To say artificial intelligence has gotten a lot of buzz recently would be an understatement. Google and Facebook have embarked on major AI research and development undertakings. Investment has come from the likes of Goldman Sachs and some of Silicon Valley’s biggest venture capital firms. The media churns out article after article about the potential implications of “thinking” machines; many of these are peppered with quotes from tech’s biggest thought leaders. But what about the practical applications of this technology?
Most of us have heard by now that some forms of AI – such as recommendation engines and spam filters – have already entered our daily lives. The usefulness of these technologies is self evident. Not so when it comes to the more sophisticated AI applications that are set to revolutionize the way companies do business.
If decision makers are skeptical, they can be forgiven. The business press often disappoints by generating hype without offering concrete advice. The big data craze, for example, has lead to large-scale data collection by many organizations that have no idea what to do with it. The truth is, artificial intelligence is more than just a buzzword. Its potential benefits are so great that, in many fields, embracing it is about to become a necessity.
The Data Deluge
From emails and social media feeds to the over 1 billion websites on the internet, information has never been more plentiful – or more overwhelming. For organizations around the world, this deluge of data is hindering productivity and resulting in major missed opportunities. Some of the biggest losses occur when big companies and corporations fail to make the most of their data.
The introduction to this post touched on big data, a concept that isn’t always thought through by those who try to leverage it. Big data can be very useful when it’s analyzed carefully, with specific organizational goals in mind. Of course, human beings aren’t capable of processing the very large data sets that constitute big data, a task that requires significant computing power. But how does a computer perform careful analysis? How does a computer “understand” the many factors that go into achieving a complex goal? These are the kinds of problems data scientists have been struggling with.
Of course, not every organization is equipped to perform these types of analyses; but companies of all sizes are struggling to manage information on a more basic level. As a concept and a frequently-experienced phenomenon, information overload has been around for awhile. It really entered the public consciousness in 2011, when Jonathan Spira’s book “Overload! How Too Much Information is Hazardous to your Organization” was published.
In one of the book’s most staggering statistics, Spira claims that over 28 billion hours are lost to information overload in the U.S. each year. Some of the biggest time sinks relate to “filing, deleting and sorting information”. Analyzing business-critical intelligence can be even more time intensive. The task of extracting truly relevant information – especially from the unfathomable depths of the world wide web – is the definition of overwhelming.
Whether a company is struggling with big data, information overload, or both, AI-enabled platforms and tools may offer whole or partial solutions.
Artificial Intelligence: a Practical Solution?
Recent events at Gnowit provide a good example of the difference artificial intelligence can make in processing large volumes of data. For those who don’t know, the company performs automatic monitoring of sources relevant to the people and organizations that use it. Formerly, these sources were curated manually. But with the volume of information disseminated through online channels expanding rapidly, manual source addition wasn’t cutting it anymore. All over the world, companies that deal with data are facing similar challenges.
The solution was a web crawler capable of not only adding sources to Gnowit’s system, but deciphering between relevant and irrelevant pages within these sources. This crawler illustrates the often perfect marriage between automation (which brings speed into the data-processing equation) and artificial intelligence (which ensures that, in addition to being collected rapidly, results are accurate). Specifically, the type of AI used was machine learning. Consider this from a business perspective. Gnowit saw a massive leap in productivity when this technology was implemented. In less than a week, the crawler was able to process a volume of information that, formerly, a team of three processed in 183 days. Not only that, but we saw a 20% reduction in errors.
I use this example not to brag about the work that the company is doing (or at least, that’s not the only reason) but because it clearly illustrates how artificial intelligence can be used to solve complex business problems, even in small and medium-sized companies. Not every company is Google or Facebook, and AI is used to build a lot more than robots that can carry on a conversational. It’s capable of extracting useful insights in a fraction of the time it would take a team of human analysts to do so. Given that, in the world’s history, there’s never been so much information, this capacity is very significant.
And what about information overload, that affliction that corrodes organizational efficiency and (according to a slew of studies) makes individual employees unproductive and unhappy? Artificial intelligence can help there, too. This post already mentioned the role AI plays in sophisticated spam filtering. Any platform or tool that uses filtering capabilities – or those related to sorting, synthesizing, and analyzing large quantities of data – can make analysts more productive. But without the accuracy provided by sophisticated, AI-enabled tools, these functions can lead to unreliable results (and, consequently, missed opportunities).
Making the Most of AI
Media monitoring, social media monitoring, web intelligence, business intelligence– ultimately, all of these solutions provide companies with the same benefit: the power to make better decisions. More and more, the information leading to these decisions is powered by artificial intelligence. SaaS innovators have been quick to fill the demand for these technologies. Faced with so many choices, how can organizations ensure they choose the right tools?
There are a few basic considerations. This may seem like an obvious point, but it bears repeating: understand why you’re seeking insight in the first place. Is it for competitive intelligence, issues management, or lead generation? Do you want to better serve customers by anticipating their behaviour? The answers to these questions will determine whether you should look for ways to analyze your own data sets, or collect information from online sources. Either way, it’s important for organizations to understand the scope of sources they need to monitor, and find a system capable of dealing with the appropriate data formats and volumes.
Of course, usability is important. Ultimately, an intuitive interface will always be beneficial. Accurate, computer-generated insights are impressive, but they should be accessible at the click of a button. Having to search a disorganized interface for the piece of information you need partially defeats the purpose of artificial intelligence.
This brings us to one final point. In business – and in other areas of life – AI helps us find the information we need, quickly. It brings accuracy to the superior processing speed of machines (and in some cases, it makes information easier to digest). In short, it’s a tool. Sure, it’s incredibly promising – revolutionary, in fact. But, like all sophisticated tools, its use requires the oversight of one or more people with vision.
Feature Image Courtesy: mark Miller