TempRocket CTO Dr Fraser Greenroyd explains why people, not machines, must always have the final say in recruitment.
The topic of Artificial Intelligence (AI) comes up in almost every sector these days, from healthcare imaging to self-driving cars. AI is all around us when we use almost any form of technology in varying forms and disguises.
When we speak to Cortana, Siri, or Alexa, we’re using AI through voice recognition to allow the machines to understand what we’re asking them to do. However, there’s still a key player in these interactions, the person using the platform to interact with the machines is still the key decision maker and holds the final say in how these devices execute their instructions.
We’re frequently presented with a future where AI-driven robots run the world, doing all our work and making all our decisions for us. But this is not only a future that many of us do not want, it’s also actually highly impractical in many situations.
A sub-division of AI, sometimes referred to as Computational Intelligence (CI), can learn how to perform a task from a dataset or experimental observations, removing the need for human interaction in a machine’s decision-making capability. This could result in Alexa ordering milk deliveries based on the amount of milk remaining in the fridge, or tea bags based on the amount left in the packet. Why do we need humans to tell us whether they think there’s enough milk left in the fridge when our computers can do it for us?
Well, we probably don’t, but that’s not necessarily the case with every task. Take recruiting the right employee for your business, for instance. In the temp side, if we remove the human decision making element from the system, we could, in theory, speed the booking process up immensely.
If we don’t need the contractor to confirm they can undertake an assignment because we’ve linked a computer up to their calendar directly, then we could save precious minutes in the booking process. If we used CI to predict when a company needs temporary workers, then we could have them booked in a matter of seconds with automatic bookings made across the system. This would be a great, right? Well, except for the dystopian machines-running-the-world issue…
However, removing the human from the decision-making process is a mistake in systems where human interaction is key, and the recruitment industry is a case in point. If we remove just one element from the booking process the entire system will collapse.
We need the contractors to confirm their availability, the agencies to confirm their willingness to supply, and the hirers to confirm they need temporary workers. A contractor isn’t going to want to have their schedule and calendar dictated to them by a machine. A hirer isn’t going to want a machine second-guessing their staffing requirements. Thus, involving the decision makers is key, and that’s exactly what the new online temporary recruitment platforms achieve. The best ones combine inputs from agencies, hirers, and contractors with AI to provide a streamlined workflow saving users time and money. This optimised workflow allows a booking to occur in under a minute while continuing to engage with the key stakeholders.
This is the right level of AI to have in a system relying on organising human livelihoods. It is crucial to remember the difference between the machines helping our decision making (AI) and making decisions for us (CI) and recognising when it’s the right time to use one or the other. AI will not revolutionise the recruitment process because it’s dependent on subjective human input, but it can be used to aid our decision-making and assist in our booking processes like it does with the new temporary recruitment platforms.
Dr Fraser Greenroyd is CTO at TempRocket.
Originally published Net Imperative | January 2019