Machine Learning Isn't Mechanical

Machine Learning Isn’t Mechanical

“Can’t we just buy one of those AI programs and turn it loose? It’s so smart, won’t it just figure things out?”

This was the giddily optimistic ― but wholly naïve — question posed recently by one of my clients, an otherwise savvy corporate executive of great experience and intelligence. The question vividly illustrates a disturbing reality: Although 85 percent of CEOs believe artificial intelligence will change the way they do business in the next five years (according to a recent survey conducted at the World Economic Forum in Davos), there’s still a whole lot of misinformation out there about what machine intelligence can and cannot do. (Not to mention ignorance about the investments a useful machine learning system requires.)

I blame a lot of these techno-myths on the depiction of “intelligent” computers in popular culture. From HAL 9000 in Stanley Kubrick’s 1968 classic film 2001: A Space Odyssey to the scheming videogame avatars in Disney’s 1982 Tron to the rebellious android “hosts” in HBO’s Westworld, AI in sci-fi is depicted as having the personality traits, emotions, self-awareness, and agency of flesh-and-blood human beings. Although the executives who are investing millions in machine learning undoubtedly know science fiction is, well, fiction ―and certainly don’t expect their über-algorithms to stage an armed uprising ― many still harbor the belief this technology will display at least some of the initiative, creativity, and even intuition of the employees it’s designed to augment (or replace).

Fact vs. Fantasy

The reality is not quite this utopian (or scary, depending upon your perspective). Machine learning is, without question, an impressive and extremely valuable technology. Companies that learn to harness its potential can most certainly seize the high ground in their battle with competitors. But machines are still machines, even the really intelligent ones. No matter how much information they can process in a second ― no matter how many insights they can glean from data that might otherwise look like random noise ― they’re still “toasters.” They have no initiative. They have no agency. They feel nothing. They want nothing. They understand nothing. They’re only going to do exactly what you tell them to do. Nothing less. Nothing more.

 

Not Plug & Play

Another popular misconception is that machine learning technology is some kind of self-contained “black box” you just plug in and put to work. (Ah, if only it were that simple…) In fact, machine learning is an entire ecosystem of technology and activities that can reach into every corner of a company’s operation. It requires use/case expertise — people who know exactly what questions need to be asked and the kinds of answers they need to receive. It also requires the accumulation and curation of large amounts of content so the algorithms can be “trained”. (A company may not have the legal rights to all the content required; this is where many attempts at machine learning fail.) It requires a huge computer processing infrastructure, including:

In short, “Plug & Play” this ain’t.

Bringing out the Big Guns

In spite of the above considerations, employing machine learning is still the business equivalent of “bringing out the big guns.” When advising clients, we always offer the following guidelines:

  1. Pick a problem that can be solved using existing technologies. While machine learning is no longer in its infancy, it’s most certainly still in its messy adolescent phase. It has its limits, so be realistic.
  2. Pick a problem that is important to your business. Getting your machine learning solution operational is going to require time, money, and attention from virtually every part of your organization so make sure the investment is going to be worth it.
  3. Pick a problem that can’t be solved any other way. If you can do it easier, faster, or cheaper using some other method or technology, by all means, do so. You don’t have to build a Saturn V rocket if all you want to do is go to the grocery store.
  4. Bring together all disciplines and domains. As stated earlier, implementing machine learning requires a group effort. This includes getting input from even the non-techies at your organization.

 

Learn More about Machine Learning

Want to know if machine learning is the right choice for your organization? Click here to get started. Or contact us to start a conversation.

Back To Top