to communications, security, finance, and more. 

Why Having a Deep Learning Strategy is Important

Having an AI strategy is great—and will be a necessity to keep up with the market in the near future. But without a Deep Learning strategy, your AI strategy won’t be able to keep up. It’s the advances in Deep Learning that produces that improvements in AI we see and use on a daily basis. Big names like Google are already diving in to improving their use of Deep Learning technology—just in the past 2 years the deep learning used in their apps has grown exponentially. And companies who want to stay ahead of the game will need to follow suit, too. Now is the time to start getting involved with creating and growing your Deep Learning strategy rather than waiting for the technology to catch up to what you hope or need it to be able to accomplish. Its algorithms are improving so quickly that starting sooner rather than later means you can take advantage of these improvements as they develop, rather than playing catch up after the fact. 

Where Do You Start?

Before jumping into the AI space just because it’s “the next big thing,” it’s smart to determine which tasks it would be most efficient for. Despite the advances in AI we’ve seen just in the past year or two, it still isn’t the most efficient choice for all processes. In some instances, human intervention is still the best option. If your business has tasks that:

·      Are repetitive

·      Are high volume

·      Follow patterns 

·      Require limited judgement

·      Have a low cost of mistake

Then these are the best places to start implementing a Deep Learning and AI strategy, while leaving the more complex tasks to humans. That’s not to say however, that both AI and human intervention can’t work together, though. 

For example, in the marketing space, AI can help personalize offers to consumers using micro segmentation and customization, but decisions about marketing strategy that will heavily impact revenue or market share are probably best left to humans. In the customer service space, AI systems are great for creating better responses to transactional requests, but decisions for that customer still need to be based on empathy which means human intervention is still a necessity. In sales, AI chat bots can be effective as aids to the sales process, but an actual sales person will still be much more in influencing buyer decision, especially with high ticket B2B scenarios. 

However, no matter the sector, AI is sure to disrupt industries more than we can probably even foresee today since it becomes more advance by the day. As an example, scientists in the U.K at the University of Essex and experts at Orbital Media just rolled out a new AI system to aid the NHS in providing customized medical advice and tips to patients. It’s easing the burden on doctors, helping people with self-treatable and minor issues find solutions to their problems without requiring human intervention.

 

Copyright 2015 Appsconsultant.com. All rights reserved.