AI in business: 'How can we do things completely differently now?' - RNZ

TEXT START: Lack of skill and fear of job losses are slowing down businesses using artificial intelligence (AI) to transform their operations, a technology consultancy says.

The Dissection

This is a snapshot of corporate hesitation at the threshold: 91 percent of surveyed organisations use AI somewhere, but only four percent use it to transform core operations. Firms are attaching AI to existing workflows while protecting the labour architecture that AI threatens. “How can we do things completely differently now?” is the question the survey shows they are avoiding.

The Core Fallacy

The article treats transformation failure as a skills and governance problem: train workers, appoint AI officers, and create standards. Under the Discontinuity Thesis, those are lag defenses. The central issue is not whether employees can use AI, but whether their work remains economically necessary once AI can perform it more cheaply and efficiently. Fear of job losses is not irrational friction. It is recognition of the actual mechanism.

Hidden Assumptions

• AI will primarily augment existing employees rather than eliminate economically unnecessary roles.
• Reskilling can absorb displaced workers at sufficient speed and scale.
• Regulation can preserve stable human economic participation.
• Organisations can choose the pace of transformation instead of being forced by competitive cost pressure.
• New AI leadership roles will distribute power rather than concentrate control of AI capital.
• Better adoption will benefit workers rather than mainly increase output while reducing labour demand.

Social Function

Transition management and ideological anesthetic, with partial truth. The article accurately records adoption inertia, skill shortages, fear, and regulatory pressure. But it packages a structural conflict as an implementation problem. Management can discuss “people and processes” without confronting ownership, control, or who receives the productivity gains. Incipient labour displacement is converted into a request for better change management.

The Verdict

Under DT logic, this is evidence of a lag defense, not a counterexample. The four-percent figure shows that firms have barely begun transforming their core machinery; it does not establish a limit on what they will eventually do. The 91-percent adoption figure is economically misleading: widespread use of AI features can coexist with preservation of the old employment shell. If AI achieves durable cost and performance superiority, competitive pressure will turn the gap into a race. Skills programmes, AI officers, and standards may manage the transition, but they cannot preserve mass productive participation when core workflows no longer require most human labour. This article is a report from the waiting room, not evidence that the patient is recovering.