< back to news
June 6, 2018

Machine Learning & AI – Fundamental Understanding [1100]

Artificial Intelligence (AI)


You might already have seen this, but we have added a new functionality to our website – a friendly chatbot to improve your experience the second you land on our homepage. In honor of our marvelous, friendly chatbot, we have decided to trace the origins of the technology that’s made it possible to interact with sophisticated code.

Yes, we’ll be talking about Artificial intelligence (AI) and machine learning. Popular culture gives us a high version of AI that comes close to being indistinguishable from human beings, and while androids might be in our future… The present is rather more mundane.

AI stands behind self-driving cars, the connective functionality of The Internet of Things and even image processing algorithms.


A Point of Origin

Computer scientists and engineers have been testing the limits of programming and algorithms since the infancy of the modern computer. Artificial intelligence first appeared as a term in the mid-50s and its main purpose was to solve problems. It was only later, as hardware technology evolved by leaps and bounds in the 1960s, that AI steered into replicating basic human reasoning. DARPA (Defense Advanced Research Projects Agency) has pioneered the concept of the digital personal assistant before Apple and Microsoft turned the technology into a widely available commodity.


If we’re to look at the process in terms of historical stages, three distinct phases emerge:

  • The Inception (1950s – 1970s); first steps are taken towards the creation of AI and the first successes include the computers that can play chess.
  • Machine Learning (1980s – 2000s); computer technology gives rise to more sophisticated algorithms and machine learning (discussed later in the article) is on the rise.
  • Deep Learning (2000s – now); AI has become prevalent in most industries today thanks to advances made in the field of deep learning (the next progression in machine learning) and it’s the reason we have personal assistants, search engines and chatbots.


Where Would We Be Without AI?

One word – nowhere!

Modern industry (and life as we currently know it) relies on one key concept – automation. Early forms of automation concern the automation of manual tasks, or otherwise referred to as robotic automation. It’s this process that allows us to build cars, planes and basically anything in a matter of hours, rather than months.

AI compliments this robotic automation by automating data processes (discovery through data and repetitive learning), but whereas its predecessor is autonomous, AI still needs human guidance. As such, AI does not exist as a separate application and thus only improves on existing products and technology. It also explains why AI has spread widely through so many industries – the software adapts and can be applied to virtually any task.


But… Where is AI around Me?

Everywhere. Smart watches, fridges that notify you when you’ve run out of fresh milk, and voice-activated technology; these are all the accomplishments of AI.

Marketing is what we at Sales Promotions do best, but without AI we’d be helpless in positioning ourselves online to assist you in planning successful marketing promotions. It would also be next to impossible to process data on our own without AI’s speed and accuracy in analysis. AI gives deep insights into your interactions as a company with your customer base, and generates strategies for improvement. AI knows what products to suggest to customers based on their prior purchases, and AI is also behind massive search engines like Google and Bing.

Hospitals, universities, banks, financial organizations, social media – everything runs thanks to the inclusion of AI.


You might say AI is taking over the world… metaphorically speaking, of course. While alarmists view AI like Siri and Cortana as the first step towards robots’ global domination over humanity, it’s a lot more complicated than that. Remember when we said AI needs human input. That’s still very much valid no matter how sophisticated the processing capabilities of AI become.

Also fears of AI stems from a general lack of understanding of its basic nature.


You Have to Understand How AI Works

This is your warning that the subject matter is going to become a lot more technical.

There are three main components to AI, speaking in the roughest sense: (1) vast amounts of raw data, (2) complex algorithms, and (3) iterative processing at high speeds. Algorithms are trained to go through the raw data, searching for patterns or features based on the task at hand. However, take note that this is an oversimplified definition as artificial intelligence contains multitudes when it comes to technologies, methods and approaches. The field of study has become so large it has already spawned several subfields:


  • Machine learning – a process of analytical model building automation, which is the main foundation of AI. It’s through machine learning that AI identifies hidden insights in data with no prior guidelines or instructions programmed into the process.
  • Neural networks – builds on machine learning through the creation of a pseudo nervous system comprised of interconnected units that yields greater analytical results in locating connections within raw data.
  • Computer vision – its AI trained for pattern recognition in order to understand all matters of visual information in a manner similar to how humans interpret pictures and video.
  • Natural language processing (NLP) – this field preoccupies itself with the analysis and the understanding of human language. This is the technology that allows chatbots to function, as it not only allows the AI to comprehend what a person communicates through the use of human written language, but gives the AI the capability to respond back in kind. All set within strict parameters as chatbots have a limited dictionary and range of responses.


AI of the Future

Artificial intelligence maintains an upward trajectory as we build better, faster and more powerful computers that give us the processing power to increase the capacity of algorithms to go through bigger volumes of data. Maybe one day we will see an autonomous AI android that will mimic the human mind to a point of independent thought…

More realistically, the AI of the future is going to make lives all the more convenient whether you want to invest on the stock market or find the perfect image of candid apples to pin on a Pinterest board.

Now that we have covered the introduction of AI, we’re going to be moving forward with a new topic next week – conversational marketing, which is the marketing principle that has led to the creation of chatbots in the first place. Once we discuss how conversational marketing caused a veritable revolution, we’re going to conclude the series with an article illuminating the nature of chatbots. What they are, what they aren’t and what they could be.