'Work is going to look very different': The future of AI in the workplace - HR Leader

TEXT START: A recent discussion at Mallesons’ Digital Future Summit 2026 dispelled misconceptions about AI use and explored what the next decade of work might look like.

The Dissection

This is transition-management literature dressed as a forecast. It acknowledges rapid deployment, genuine savings, restructuring, and reduced human workload, then redirects the reader toward prompts, leadership, redesign, oversight, adaptability, and “human” strengths.

The article correctly identifies the visible mechanics: AI scales quickly, institutions lag, and inserting AI into one process creates bottlenecks elsewhere. But it treats the central question—whether humans remain economically necessary—as an organizational design problem rather than a structural rupture.

The Core Fallacy

The article mistakes complementarity for permanence.

Creativity, judgement, empathy, expertise, and social coordination may be valuable during the transition. Under the Discontinuity Thesis, that does not make them durable employment moats. If AI becomes cheaper, faster, and broadly superior across cognitive work, those traits can be simulated, embedded in systems, or performed by a much smaller supervisory class.

“Greater expertise is in greater demand” can coexist with a collapse in total demand for human labor. A few high-value operators do not preserve a mass employment system.

The article also treats uncertainty about the net job impact as evidence of genuine ambiguity. Under P1, P2, and P3, that uncertainty is largely deployment lag: institutional resistance delays the consequences but does not reverse them. The Industrial Revolution analogy is therefore incomplete. Earlier machines displaced tasks while leaving humans necessary to operate the economy and consume its output. AI can attack the cognitive coordination layer itself.

Hidden Assumptions

• Human oversight will remain cheaper and necessary at scale.
• Expertise will remain scarce after AI can reproduce and distribute it.
• Displaced workers can move laterally into expanding roles in sufficient numbers.
• Organizational redesign will create enough new human work to replace deleted work.
• AI failures are mainly prompt failures rather than limits of the system or failures of human oversight.
• Institutions can coordinate a fair distribution of AI-created gains.
• “Sharing the benefits” will occur through policy rather than ownership and power.
• Human social cooperation will retain economic value even after machines can coordinate much of the work.

None of these assumptions is demonstrated. They are the load-bearing beams of the article’s reassurance.

Social Function

Primary classification: transition management and ideological anesthetic, with a substantial partial-truth component.

The text is not crude copium. It admits that some work will disappear and that restructuring is real. That concession gives it credibility. But it converts a distributional crisis—who owns the systems that replace workers—into a procedural challenge: redesign processes, train people, improve prompts, and manage change.

That framing lets institutions appear responsible without confronting the likely result: productive participation becomes concentrated among Sovereigns and a narrow Servitor layer, while the majority are maintained through transfers, precarious niches, or administrative containment. “Support people through the changes” is an aspiration, not a mechanism.

The Verdict

The article accurately records the early symptoms: rapid diffusion, institutional lag, workflow redesign, and real labor savings. It fails at the terminal diagnosis.

Under DT mechanics, the stated trajectory leads from P1 to P2 to P3: cognitive automation dominates, human-only economic domains become difficult to preserve, and mass access to necessary labor collapses. Human qualities may remain culturally admired and selectively monetized, but admiration is not employment.

This is not a refutation of obsolescence. It is a polished memo explaining how management will administer the transition while postponing the ownership question. The workforce is not being gently upgraded. Its economic centrality is being dismantled, one “complementary” task at a time.