Artificial Intelligence (AI) is all the rage
in tech circles these days. In fact, in its “Top 10 Strategic Technology Trends for 2018” report, the global IT research and
analysis firm Gartner revealed that calls from clients regarding AI-related
topics increased more than 500% during the last year.
When tech topics become this hot,
distinctions between terms and jargon can become blurred. So, to clarify for
our readers, we offered these simple definitions basic AI elements in one of our previous posts:
- Machine Learning -- Software applications
that become more accurate in predicting outcomes by using algorithms for
- Chatbots – Computer programs that can respond to texts or chat
messages from people, emulating the conversational language of humans.
- Mobile Messaging – Related to Mobile
Instant Messaging (MIM), an app that applies AI such as machine learning and/or
chatbots to communicate between computers and people through mobile devices.
Each of the technologies has multiple
nuances, but machine learning especially seems to inspire confusion and
fallacies. Perhaps because it’s become a sort of vanguard of AI, introducing
businesses to the realm of AI. As CIO
contributor Mary Branscombe writes in a recent column: “Machine learning is proving so useful
that it's tempting to assume it can solve every problem and applies to every
So, we reviewed Branscombe’s article and
cherry-picked items from her list of “machine learning myths.” Here are three
points we believe are particularly pertinent to business leaders sizing up AI
for their organizations:
- All AI is not machine learning
Beware any discussion that
uses the terms AI and machine learning interchangeably. AI is involved in areas
such as robotics and natural language processing, for example. But neither of those fields
must necessarily involve machine learning. Branscombe’s advice for
understanding the role machine learning plays is “think of it as anything that
makes machines seem smart.” Machine learning essentially is about learning
patterns and predicting outcomes. “The results might look ‘intelligent’,” she
explains. “But at heart it’s about applying statistics at unprecedented speed
- Machine learning is objective
The old data-processing
adage “garbage in, garbage out” applies to machine learning. If data sets include
some sort of bias, machine learning will replicate the bias or even amplify it.
Branscombe offers the example of a machine searching for photos of CEOs on the
internet. The machine likely will return with pictures of white males because
today most CEOs are indeed white males. But all
CEOs are not white males. So, these results reflect a pattern but are not a
complete, accurate or objective depiction of reality. As Branscombe explains,
data sets must be “trained” to be useful in machine learning applications.
- Machine learning will eliminate jobs
Machine learning systems
enhance efficiency in repetitive tasks, improve compliance with complex regulations and reduce costs in the process. So, yes, some current jobs may become
obsolete. But like previous generations of automation, machine learning also
will create new business opportunities in areas such as customer service and
product innovation, which means new positions will be created, too. In fact,
some analysts believe machine learning – and AI in general – will generate a
net gain in employment over the long run.
Whatever form machine learning may come to
your company, the support of an IT Managed Services Provider (MSP) can help integrate this
emerging tech into your everyday operations.