Big Data, Artificial Intelligence and Deep Learning: The perfect technology trident (Part 1 of 2)
Technology is a complex subject. It constantly changes, so much so that new products and services are always being developed. Every now and then technological advancements are discovered. Given the Internet’s supreme dominance over our daily lives, you’ll understand why technology can be truly powerful. Sometimes, with everything that it is, technology becomes too difficult to understand.
Take for instance three of the major technological innovations staring us in the face right now:
- big data – refers to complex or massive data that cannot be processed through traditional software and database
- Artificial Intelligence – more popularly known as AI, is all about making computers act or behave like humans
- deep learning – is a technology that simulates the brain and trains machines to recognize, classify and categorize data patterns
But the question is – Do we really understand these technologies?
Big Data and Artificial Intelligence
Artificial Intelligence has been around for years. A good example of AI in its early days are computer programs that were intended to perform tasks that solved problems humans found quite easy to handle, such as how to recognize an object contained in an image. These were followed by computer programs that were described as expert systems since they helped solve problems through an “expert” that asked questions. This is what classic AI is all about.
Today, Artificial Intelligence, or machine learning, has developed by leaps and bounds. One of the most popular examples is Siri. To get Siri to talk to you, all you need to do is ask a question. Whatever your question may be, Siri comes back with an answer. Even a simple “Hi, Siri.” will get a “Hello!”. Siri will even call you by your first name!
Another good example is Google’s “personal assistant” called Google Now. To use it, just install the Google Now Launcher and, on your home screen, say the words “Ok Google”. As it is a personal assistant, Google Now will remind you of events and appointments, give updates about your favorite websites, help you plan your travels and assist you with your commute time. The most interesting feature of Google Now is its voice command, which will allow you to ask Google practically anything.
This is the kind of development that machine learning or AI brings to big data. Let’s examine this by looking at a company using big data.
For example, your company sells different food products. With big data, you are able to collect information from your customers. This information is what you will use to understand your customers and find a way to give them what they want and need. Eventually, you will realize that something is missing; something that will drive you to more growth and help you go up the competition ladder. You cannot rely on just the customer data that you are getting, you also need intelligence. The kind that will help you understand complex matters and perform analysis faster than any human being can. This is where Artificial Intelligence comes in.
The idea is for AI to make machines think like humans, or even better. Primarily, the intention is to make human tasks easier and more efficient.
Deep Learning: The road to unsupervised data
Artificial Intelligence is related to deep learning, in that AI can provide deep learning. Once data is given to the machine, it will find a way to do what needs to be done, so we humans do not need to tell them what to do anymore. This is what is known as unsupervised data.
So far, the biggest advances in deep learning, and even in big data, are speech and image recognition. You can dictate into your phone (Siri and Google Now) and detect faces in photos (Facebook).
What Should We Be Excited About?
Despite Siri, Google Now and all the other AI and big data advantages we are currently enjoying, a lot still needs to be done to make Artificial Intelligence really take off. According to Geometric Intelligence CEO & founder and professor of psychology Gary Marcus, while there is a lot of progress and interest in AI, it is not really headed towards the right direction. Additionally, everybody seems to be excited about deep learning and big data. But, the tiny advances we have experienced are not really bringing us closer to the ultimate AI goal.
Artificial Intelligence, along with deep learning and big data, has the potential to instill change and drive inventions. This will need machines that can go beyond the data. Yes, we have robots that can perform tasks at home (and in the office). Our technology has also allowed us to power driverless cars. But these are not enough. For Marcus, there is still an element lacking. A home robot, for example, can only do so much. It can clean the house, wash the dishes and even take care of the kids. But, it can also put your dog in the trash or the dishwasher. The point is, there is a limit to the data it can process. For AI to work to its full capacity, it has to go beyond the norm.
According to Marcus, the ultimate Artificial Intelligence is the one that combines some of the best of what people do with some of the best of what machines can do. To be able to do this, we need to go back to psychology and study human beings. If we do so, we will find a better understanding of humans and use this to strategize on how to come up with the ultimate AI.
Moreover, Marcus believes that Artificial Intelligence is capable of radically changing the fields of science, medicine and technology.
Even if there’s a lot that still needs to be done, Artificial Intelligence and big data can still create a big impact, especially with deep learning tagging along. Already, it has made a splash in stock market trading by synthesizing data and increasing speed. In time, it will transform an even bigger industry and, perhaps, a whole community.
Stay tuned for the second part of this article.
Photo courtesy of Kārlis Dambrāns.