AI is not your employee—it's a performance-enhancing tool

Ruba Borno, Worldwide Vice President of Global Specialists and Partners at AWS, argues that the language used to describe AI in the workplace—calling AI a 'colleague,' or referring to agents as 'team members' and humans as 'managers'—has led businesses astray in their thinking about how to deploy the technology. While there is some value in borrowing management concepts from how people oversee autonomous systems, the metaphor fundamentally breaks down when applied to workforce decisions.

Borno proposes an alternative mental model by drawing on sports history. In 1961, the fiberglass pole replaced the metal one in pole vaulting, and athletes broke the world record 19 times in a single decade. In cycling, better aerodynamics alone accounted for nearly half of all performance improvement over the sport's history. Graphite rackets transformed tennis, and polyurethane swimsuits became so effective in swimming they were eventually banned (though some records from those two years still stand). None of these technologies competed with athletes; each was a performance-enhancing tool that allowed extraordinary humans to exceed what was previously possible. This, Borno argues, is the right mental model for AI.

She illustrates this with examples from her own work and beyond. Within her engineering organization, teams increased shipping throughput 1.65x year-over-year and improved time to production by 21% with more than 76% of code changes AI-assisted. Through Kiro, an AI coding assistant, every employee can now build workflows, sort data, and automate administrative work that used to consume their weeks. The World Surf League provides another case: for 50 years, performance data was buried in spreadsheets dating back to the 1970s, and commentators spent hours before each broadcast hunting through fragmented files for athlete stats and career histories. Working with AllCloud, the league unified all of it into a real-time AI platform, cutting broadcast prep time by 80% and freeing broadcasters to do what they do best—tell stories rather than dig through spreadsheets. United Airlines moved more than half a million passengers through its network each day and initially drafted delay messages by hand; today, AI drafts more than half of those delay messages, enabling a customer experience at a volume no team of any size could have delivered.

Borno concludes that in every case, the pattern is identical: AI automates the process, humans deliver the new value, and once they do, that standard becomes the new floor. She argues that too many companies are still treating AI as a hiring decision, asking 'how many people can we replace?' rather than the bigger question: what can our people accomplish now that couldn't before? The companies pulling ahead are not asking how to do the same work with fewer people; they are asking where the bar is today and how to set it higher than anyone else can reach. In this era, the standard is not 'keep up,' but 'raise the bar so high that keeping up is the other guy's problem.'