AI in the Workplace: Making the Already Broken Even Dumber
There was a time, not long ago, when artificial intelligence was sold as the great liberator of modern work. It would automate the drudgery, surface insights buried in data, free humans for creative and strategic thinking, and finally fix the chronic dysfunctions plaguing offices, factories, and service industries. Productivity would soar. Bureaucracy would shrink. The knowledge worker would be reborn smarter, faster, and less stressed.
Reality has delivered a more familiar story.
A recent analysis in The Australian captures the growing disillusionment: despite billions poured into AI tools, chatbots, automation platforms, and "intelligent" workflows, many workplaces feel no better, and in some cases, demonstrably worse. Far from solving broken systems, AI is often amplifying the underlying stupidity.
The pattern is now clear. Organisations with messy processes, poor management, misaligned incentives, and cultures of risk-aversion throw AI at the problem like a technological bandage. The result is rarely transformation. Instead, you get faster production of low-quality output, more sophisticated ways to avoid accountability, and new layers of digital theatre that mask the same old problems.
Email chains become endless AI-generated summaries. Reports that no one read are now longer and more polished. Customer service scripts grow more robotic while real problems remain unresolved. Middle managers, instead of using AI to make better decisions, use it to generate impressive-looking slides that obscure their lack of judgment. The machine learns the organisation's existing mediocrity and optimises for it, producing higher volumes of the same flawed thinking at greater speed.
This is not an AI failure so much as a human and institutional one. Artificial intelligence excels at pattern matching within defined parameters. What it cannot easily fix is unclear goals, misaligned incentives, risk-averse cultures that punish initiative, or the basic incompetence and politics that plague many large organisations. When you automate a dumb process, you simply get dumb results faster and at scale. When you ask an LLM to summarise nonsense, it returns elegant nonsense; garbage in; garbage out. When middle management uses AI to avoid hard calls, the organisation becomes simultaneously more verbose and less decisive.
The deeper issue is cultural. Too many workplaces were already suffering from "bs jobs," process obsession, and signalling over substance long before AI arrived. AI supercharges the signalling. Employees learn quickly how to prompt for the desired corporate tone, optimistic, compliant, data-heavy but insight-light. Meetings about AI initiatives become more common than actual improvements. Consultants and vendors thrive on the hype cycle, selling the next tool that promises to fix what the last one didn't.
None of this means AI has no value. In narrow, well-defined domains with strong data and clear objectives, medical imaging, logistics routing, certain coding tasks, it delivers real gains. The problem emerges when it is deployed as a general-purpose fix for broken human systems. Once again: garbage in, gospel out. Or worse: garbage in, beautifully formatted garbage out that senior leadership mistakes for progress.
The lesson is old but worth repeating in the age of silicon: technology amplifies what already exists. A disciplined, high-trust, high-judgment organisation can become significantly better with AI. A confused, low-accountability, politically driven one will simply become more efficiently confused. The latter description fits far too many modern workplaces, like universities.
We should therefore lower our expectations about AI magically reforming organisational life. The real bottlenecks are rarely technical. They are human: incentives, culture, leadership, and the uncomfortable willingness to make trade-offs and enforce standards. Until those are addressed, pouring more artificial intelligence on top of institutional stupidity will continue to produce exactly what we're seeing: workplaces that feel dumber, faster, in the race to the bottom.
