The Ice Cutters and the Executive Floor

In 2019, during my Artificial Intelligence class at Smith School of Business at Queen's University , Jonathan Aikman opened the first lecture with a black-and-white photograph of ice cutters in Quebec. Men stood on the frozen St. Lawrence River, harvesting blocks of ice for household and industrial use. Before mechanical refrigeration, this was skilled labour. It had tools, timing, discipline, logistics, and danger. Ice moved from river to warehouse to kitchen. A whole chain of work existed around winter.

Cutting ice along the st. Lawrence River near Montreal, 1869, photograph by Alexander Henderson

Then refrigeration arrived.

The photograph stayed with me because it avoided melodrama. No villain appeared in the frame. No foolishness, no laziness, no failure of character. The men were competent. The trade was real. The need for cooling remained. The economic arrangement around cooling changed.

That image returns whenever I read another announcement about artificial intelligence, automation, and job cuts. Standard Chartered's reported plan to cut approximately 7,800 back-office roles by 2030, more than 15 percent of its global support functions, sits inside that longer history of labour reorganization. The affected areas reportedly include HR, risk, and compliance, and the language used by CEO Bill Winters about replacing “lower-value human capital” with financial and investment capital has understandably drawn attention. The bank also set targets to increase income per employee and improve returns. 

Standard Chartered was once part of my own professional map, so I read the announcement with more recognition than surprise. Large banks do not move like sentimental organisms. They move through targets, ratios, operating models, investor expectations, and carefully chosen language.

Investors received the announcement favourably, according to the same reports.

The market heard productivity. Thousands of households heard a different sentence.

Is there any surprise, really? Banking has automated for decades. Manufacturing moved through the same pattern with robotics, sensors, and software. Agriculture did it earlier. Logistics did it brutally and efficiently. Anyone who enjoys modern abundance lives inside the consequences of past disruption. Our food, medicine, transportation, communication, and financial systems all depend on forms of automation that once displaced someone’s craft, desk, counter, route, or workshop.

The easy thought piece would rage against AI eating jobs. That version is emotionally satisfying for about eight minutes. It also leaves the harder questions untouched.

A more honest discourse begins by accepting the historical pattern. Capitalism reorganizes labour. Technology changes the price of coordination. Work once considered permanent becomes temporary. Tasks once requiring teams become buttons. Skills once carrying status become background plumbing. This is harsh, and also ordinary. 

Yet the ordinary nature of disruption gives leaders no moral exemption. A predictable event can still be badly handled. A financially rational decision can still be socially clumsy. A market-approved announcement can still sound cold enough to make the old ice trade feel warm.

The phrase “lower-value human capital” deserves attention because it reveals the compression of human contribution into an accounting category. Of course, executives speak to investors in the language of capital allocation. They are paid, in part, to translate complexity into numbers. Still, language does work. It prepares the mind for certain actions. It makes some choices feel clean. It turns people into a line item before the spreadsheet has finished loading.

Human beings who work in HR, risk, compliance, finance operations, reporting, reconciliation, onboarding, controls, and administration usually know something that the organization only notices after they leave. They know where the process bends. They know which policy looks elegant in a manual and fails on a Tuesday afternoon. They know which senior person needs three reminders, which system produces false confidence, and which regulatory process carries hidden fragility. Much of that knowledge lacks glamour. A great deal of institutional competence lacks glamour. The absence of glamour should never be confused with the absence of value.

Support functions became the first large target because they are legible. Their tasks can be mapped. Their workflows can be timed. Their cost can be counted. Their prestige is lower than their operational importance. In many organizations, that combination creates vulnerability. A role becomes easier to remove when its contribution appears as friction rather than judgment.

This is where the executive suite should pause.

The first wave of AI restructuring may reach analysts, administrators, operations teams, HR functions, and compliance desks. The next wave will climb. Executive work carries a thicker layer of symbolism, status, and boardroom ritual, yet many executive tasks also involve synthesis, prioritization, resource allocation, performance interpretation, risk framing, and communication. AI already performs large portions of those activities at growing speed and falling cost.

Modern executive life contains a surprising amount of structured repetition. Quarterly packs. Forecast revisions. Talent grids. Risk heat maps. Competitive scans. Town hall scripts. Strategy refreshes. Committee papers. Board pre-reads. Investor narratives. Transformation updates. The titles are grand; the mechanics are often procedural. A capable AI system connected to clean internal data can produce a first draft of many executive artefacts faster than a small army of highly caffeinated humans.

That observation should produce humility at the top. The support function employee may be the ice cutter today. The divisional executive may be the ice warehouse owner tomorrow.

The C-suite has long justified itself through judgment under uncertainty. That justification remains valid where judgment truly exists. Real leadership involves moral responsibility, courage, timing, trust, and the capacity to hold competing pressures in full view. It also involves the willingness to absorb consequences rather than outsource them to a dashboard. AI can model options. It can summarize scenarios. It can detect patterns. It can expose inconsistencies. It cannot carry moral agency on behalf of the institution. At least, the law, the regulator, the employee, and the public will continue to ask a human name to appear on the door.

This distinction gives executives a serious obligation. They should use AI to sharpen judgment, reduce waste, and remove unnecessary drudgery. They should resist the temptation to describe all displacement as progress and all savings as wisdom. The presence of a new tool does not automatically make every cut intelligent. A chainsaw can prune a tree and ruin it, depending on the hand using it.

An AI transition led well would begin with a different discipline. Before announcing headcount reduction, leaders would explain which work is disappearing, which work is changing, which skills are being rebuilt, and which savings will be reinvested in people rather than absorbed entirely into return targets. They would treat redeployment as a design requirement rather than a paragraph in a press release. They would publish internal principles and then allow employees to test decisions against them. They would train managers to speak plainly, because vague reassurance becomes cruelty when everyone already sees the spreadsheet.

They would also examine executive cost with the same seriousness applied to support functions. If AI creates productivity gains across the organization, senior layers should demonstrate restraint before asking lower layers to absorb the first pain. That could mean slower growth in executive compensation, smaller leadership teams, fewer ceremonial committees, tighter spans of accountability, and a clearer distinction between genuine decision authority and expensive coordination theatre. Markets may eventually reward that too. Investors have a long history of discovering courage right after someone else proves it is safe.

Boards should ask sharper questions as well. Which executive roles exist because the work requires human judgment? Which roles exist because old complexity created old hierarchy? Which committees survive through habit? Which leadership structures remain from a time when information travelled slowly? 

There is a strange asymmetry in many AI conversations. Junior and mid-level workers are told to reskill, adapt, and stay relevant. Executives are told to sponsor transformation. That language gives one group homework and the other group applause. The same standard should apply across the building. Every role now requires a defensible answer to a blunt question: what uniquely human value does this position add when machines can process the information faster?

Some executives will answer that question well. They will become better leaders because AI will remove some of the noise around them. They will see more clearly, listen more carefully, and spend more time on decisions requiring judgment rather than performance. Others will discover that their authority rested heavily on information asymmetry. The old executive advantage often came from access: access to data, access to meetings, access to interpretation, access to the room where meaning was assigned. AI weakens that advantage. When more people can ask better questions of the same data, hierarchy loses some of its mystique.

This could become healthy. Organizations have carried too much theatre for too long. Many employees already know which meetings exist to create the appearance of control. Many managers already know which dashboards decorate uncertainty rather than clarify it. Many executives, privately, know it too. AI may force a cleaner conversation about work, contribution, and authority. The danger lies in applying that conversation downward only.

For the newly unemployed, the situation demands sobriety rather than slogans. Being displaced by technology is painful, even when history can explain it. A role can vanish while the person remains capable, intelligent, and serious. The ice cutter’s skill did not become stupidity because a compressor arrived.

The practical response begins with translation. Workers coming from HR, risk, compliance, operations, finance, and back-office functions often possess deep process knowledge. That knowledge should be converted into AI-adjacent capability. Learn the tools, yes, but learn them through the problems you already understand. Become the person who can say, “This output is plausible and wrong.” Become the person who knows where automation will fail, which controls must remain human, where data lineage affects accountability, where a workflow hides regulatory exposure, where an employee experience cannot be reduced to a chatbot script.

The next labour market will reward people who combine domain fluency with machine supervision. The phrase sounds dry yet still true. The opportunity is real. A compliance professional who understands AI validation, model governance, audit trails, and regulatory accountability becomes more valuable. An HR professional who understands workforce analytics, AI-assisted case management, employee relations, and the ethics of automated decision-making becomes more credible. An operations professional who can redesign processes around human-machine collaboration becomes harder to dismiss as overhead.

There is also a psychological task. People must separate identity from title faster than corporations deserve. Many institutions ask for loyalty and return indifference during restructuring. This creates anger, and anger can become useful if it hardens into discipline rather than bitterness. The question becomes: which part of my knowledge survives the disappearance of this role? Which judgment did I build that a job description never fully captured? Which problems can I now solve in a new language?

The ice cutters did not control refrigeration. They also did not control how later generations would read their photograph. I read it as a warning against arrogance. Every economic arrangement feels natural to those standing inside it. Every status system presents itself as permanent. Every generation assumes some forms of work sit above technological weather.

The executive floor should avoid that comfort. AI will challenge support functions first because the business case is easier to write. It will challenge executive structures next because the same logic will keep moving. Once an institution accepts that machines can perform expensive cognitive work at scale, the question will travel upward through the org chart with quiet efficiency. It will pass through HR, risk, compliance, finance, strategy, transformation, and eventually reach the rooms where the word “human capital” is spoken with professional calm.

A humane AI transition requires more than reskilling budgets and investor-day slides. It requires leaders who can admit that efficiency gains carry human cost, that language can wound, that dignity deserves protection even during restructuring, and that executive privilege offers no permanent shelter from the machine logic now being applied elsewhere.

The ice cutters were skilled men standing inside a system about to change.

So are many others. So, perhaps, are the people making the announcements.