Has AI inspired you to do more at work?

There is a useful way to test the impact of AI at work.

Do not start with the big question.

Do not ask whether AI will replace jobs. Do not ask whether AI is coming for your industry. Do not ask whether a model can do everything a person can do.

Start smaller. Ask this: has AI inspired you to do more at work?

Not search more. Not type faster. Not produce longer documents.

Actually do more.

Pitch the idea. Write the first draft. Prepare better for the meeting. Ask a sharper question. Build the proposal that would previously have stayed in your head. Finish the document that would normally have waited for next week. Turn a vague thought into something someone else can react to.

That is where the AI conversation becomes more interesting. Because the most important workplace shift may not be that AI gives people answers. It may be that AI gives people a stronger starting point.

And a stronger starting point changes behaviour.

AI is not only compressing the time between idea and output. It is compressing the time between confidence and action.

The first visible impact is not replacement. It is activation.

Much of the public debate around AI still sits in a replacement frame.

Will AI replace copywriters? Will AI replace analysts? Will AI replace developers? Will AI replace customer support agents? Will AI replace junior staff?

Those questions matter. But they are not the only useful questions.

Inside organisations, the early impact often feels different. AI is helping people activate work that was previously stuck.

A blank page becomes a rough outline. A messy thought becomes a structured argument. A long document becomes a summary. A difficult message becomes a draft. A meeting becomes something people can prepare for in minutes rather than hours.

This matters because a great deal of knowledge work does not fail at the finish line. It fails at the starting line.

People delay because the first version feels too heavy. They wait because they need structure. They pause because they do not want to look unprepared. They avoid the task because they do not know where to begin. They ask for a meeting because they have not yet shaped their own thinking.

AI reduces some of that friction. It does not remove the need for judgement. It does not remove accountability. It does not remove expertise. But it can help people move from intention to action.

That is not a small productivity improvement. It is a behavioural shift.

The evidence supports the productivity story, but with important limits.

There is credible evidence that generative AI can improve productivity in some workplace tasks.

A widely cited experiment by Shakked Noy and Whitney Zhang tested ChatGPT on mid-level professional writing tasks. The study found that access to ChatGPT reduced average completion time by 40% and increased average output quality by 18%.[1]

That is important because writing is not a niche activity in modern organisations. Emails, briefs, summaries, proposals, policies, updates, presentations and notes all rely on the same underlying ability: turn thinking into language.

If AI helps people do that faster and better, the effect is not limited to "writing". It touches execution.

Another major study, published in the Quarterly Journal of Economics, examined the use of generative AI assistance by more than 5,000 customer-support agents. The study found a 15% average productivity increase, measured by issues resolved per hour, with the largest gains among less-experienced and lower-skilled workers.[2]

That finding matters for a different reason. It suggests AI can raise the floor. It can help people with less experience perform closer to more experienced colleagues, at least in some structured work environments.

But the evidence is not a licence to claim that AI improves everything. The same body of research also warns us to be careful.

The Harvard Business School and Boston Consulting Group study on the "jagged technological frontier" found that consultants using GPT-4 performed better on tasks within AI's capability range, but could perform worse on tasks outside that range when they trusted AI too much.[3]

That is the nuance. AI can help people move faster. AI can improve first versions. AI can raise the baseline. But it does not remove the need to know when the output is useful, incomplete, generic, misleading or wrong.

The productivity story is real. The judgement story is just as real.

AI is changing the economics of starting.

For many people, the biggest change is not that AI completes the task. It helps them begin.

This is where the workplace impact becomes subtle but powerful.

Before AI, starting often carried a cost. You needed to open a blank document. You needed to decide the structure. You needed to remember the background. You needed to make the first sentence work. You needed to know what "good" might look like. You needed to ask someone for help if you were stuck.

That friction slowed work down. Sometimes it improved quality. Struggle can be useful. But often it simply delayed progress.

AI changes the economics of starting because the first version becomes cheaper. Not cheaper in the sense of being less valuable. Cheaper in the sense of being easier to produce.

A person can now ask for a first structure. They can ask for a set of questions. They can ask for an executive summary. They can ask for risks and trade-offs. They can ask for a sharper version of a weak paragraph. They can ask for a comparison they can then verify.

The first version is no longer the expensive part. The value moves to what happens next.

Do you challenge the answer? Do you improve it with context? Do you verify the claims? Do you adapt it for the audience? Do you know what should be removed? Do you add the judgement that only comes from experience, customer reality and commercial context?

That is where the human premium sits.

The real unlock may be confidence.

One of the most under-discussed effects of AI at work is confidence.

People do not only use AI because they are lazy. They use it because it reduces the social and psychological cost of trying.

A junior employee may ask AI a question they would be embarrassed to ask in a meeting. A second-language English speaker may use AI to express strong thinking with more confidence. A manager may use AI to prepare for a difficult conversation before speaking to a team member. A salesperson may use AI to turn scattered notes into a customer-ready follow-up. A founder may use AI to shape a pitch before sharing it with investors. A busy executive may use AI to pressure-test a memo before sending it to a board.

In each case, AI is not simply generating text. It is lowering the activation barrier.

That is powerful because many organisations do not only suffer from a lack of ideas. They suffer from unexpressed ideas.

The proposal that is never written. The process improvement that is never framed. The customer concern that is never escalated. The meeting question that is never asked. The strategic thought that never leaves the notebook.

AI can help move those thoughts into the world. That is why the question "Has AI inspired you to do more?" matters. It gets closer to the human behaviour underneath the tool.

The workplace is already adopting AI faster than policy can catch up.

The adoption data shows that AI is no longer a fringe workplace tool.

Microsoft and LinkedIn's 2024 Work Trend Index reported that 75% of global knowledge workers were already using AI at work, and that use of generative AI had nearly doubled in the previous six months.[4]

The same report found that many employees were bringing their own AI tools into work, which tells us something important: adoption is not only being driven by top-down corporate transformation programmes.[4] It is being pulled into the workplace by individuals.

People are using AI because they feel the pressure of work. They are using it because they need leverage. They are using it because the tool gives them a way to start, structure, summarise and improve.

That creates a challenge for organisations. AI adoption is happening whether companies are ready or not. The question is whether leaders turn that adoption into better work, or simply allow unmanaged AI use to produce more noise.

Because doing more is not automatically the same as doing better.

More output is not the goal.

This is the part that matters most.

AI can make it easier to produce. More drafts. More summaries. More emails. More slides. More meeting notes. More documents.

But volume is not value. A workplace drowning in faster documents is not necessarily a better workplace.

The goal should not be more content. The goal should be more useful contribution.

Better preparation. Clearer thinking. Faster alignment. Sharper decisions. Stronger customer communication. More thoughtful leadership. More people confident enough to participate.

AI is valuable when it improves the quality of movement. It is less valuable when it simply accelerates activity.

That distinction will become more important as AI becomes more common. When everyone can generate a decent first draft, the differentiator is no longer the first draft. The differentiator is the judgement behind the final version.

The new question for leaders

For leaders, the question is not only whether teams are using AI. That is too basic.

The better questions are: What work is AI helping people start? What ideas are surfacing that used to stay hidden? Where is AI improving confidence? Where is it raising the baseline? Where is it creating more noise? Where are people accepting answers too quickly? Where do teams need better verification habits? Where does human judgement still need to lead?

AI should not become a shortcut around thinking. It should become a tool for better thinking.

That requires a different kind of management conversation. Not "Did you use AI?" But: What did it help you see? What did you reject? What did you verify? What did you change? What did you add that AI would not know? How did it improve the work?

Those questions turn AI from a private productivity hack into a visible learning tool.

The opportunity is bigger than speed.

The easiest AI story is speed. Faster writing. Faster research. Faster summaries. Faster preparation. That story is true, but incomplete.

The more interesting story is agency. AI can give people the confidence to attempt work that previously felt too complex, too slow or too exposed.

That matters. Because when more people contribute, organisations can become more inventive. Not because AI has all the answers. Because more people can get their thinking into a form that can be debated, improved and used.

That is the positive version of the AI workplace. Not a world where humans stop thinking. A world where more humans can participate earlier, with better preparation and stronger tools.

The risk is shallow acceleration.

There is also a risk. AI can make weak thinking look stronger than it is.

It can make a generic idea sound polished. It can make a thin argument look structured. It can make a confident answer feel more reliable than it deserves. It can create a false sense of progress.

That is why the answer is not blind enthusiasm. The answer is disciplined adoption.

Use AI to start. Use AI to explore. Use AI to structure. Use AI to improve. But do not outsource accountability.

Do not confuse fluency with truth. Do not confuse polish with insight. Do not confuse output with contribution. Do not confuse speed with progress.

That is the balance. AI can inspire people to do more. The leadership task is to make sure that "more" becomes better work, not just more work.

The question

So here is the question. Has AI inspired you to do more at work?

Not search more. Not type faster. Not produce longer documents. Actually do more.

The idea you pitched. The proposal you finally drafted. The customer note you improved. The meeting you prepared for properly. The difficult conversation you structured. The concept you finally understood. The work that moved from "one day" to "today".

That is where the real signal is. Not in the hype. In the work people are now brave enough, prepared enough and equipped enough to start.

---

Sources

[1] Shakked Noy and Whitney Zhang, "Experimental evidence on the productivity effects of generative artificial intelligence", Science, 2023. [2] Erik Brynjolfsson, Danielle Li and Lindsey Raymond, "Generative AI at Work", Quarterly Journal of Economics, 2025. [3] Fabrizio Dell'Acqua et al., "Navigating the Jagged Technological Frontier", Harvard Business School working paper with Boston Consulting Group, 2023. [4] Microsoft and LinkedIn, "2024 Work Trend Index: AI at Work Is Here. Now Comes the Hard Part", 2024.