The way Amazon is revolutionising how it recruits for large-scale jobs and automates its workforce is through a new breed of artificial intelligence tools that will make machines seem more human-like and eliminate some traditional processes. Speaking at a recent Amazon Web Services cloud computing conference, the company unveiled its most recent attempts at what is increasingly referred to as agentic AI, a type of sophisticated software that is able to operate independently with little to no human intervention.
The development signals the latest trend among tech giants in the field of automation. Rather than create machines that simply do the job, Amazon is trying to create AI that is seamlessly integrated. This approach, called “humorphism”, seeks to make technology work with human processes, rather than the other way around. It is a slightly different reframing of artificial intelligence, particularly in an era of growing fear about the loss of jobs.
A particularly impressive tool that has been released is Connect Talent, a recruitment system that promises to dramatically streamline the hiring process. For an organisation like Amazon, which employs hundreds of thousands of seasonal employees for the holiday shopping season, the ability to hire quickly can have a major impact. The company hired around 250,000 seasonal workers last year, an indication of the scale and speed required in its recruitment process.
Connect Talent employs artificial intelligence to handle various aspects of the hiring process. The system can assess potential applicants, interview them and produce reports for human recruiters, without human intervention. This tool is a 24/7 hiring system, unlike the more traditional approach that is based on appointments and face-to-face meetings. The elimination of the one-on-one interview is a significant change, one that could alter expectations for both companies and applicants.
Amazon, however, has stressed that the system is transparent. Applicants will know they’re speaking to a machine rather than a person. Colleen Aubrey, senior vice president of applied AI solutions at Amazon Web Services, said the technology is still in development, especially in terms of making interactions with it seem natural. “The experience continues to get better and better each iteration we go through,” she said in a pre-event briefing. “There’s a little bit of art in making that voice interaction natural and human.”
That statement highlights a problem in artificial intelligence development. Although computers can analyse data at unprecedented rates, they are still far from being able to mimic human speech. Nuance, empathy, and spontaneity are hard to program and yet are crucial to communication. Amazon’s strategy suggests that it recognises this challenge and is seeking to overcome it.
The use of agentic AI also brings up issues of control and oversight. These AI systems make their own plans, decisions and actions, which can lead to greater efficiency but also potential problems. If not carefully controlled, automated systems could make mistakes on a large scale or operate in a “black box” manner. This balance between progress and accountability is a growing trend in the field of AI.
Amazon is not the only company taking this path. Other firms such as Alphabet and OpenAI are heavily focusing on the development of agent-based systems that can autonomously complete complex tasks. And Microsoft is forging ahead with partnerships and the development of its AI technologies, further adding to the competition. The speed of innovation in the field indicates that companies in many sectors are gearing up for a world where AI will be a key part of the decision-making and operational processes.
Amazon has plans for more than recruitment. It has also developed Connect Decisions, a supply chain management system. This platform is designed to crunch data and help with planning and purchasing, serving as a smart assistant for supply chain professionals. With its extensive network of fulfillment centers around the world, even slight gains in supply chain efficiency can have a substantial impact on cost reduction and speed of delivery.
Aubrey said the company’s own experience was instrumental in developing this product. Drawing on its experience in managing complex supply chain operations, Amazon has created software it hopes will help other companies with similar challenges. “How do we get these human behaviours of collaborating to be a product?” she said. “Which is what we are trying to do and hopefully you’ll see it.”
But there’s an inherent irony in these developments. Amazon’s AI ambition is positioned as a means to complement human effort, but the company has also downsized its corporate workforce in recent years, some of it due to automation. This is part of a broader trend in the tech industry, where new technologies can be both a boon and a curse.
On the face of it, the benefits of AI-powered recruitment are obvious. They can streamline hiring, remove scheduling constraints, and automate grading. For jobseekers, the flexibility of being able to interview at any time may be appealing, especially for those juggling multiple life commitments. However, the lack of human contact may also result in candidates feeling isolated or unsure about how they’re being evaluated.
And there’s the issue of bias. Although AI is heralded as more impartial than humans, it is only as good as its data. To promote fairness and inclusion, systems need to be monitored and adjusted, which Amazon seems to recognise as it continues to develop its system.
This evolution is not one of machines taking over, but rather one of partnership and co-evolution. Amazon’s philosophy of “humorphism” seems to be an effort to humanise technology at a time when it can become dehumanised. We shall see if that is possible.
