At Scout our recruitment marketplace is driven by AI. We use it to match specialty recruiters with organizations that need specific recruiting skill sets at a specific moment in time. We use machine learning and AI to develop recruiter rankings that help employers identify best-fit recruiters. Machine learning also helps us to identify trends in the job market—whether it’s jobs with rapidly rising salaries or a sudden niche demand among employers. On the requisition front, we use AI to improve job classifications and predict needed skills.
Because we were early adopters of AI in recruiting we have a vantage point others don’t. We also know there’s a lot of chatter out there, so we thought we’d take this time to dispel the top five myths on the topic.
1. AI, machine learning and predictive analytics are the same thing
We put this first for a reason: There’s a lot of confusion out there about what AI is and isn’t, and related words are often used interchangeably. Artificial Intelligence, or AI, is the umbrella term for the scientific study of machines that perform tasks historically carried out by people. A lot of practices fall under the broader category of AI.
- Machine learning, for example, encompasses the algorithms and statistical models that allow a computer to learn from itself. Scout uses machine learning to assess recruiter effectiveness.
- Natural language processing is the field dedicated to the classification, translation and use of human language by computers. Chatbots are an everyday application of this. And yes, one call to the cable company’s automated operator demonstrates just how far we have to go with this particular technology.
- Predictive analytics is the practice of using historical data to predict future results. AI deployments often rely on predictive analytics to get the job done. For example, at Scout we look at a recruiter’s past performance in a specific area (say, job role, industry or level) to predict how successful they might be in that arena in the future. The more data we have, the more accurate the predictions.
Image recognition, machine vision and robotics also fall under the AI banner. So do any number of new technologies that are just starting to be imagined. If you’re interested in a good primer of the branches of AI today, take a look at the chart in this legal blog.
2. AI eliminates bias
AI offers great potential to reduce bias, but the algorithms are still written by people who need to be able to anticipate and control for bias. Unless you correct for bias, AI will never deliver on its potential. Consider this: Amazon famously built an automated resume review tool that preferred men. It was built on ten years of historical data during a period dominated by male hires. As the system learned to identify predictors of success, it taught itself to weed out women. Don’t buy it? Apparently, the tool downgraded resumes simply because they included the word “women.”
3. Everyone is doing it
A recent PwC survey found that only 27 percent of companies have implemented AI in multiple areas of their companies. So while it isn’t everywhere yet, 100% of the surveyed companies are somewhere on the continuum from considering to implementing AI. It simply can’t be ignored by forward-looking companies. Among AI adopters, 80 percent say their investments have resulted in a financial return.
4. AI replaces people (or at least it will)
AI is a way to determine and predict efficiencies and best outcomes. In the case of Scout it assigns the right jobs to the right recruiters but only they have the human relationships with candidates that determine the best fit with a hiring organization. Because AI is constantly learning, our recruitment marketplace is getting smarter with each job that is filled.
As for other fields, automation will certainly replace some jobs, but it’s likely to create as many jobs as it takes away. What’s more, we’re already seeing a significant demand for AI talent, a good indicator that this is one trend that won’t be going away.
5. AI is always right
Refer back to myth #1. AI is only as useful as the people who program it. And, while AI is getting more sophisticated by the day, there’s a long way to go yet. We can now access and process massive data sets, but machines can’t address multiple problems with limited data sets. For the time being, only people can do that. With major investments being made, it’s natural to want the machines to be right all the time, but we still have to rely on people to make the final decisions.
AI is only one tool—albeit a critical one—in an environment where keeping pace with the competition takes everything you’ve got. At Scout, our business literally depends on the quality of our AI deployments, but we never lose sight of the fact that it relies just as much on the quality of our recruiters and search firms, as well as the intelligence—artificial and otherwise—that we deliver to clients.