Deep Learning for Exponential Executives: What to Expect?

The Exponential Organization Executive (ExO) of the future will not just be the CIO, CISO, or VP of IT. ExOs will be business leaders who understand how to take advantage of information in their businesses.

They will understand Data (big data) both within and with their cloud providers. They will understand that IP intellectual property of this data can be exposed, developed, segmented, etc. by using APIs, Algorithms, Machine Learning, Deep Learning, and Artificial Intelligence. ExO Executives who “grock” this will apply significant advantage personally and to their businesses.

The Impact of Deep Learning.

In 2011 Google began a project known simply as Google Brain.

The intention of this has been to mimic certain aspects of the human brain, and in doing so create an artificial intelligence the likes of which we’ve never seen.

What started as a handful of experiments has now reportedly evolved into over 600 different projects.

One of which, TensorFlow is an open-source AI software that allows external teams to work with and train Google Search Engine. It can now translate between dozens of languages, read handwritten text and believe it or not, create original artwork.

This program is providing Google an unbelievable advantage and the future implications of these technologies are huge.

But Google isn’t the only company investing in this type of Artificial Intelligence, known as deep learning. Facebook, Microsoft and Chinese Search Engine Baidu are all developing this technology, which is only going to propel it forward in the coming years.

So Exactly What is Deep Learning?

First you have to look at machine learning which is essentially the ability of computers to learn without being programmed. Deep learning is a subset of computer science that evolved from pattern recognition methodologies in artificial intelligence.

Deep learning is a subset of machine learning that has been inspired by the structural neural networks of the human brain.

It is already playing a part in our lives in the areas of:

  • Automatic speech recognition
  • Image recognition
  • Natural language processing
  • Drug discovery
  • CRM
  • Bioinformatics

Deep learning works by having human beings input data for analysis, the machine then has error-correcting feedback included, and the system is able to improve itself (theoretically without the need for human supervision).

One way to think about it is to picture a machine that learns in layers – using a series of processes to attempt to build and train neural networks.

To put it simply, the machine take a set of numbers as inputs, which can represent pixels, audio waveforms, or words, and runs an array of algorithms to try and predict outputs. Facial recognition software is a popular example of deep learning.

How is it Applicable to Exponential Organizations?

The future of deep learning is looking positive, and many believe that we will come to solve big problems in the field of AI on the road to singularity.

As deep learning has quickly developed from being a special purpose machine to a much more general tool, we’re seeing broad implications for organizations across the board.

For example, any CIO whose team is reliant on data analysis should understand the importance of automated algorithms which are really at the heart of deep learning.

Whether you are working for a company that relies on domain expertise for your competitive advantage or a start-up that is looking for hidden gaps to enter tough markets, deep learning could provide enormous opportunity.

Here are some of the things you can expect from deep learning in the near future.

1. IoT and Big Data will Become More Accessible

Right now most small businesses and start-ups don’t have access to the kind of data that powerhouses like Google and Facebook do.

However, with advances in algorithms, crowdsourcing and open source tools, deep learning will be accessible to a wider market and you will be able to gather much more accurate data sets in regards to your customers.

This will only speed up as the trend of open-source software becomes increasingly accepted and an abundance mentality begins to dominate the business-tech landscape.

2. Deep Learning will Become More Mobile-friendly

The models of deep learning that are prominent now are mainly optimized for computers, though this is changing.

Google, for example, has partnered with machine vision tech company Movidius to contribute to a neural network technology road map that could in the long run allow smartphones to quick and accurately recognize images, audio and video.

This means that you would have applications for speech, text, image, video and other recognition sources running offline on devices from mobiles to cameras to medical devices and cars.

One example of this could be the advancement of street view detection technology.

Consider for a second getting lost and only needing to snap a photo of where you are to have your mobile phone give you directions home.

3. Improvements in Unsupervised Learning

The intention of deep learning is to create machines that can learn on their own.

The most powerful recent example of this is a recent Deep Learning Algorithm. As error-correcting algorithms continue to advance, we can expect AI to become more and more competent when it comes to improving their own systems.

This increase in complexity, through improved network architectures, will have an impact on various domains from data-mining to robotics. As more and more data is able to be labelled you will also see wider applications and new opportunities.

A problem that you face now is having more data than you know what to do with, but deep learning will hopefully assist you in best filtering, sorting and utilizing this data.

4. Advanced Video and Text Recognition

As Google Brain consultant and noted computer scientist Geoffrey Hinton explained:

“I think that the most exciting areas over the next five years will be really understanding videos and text. I will be disappointed if in five years’ time we do not have something that can watch a YouTube video and tell a story about what happened.”

Think facial recognition software but for videos. In fact Facebook is reaching this point as it can recognise the variance between number of different sports, tagging and categorizing them.

But the implications reach further than social media. Imagine for example a text and video recognition software that can read any words (on a video) aloud for a blind user and help them navigate the web, or instantly translate street signs in videos into a different language.

While huge companies such as Google, Apple, Microsoft and Facebook will compete to dominate the machine learning market, smaller organizations can benefit greatly from staying up to date on industry trends.

Deep learning and AI in general is shaping the world and CIOs who can accumulate knowledge in this relatively novel field will be in a position to usher their organizations into this new age with an undeniable competitive advantage.

AI, Machine Learning, and Deep Learning need to be tracked closely by you so that you can start to understand where these technologies can be applied as a competitive advantage for you.

Reduce the Ceiling of IT Security Complexity with the CIO Security Scoreboard. Learn more here