The Colleague That Never Sleeps

Stop thinking of AI as software. Start thinking of it as a colleague.

Not a particularly warm colleague, admittedly. It does not remember your birthday or ask about your weekend. But it is always available, never tired, endlessly patient, and capable of holding the entirety of your organisation's knowledge in its working memory simultaneously. Try getting that from your actual colleagues.

The mental model matters. When we think of AI as a tool – like a spreadsheet or a search engine – we use it transactionally. We ask a question, get an answer, close the tab. But when we think of it as a collaborator, something shifts. We begin to work with it rather than merely on it. And that shift in posture unlocks an entirely different category of value.

The always-on thinking partner

Here is what the colleague that never sleeps actually does in practice.

It holds context. You can share an entire strategy document, a competitive analysis, a set of client requirements, and a budget constraint – and it will hold all of that simultaneously while you think through the problem. Human colleagues are brilliant, but they cannot maintain perfect recall of a 40-page document while simultaneously cross-referencing it against six other sources.

It red-teams your thinking. Ask it to poke holes in your strategy. Ask it to argue the opposing case. Ask it to identify the three things you have not considered. This is not about AI being smarter than you. It is about AI being relentlessly, tirelessly comprehensive in a way that even the best human advisors struggle to match under time pressure.

It drafts at speed. Not to replace your writing – your voice, your judgment, your editorial instinct remain essential – but to give you something to react to rather than a blank page. The cognitive difference between creating from nothing and editing from something is enormous. AI eliminates the blank page problem entirely.

It synthesises across domains. The most valuable insights often emerge at the intersection of fields. AI can draw connections between your marketing strategy and an academic paper on behavioural economics, between your client's industry trends and a regulatory development in another sector. It does not replace domain expertise. It augments it with breadth that no individual human can match.

What changes about knowledge work

When AI becomes a permanent intellectual collaborator, the nature of knowledge work shifts in ways that most organisations have not yet fully grasped.

The value of knowing things decreases. If AI can recall, summarise, and synthesise any body of knowledge on demand, the premium on memorisation and recall collapses. The person who has read every regulatory document is no longer more valuable than the person who can prompt AI to find the relevant clause in seconds. What remains valuable is the judgment to know which clause matters and why.

The value of asking questions increases. AI is extraordinary at generating answers. It is mediocre at generating the questions worth asking. The ability to frame a problem precisely, to identify what is actually at stake, to ask the question that nobody else has thought to ask – this becomes the defining skill of the knowledge worker in an AI-augmented world.

First drafts become free. The economics of content production, analysis, and documentation shift fundamentally when the first draft costs nothing. This does not eliminate the need for skilled professionals – quite the opposite. It means that the finishing, the judgment, the editorial refinement become the entirety of the value proposition rather than a fraction of it.

Collaboration becomes continuous. The traditional model of knowledge work is episodic – you meet, you discuss, you go away and work, you reconvene. AI makes collaboration continuous. You can iterate on a strategy document at 11pm, have AI pressure-test it, refine it further, and arrive at the morning meeting with something three revisions ahead of where you would otherwise have been.

The human premium

None of this means that human knowledge workers are obsolete. It means the opposite – but with a crucial reframing of what "human" means in this context.

AI can synthesise information. It cannot exercise genuine judgment about what matters in a specific organisational context. It can generate options. It cannot take responsibility for the choice. It can draft communications. It cannot read the room, sense the political dynamics, or calibrate the tone to the precise emotional temperature of a difficult conversation.

The human premium in an AI-augmented workplace is not expertise – AI can approximate that. It is not productivity – AI multiplies that. The human premium is accountability. The willingness to make a decision, own the consequences, and stand behind the work. In a world where AI can generate a plausible answer to any question, the person who says "I believe this is the right answer, and here is why" is more valuable than ever.

There is also the irreducible value of human connection. Clients do not hire consultancies because they produce good PowerPoint decks. They hire them because they trust the people. That trust is built through shared experience, demonstrated empathy, and the kind of relational intelligence that AI cannot simulate. Not yet. Perhaps not ever.

How to work with the colleague that never sleeps

The practical advice is straightforward, even if the cultural shift is not.

Give it context, not just questions. The more context AI has about your situation – the organisation, the constraints, the history, the politics – the more useful its contributions become. Treat it like a new team member who needs to be briefed, not a search engine that needs a keyword.

Use it as a red team, not just a yes-man. AI has a tendency to be agreeable. Actively prompt it to challenge your thinking. "What am I missing?" and "Why might this fail?" are more valuable prompts than "Write me a summary."

Maintain editorial sovereignty. AI drafts. You decide. The moment you start publishing raw AI output without human judgment, you have surrendered the one thing that makes you valuable. The colleague that never sleeps is brilliant at preparation. It is terrible at taking responsibility.

Build the habit. The organisations that extract the most value from AI are those where it is embedded in daily workflows, not reserved for special projects. Meeting preparation, research synthesis, document review, strategic thinking – AI should be your first port of call, not your last resort.

The colleague that never sleeps is already at your desk. The question is not whether to work with it. It is whether you will learn to work with it well enough to stay ahead of those who already have.