Four enterprise leaders, four labels: the "specialists aren't required" claim doesn't survive contact

Four named enterprise leaders told ZDNET that the tightly defined single-domain specialist is losing ground to people who can work with agents across several areas at once ^{[1][4][8][11]}. That claim lands directly in the hands of anyone rewriting job families this quarter, and it does not hold together as cleanly as the framing suggests.

Kathy Pham, vice president of AI at Workday, told ZDNET that professionals who can work with agents across multiple areas simultaneously will be crucial ^[1]. Her stated reason is capacity rather than headcount: "These are the gaps that we need to fill, and maybe we can then redirect our energy to higher-value areas, because we've now automated the less valuable parts of our working lives that took our time" ^[2]. ZDNET reports the leaders it spoke to converged on two characteristics: adaptability to AI-enabled change, and the confidence to make decisions agents cannot ^[3].

Chris Kairinos, senior director of global modern workplace technology, client services, and technology operations at A+E Global Media, put it as a substitution: expertise in one area is being replaced by the aptitude to work across many ^[4]. He also said mastery is no longer available. "I don't think you need to master technology, because I don't think there is a way to master it now" ^[5]. His analogy is a piano gaining thirty keys, three layers of keys and three extra pedals, on the argument that an instrument changing constantly cannot be mastered, only used adaptably ^[6].

Then the vocabulary breaks. David Minahan, director of digital, data, and technology at the UK charity Young Lives vs Cancer, told ZDNET that "generalists are too broad" and that specialists are not required in most businesses anymore, because AI puts knowledge to work differently ^[8]. His answer is a third category he calls versatilists, people expected to work across the tech stack and to sit in conversations with business stakeholders about which digital transformation projects are worth doing ^[9]. He frames the label as a career-mobility fix as much as a capability one: IT has a history of pigeonholing people, and the new language is meant to be a pathway out ^[10].

Ankur Anand, group CIO at the recruiter Harvey Nash, rejects the same word from the opposite direction. According to Anand, "every professional who's actually pulling ahead right now is an expert whose specialism has stretched to include directing, questioning, and correcting AI that works alongside them," which he distinguishes from becoming a jack of all trades ^[11]. His test is a depth test: whether your expertise is deep enough to know when an agent is making the wrong choices ^[12].

So three distinct labels are in play across four leaders in one article ^[13]. That is not a semantic quibble if you are writing a role description, because the words select different candidates. "Works across the stack" screens for range and will let through people who cannot audit an agent's output in any domain. Anand's version screens for one deep domain plus supervision skill, and will reject the person Kairinos describes as adaptable but not a master ^[4][11][12]. The one point of agreement is the failure mode: something has to catch the agent when it is wrong ^[3][12].

What to watch: whether these orgs publish anything about how breadth is assessed and levelled, rather than asserted. The source material contains no interview loops, no rubrics, no pay bands and no attrition or productivity numbers attached to any of the three labels. Until one of those appears, "specialists aren't required" is a hiring philosophy, not evidence.