This is Why You Need Machine Learning / Artificial Intelligence
I find it interesting how some terms are so well recognized, yet so often misunderstood. Take Artificial Intelligence or “AI.” We all have a pretty good idea what AI is, and we certainly know what can go wrong if a computer program with Artificial Intelligence starts to think on its own. It could be the end of the world, as portrayed by Skynet in the Arnold Schwarzenegger movie “The Terminator.” But what is AI, and how does it relate to Machine Learning? Why is it such a hot term today?
What is Artificial Intelligence?
According to the Merriam-Webster dictionary, AI is defined as:
- A branch of computer science dealing with the simulation of intelligent behavior in computers
- The capability of a machine to imitate intelligent human behavior
This is a start, but it is a bit vague. What does it mean to “simulate” intelligent behavior in a computer? It seems like many things could be categorized as AI based on this definition. Looking to John McCarthy, based on a paper he wrote while at Stanford University in 2007, he defined Artificial Intelligence as:
- The science and engineering of making intelligent machines, especially intelligent computer programs
- He then defined “intelligence” as “the computational part of the ability to achieve goals in the world. Varying kinds and degrees of intelligence occur in people, many animals and some machines”
Based on his definition, Artificial Intelligence is “manufactured” intelligence accomplished with a computer to achieve a goal. AI is more than just a computer program that responds to inputs. It is a machine / program that can collect data, make inferences about that data, and then provide new intelligence that is “concluded” or “predicted” from vast amounts of data that collected over time.
From a business point of view, this is where things get interesting.
What is Machine Learning?
Arthur Samuel, an American pioneer in the field of computer gaming and artificial intelligence, coined the term "Machine Learning" in 1959 (source). He created a Checkers-playing computer program that appears to be the world's first self-learning program, as a pioneer in the world of Artificial Intelligence (AI). His definition of Machine Learning was:
- A field of study that gives computers the ability to learn without being specifically programmed
I like this definition. It is easy to understand. I also like the example. Checkers seems like a strategy that could be programmed (at least back in 1959) without having to try and collect a summary of every conceivable move that could occur in a game of Checkers, to then provide a corresponding next move, which would certainly be a lot of computer modeling.
But, doesn’t this definition sound a lot like that of Artificial Intelligence? The answer is “yes.” It turns out that Machine Learning (ML) is almost the same as AI, with the only difference that it is simply not as well known a term, and perhaps not quite as wide a scope. Thank Hollywood for this misinformation. Technically, if you’re a data scientist, you’d define AI as being a broader field, with Machine Learning a subset of AI.
For this discussion, let’s acknowledge that ML = AI.
Why Should I Care About Machine Learning?
Machine Learning is becoming an increasingly sought-after analytics capability for one simple reason: the explosion of “Big Data.” As part of living in the digital age, we have witnessed an incredible surge in volume of data being collected, stored and transmitted. Big data is a term that was created to define data sets that are so large or complex that traditional data processing applications are inadequate to deal with them.
The below graph (source) helps to visualize what a geometric growth curve digitization has had on data generation – and this chart is a few years old. Amplify this concept across every business, and it is easy to see why there is a data revolution underway right now.
Further amplifying this growth has been the Internet of Things, which when coupled with data collection strategies and devices, has put incredible new volumes of data across every large enterprise, and most mid-sized and smaller organizations too.
This means it’s no longer possible to manually sort through this data to draw conclusions. A whole new analytics industry to process Big Data has been created to draw conclusions and help predict future events.
Here is where it gets interesting for business owners or department heads.
If future events can now be predicted with even a little better accuracy, or if new decision support intelligence can be extracted to help improve how a business is run, then serious business value has just been unlocked. This is exactly what is behind the current AI “boom” now underway, and why the AI market is estimated to be $12.5 billion by the end of 2017, according to market research firm IDC.
Machine Learning Use Case #1 – Data Security
Given the recent ransomware security attacks we have all read about this year, it should come as no surprise that Machine Learning is an important part of new data security solutions.
In 2014, Kaspersky Lab said it had detected 325,000 new malware files every day. But, institutional intelligence company Deep Instinct says that each piece of new malware tends to have almost the same code as previous versions — only between 2 and 10% of the files change from iteration to iteration. Their learning model has no problem with the 2–10% variations, and can predict which files are malware with great accuracy (source).
What this means is that some data security solutions can see common attributes of what code might be malicious, and then act accordingly to shut down the program – before that virus has even been recognized. This is a true “zero day” defense strategy, which in the world of data security, is a big deal.
Machine Learning Use Case #2 – Phone Log Monitoring
Have you ever wondered while sitting on hold or calling to place a complaint who listens to all the phone calls that begin the phrase “this phone call is being recorded for quality assurance and training purposes”? I can’t imagine how long it would take for some of the larger call centers to hire someone to listen to all the calls.
The global call center industry is estimated to reach $10 billion b7 2019, based on research conducted by Technavio. A typical call center might field 500 calls in a day. That is a lot of calls. Of course, no one has the job of listening to every call. Rather, the strategy is the same as video surveillance. If there is a future program, then the call log will be listened to better understand the situation and how to avoid it in the future.
But what about all the other intelligence that could be extracted from these calls if there was a way to selectively listen to just the right ones? New product or service innovation might be uncovered. Message adjustments might be revealed. And, sales strategies could be better executed with greater success.
Here is where Machine Learning comes into play. While an entire conversation can’t be heard, select keywords can be identified to express sentiment with what is being discussed during a phone conversation – be it with a customer service representative or a sales representative.
Once a set of keywords has been identified, new intelligence could soon follow. Machine Learning or AI can work on identifying what other word correlations exist when used in conjunction with the terms first identified. This behavior can then be modeled to predict success or failure of calls, based specific objectives associated with the parties on the call.
Prodoscore recently hosted an educational happy hour at Google's New York offices with Google, Salesforce, SuiteBriar and RingCentral on how a business can leverage Machine Learning, which is now an integrated part of Prodoscore’s latest offering (see announcement). New revelations can now be uncovered with greater ease and accuracy that has never been possible before. Now it is possible to understand when messaging is not being consistently used, or, if calls are being made, but the topic being discussed is not relevant.
Learn more by watching a recording of this webcast:
Or, watch this event slide deck.
Artificial Intelligence under the moniker of Machine Learning is quickly becoming a useful and highly valuable technology breakthrough. Significant value is being unlocked by businesses that sell to other businesses, and those that sell directly to the consumer.
A huge wealth is now being amassed by companies – big data – which is being collected as part of how the business is run. Those organizations that invest in new technologies with Machine Learning are now discovering the hidden treasures in their data and learning how to make more intelligent decisions.