"Computer, define artificial intelligence"

Understanding artificial intelligence, machine learning and where it fits in HR

Last year, more than $5 billion in funding was invested in artificial intelligence (AI). It is undoubtedly one of the hottest topics across industries and functions today. Everyone from big pharma to consumer analytics is talking about what AI is, how they are using it, and what the implications might be.

In understanding this massive technology trend, the first step is to clarify terminology. There is an un-ending list of buzz words out there when it comes to digital technologies, and it is difficult to know when we are describing the same thing using different words, or different things using the same words. My favourite example right now is “artificial intelligence” and “machine learning”.

Artificial intelligence can be defined as “A branch of computer science dealing with the simulation of human behavior in computers” or "The capability of a machine to imitate human behavior". The way I think of this is that it is an outcome we are striving for: to have a machine (hardware or software) imitate humans.

Machine learning, on the other hand, is a little different. My favourite definition is "A computer program is said to learn from experience E with respect to some class of tasks T and performance measure P if its performance at tasks in T, as measured by P, improves with experience E." - clear, isn't it? Another definition, slightly more relatable, is "Machine Learning is a process that gives computers the ability to learn from data without being explicitly programmed".

The "without being explicitly programmed" part is key. A traditional algorithm is static. It produces a result to the same level of accuracy every time, until a human actor, making observations about the outcomes, intervenes and updates the code. Then the algorithm performs at the new level of accuracy until the next update, and so on.

In contrast, a machine learning algorithm includes a continuous feedback loop between the calculation, the result, and accuracy of outcome, which allows it to refine its calculation to improve accuracy each time. The algorithm can do this constantly, without human intervention. It literally learns the same way a human does - thus achieving the outcome of artificial intelligence.

This self-improving capability is the driving force behind much of the cool stuff we hear about in the media - suggesting the next Netflix episode, your next Amazon purchase, your next online learning course in your career development plan, as well as everything awesome about processing, understanding and reacting to unstructured content.

[Aside: structured content is clear numbers or words in a pre-defined format. Unstructured content is fuzzy content like pictures, video and natural (conversational) language]

So: artificial intelligence is the outcome. Machine learning is the engine that enables it. 

That's all very well and interesting. But what is the business value? How do we take this out of the realm of opening the pod bay doors, and into the reality of daily operation?

This has fascinating implications for all aspects of the organization, and I'm particularly interested in applications in HR. As one of the most human-focused and human-effort intensive departments, it's intriguing to think about how it will be transformed by digital technologies like artificial intelligence.