My reflections from a Lightcast-sponsored dinner discussion on the question: "How might we practically balance the strategic capability building with the near-term, tactical execution? How do you keep decisions aligned with the future impact of AI on your roles and workforce?"
There was a useful honesty in the discussion around this question. Most organisations know they need to think longer term about AI, capability building and workforce shape, but most are still governed by near-term pressures, quarterly expectations and leadership incentives that reward short-term delivery. That tension is not new. What is new is the speed at which AI is beginning to reshape the work itself. That means the old habit of treating strategy as a separate, slower conversation is becoming harder to defend. If roles, tasks and decisions are changing now, then the balance between tactical execution and strategic capability building cannot sit in two different rooms.
One of the strongest ideas raised was that the lever for longer-term thinking is not simply better planning. It is a different lens on value. If the organisation is driven purely by shareholder capital, it will usually default to the short term. If it takes stakeholder capital seriously, it has more permission to invest in capability, experimentation and workforce redesign that may not pay back instantly but matters strategically. That is an important challenge for HR and workforce leaders. Too often, they are asked to build future readiness inside systems that still incentivise immediate delivery above all else.
A practical way of framing the issue came through in three words: who, what and how. The ‘who’ matters first. Long-term workforce decisions need leadership that understands the work, not just the reporting line. That means HR leaders, people analytics leaders or workforce researchers who are credible in business terms and capable of translating change into operating decisions. It also means going high enough in the organisation. If the conversation never rises above business-unit firefighting, it is unlikely to become strategic. In practice, future-focused workforce work needs sponsorship from leaders who can look beyond this quarter without being accused of avoiding the real world.
The ‘what’ is where many organisations still struggle. There was a clear point made that not every people analytics initiative needs a perfect ROI calculation, but at some point someone needs to be able to show whether a strategic initiative is making a difference. That is a healthy test. If doubling the budget or halving the budget would make no visible difference to outcomes, then the organisation should ask harder questions about why it is doing the work at all. In an AI context, this matters even more. Capability building cannot just be a banner headline. It needs to connect to decisions on hiring, job design, learning investment, organisation structure and where human judgement still creates value.
The how may be the most practical point of all. One contributor described the need for a consistent funnel: an intake process, clarity on priorities, and discipline around what comes in and what drops out. That sounds operational, but it is exactly what stops strategy becoming theatre. A credible model is simple: this person uses this process to generate this insight, which leads to this action, which should create this value. The more organisations can trace that chain, the more likely their long-term strategy will survive contact with day-to-day execution. Without that, future-of-work activity easily becomes a series of disconnected pilots, presentations and good intentions.
Another important observation was that long-term thinking often already exists inside companies, just not always in the places HR first looks. Foresight teams, corporate strategy functions and scenario planning groups are sometimes doing the work of imagining future business conditions, future markets and future operating models. The challenge is that these pockets of long-term thinking are often disconnected from workforce decisions. That is a missed opportunity. If AI is changing how work is delivered, then future-of-work strategy cannot sit on one side of the organisation while workforce planning sits on the other. Those conversations need to meet.
The most tangible example from the discussion focused on the analyst role in professional services. AI has already changed that role. Work that once took days can now be done in hours. Market models can be created far faster. The analyst is no longer doing exactly the same job, and in some cases the traditional entry-level role is already being hollowed out. That creates very practical questions. How many analysts are still needed? What does quality control now look like? Which tasks are genuinely value-adding for humans? What happens to graduate hiring? And if AI takes on more of the lower-level production work, does the classic pyramid structure start shifting towards something more diamond-shaped, with fewer purely junior roles and more oversight, orchestration and client-facing judgement?
That example matters because it moves the conversation away from abstract AI strategy and into work design. Too many organisations are still discussing AI at the level of tools or broad principles. The better question is more direct: which parts of which roles are changing, at what speed, and what does that mean for capability, career paths and workforce economics? That is where I would place the emphasis. The future impact of AI is not something to hold separately in a horizon-scanning deck. It should be changing how organisations write job descriptions, design progression routes, define manager responsibilities and think about the shape of the workforce.
There was also a sharp metaphor in the discussion: the image of people dragging a cart on square wheels while being told that round wheels exist, but replying that they do not have time to use them. That captures the organisational reality well. Most firms are not resisting change because they have never heard the argument. They are resisting because friction sits between the strategic ambition and the execution environment. Senior leaders may talk about the future of the business, but does middle management understand that direction, trust it and feel equipped to act on it? Do teams know what success looks like beyond immediate targets? And have they been given permission to test, flex and learn rather than simply deliver the old model faster?
That points to the real balancing act. Organisations do not balance short-term execution and long-term capability by splitting the difference neatly down the middle. They do it by creating the right conditions. Trust matters. Clear boundaries matter. People need to know the baseline goals they are expected to hit, but they also need room to experiment with new methods, tools and role designs. Crucially, performance measures need to reflect both execution and learning. If every KPI rewards only near-term throughput, leaders should not be surprised when nobody invests properly in future capability.
The discussion also hinted at an important nuance: time horizon should depend on the scale of change. If the organisation wants to change its business model, workforce architecture or core operating logic, it needs a longer strategic horizon. If it wants to improve a process, redesign an input or modernise a workflow, a shorter horizon may be enough. That sounds obvious, but many firms still muddle transformational questions with optimisation questions and then wonder why neither gets the right level of attention.
My own view is that HR has to become much more explicit about this. We need to stop treating AI workforce impact as a future scenario and start treating it as a present design challenge. That means moving from broad conversations about capability building to sharper decisions about work, tasks, judgement, value and structure. It also means being honest that not every organisation is ready for a grand five-year transformation plan. But every organisation can begin with a disciplined approach: identify where AI is already changing role content, decide where human value needs to move upward, connect strategic intent to operating choices, and build metrics that reward both delivery and adaptation. That is how you keep today’s decisions aligned with tomorrow’s workforce reality.
The best summary of the dinner discussion is this: longer-term thinking does not fail because organisations cannot imagine the future. It fails because they have not translated that future into concrete choices about work. The opportunity for HR and business leaders is to close that gap. Not with more abstract commentary, but with a clearer chain from foresight to role design, from insight to action, and from experimentation to enterprise value.