Get Started
Latest From Sovren

The Science (and Art) of Artificial Intelligence


Artificial Intelligence. Machine learning. Neural networks. All terms industry professionals pitch around like buzzword baseball. But what does AI really mean? Why does it matter? And, equally important, is all AI software the same? The answer might surprise you — and is essential to understanding how this technological process can boost efficiency in your business.

AI and machine learning: What’s the difference?

Artificial Intelligence (or AI) is everywhere these days. It suggests what music you should listen to, what items you should put in your digital shopping cart, what temperature your house should be — and even what movie you’ll watch tonight.

Forbes describes this technology as “machines being able to carry out tasks in a way that we would consider ‘smart’” — that is, mirroring human decision-making.

Machine learning is an application of AI. It’s the idea of defining and refining thought patterns (algorithms) by analyzing and studying data. In other words, we teach machines to think, give them access to data, then let them learn on their own.

Machine learning by itself is not sufficient. Machines make mistakes, and to be usefully accurate, machine learning needs to be human proctored and iterative. That means that human domain experts must examine the results of machine learning and refine the algorithms to increase accuracy.

A perfect illustration of why machine learning requires human proctoring and iterative refinement is what happened early on with a popular photo app that used AI to identify faces. That app misidentified people as apes, and after justifiable outcry, humans had to intervene to retrain and refine the algorithms.

Perhaps the best way to understand AI is to think of an AI system as a newborn baby that must be fed information and instructions, and must be encouraged and corrected, in order to become a productive adult. Kids raised without formal education and attentive parenting can go awry. It’s the same with AI systems.

One step further: The rise of neural networks

Neural networks apply the concept of machine learning to computer systems — again, taking a cue from the human brain and functioning on probabilities.

How it works:

  • Data is fed into a system.
  • Statements, decisions or predictions are made.
  • Feedback is given (i.e., “this statement is correct” or “this statement is incorrect”).
  • The approach is modified for the future.

The key is iterative adaptability. As the feedback loop cycles on, the system’s criteria automatically adjust so that it can provide the right results for any given input — quickly learning expectations and improving accuracy exponentially. But this process isn’t easy, and requires elements not found in any software.

Test and repeat. Learn and synthesize. Artificial intelligence is a science and an art, but for AI software to become truly powerful — and truly valuable — it also takes these critical ingredients: human intelligence, data and time.

The Sovren advantage

For the last 21 years, Sovren has mastered AI matching technology for the recruitment industry, testing and refining our algorithms against massive amounts of data. We’ve taught our engine to view candidates as more than just a string of keywords. It ranks and sorts candidates according to criteria recruiters define, recognizes matches based on strength of relevant experience and generates helpful insights no other engine can.

At Sovren, we used AI to build semantic engines – engines that don’t just understand and use relations and statistics, but that understand what things mean and what values they have. This is a far more powerful approach than AI alone.

We use humans, not machines, to confirm if the outputs are correct, then we refine the code, iterate and test again. This continual refinement, experience and sheer access to data has put us ahead of the curve and positioned us as a leader in our industry. By leveraging our tools, you can reach far better results with far less resources.

Without a doubt, AI capabilities will continue to advance. So with our legacy firmly established, we will continue to iterate and expand our AI technology, supporting the recruitment industry for years to come.

See for yourself

Learn more about the features that set our engine apart — and how they can make a difference in your recruitment strategy.

SOURCES

View All Articles