AI doesn't replace thinking. It exposes it.

What happens when AI makes information almost free?

My view is simple. AI doesn't replace thinking. It exposes it.

As information becomes abundant, discernment becomes scarce.

Judgement has always mattered

Week four argued that judgement is becoming the defining capability in an AI-enabled workplace. I still believe that. But this week isn't about why judgement matters. It is about why AI makes it much harder to hide the absence of it.

For decades, effort often disguised the quality of thinking. People spent days researching. Weeks writing reports. Hours preparing presentations. Lengthy meetings became evidence of diligence. The people who produced the most information often appeared to add the most value.

AI changes that equation. Today, almost anyone can generate a report, summarise hundreds of pages, draft a strategy, produce a presentation, analyse data and generate twenty ideas before lunch.

The cost of producing information has collapsed. That changes the economics of knowledge work. The value no longer sits primarily in creating information. It sits in deciding what deserves confidence.

AI makes thinking visible

One of the biggest misconceptions about AI is that it makes everyone smarter. The evidence suggests something more nuanced.

On structured, repeatable knowledge tasks, AI often helps less experienced people improve the most. Studies in customer support, professional writing and consulting have consistently shown that generative AI can significantly improve productivity and narrow performance gaps on well-defined work.

That is genuinely exciting. It means AI can raise the performance floor. It helps more people become productive more quickly.

But that is only part of the story. Leadership. Strategy. Hiring. Capital allocation. Product direction. Risk. These are not routine production tasks. They involve incomplete information, competing priorities, trade-offs, organisational context and experience.

AI can help generate options. It can challenge assumptions. It can identify patterns people may have missed. But it cannot decide which path an organisation should take. That responsibility still belongs to people.

The bottleneck has shifted. It is no longer producing information. It is exercising judgement.

More output is no longer the advantage

One of the biggest changes AI brings is that effort becomes less visible. For years, organisations rewarded people for producing more. More slides. More reports. More documentation. More analysis.

But when everyone can generate those things within minutes, quantity stops being a meaningful differentiator. The differentiator becomes quality. Which recommendation do you trust? Which risk actually matters? Which opportunity deserves investment? Which option should you ignore?

AI can produce plausible answers to all of them. Judgement decides which one survives.

This changes leadership

The leaders who thrive will not be the ones who generate the most content. They will be the ones who consistently make better decisions. They will recognise weak reasoning. Challenge assumptions. Understand trade-offs. Know when AI has produced something technically correct but strategically wrong.

Leadership has never been about having the longest report. It has always been about making sound decisions despite uncertainty. AI simply makes that reality much more visible.

Meetings should look different

Many meetings still revolve around sharing information. Reading updates. Presenting slides. Walking through documents everyone could have read beforehand.

AI can already prepare summaries, highlight risks, generate options and draft recommendations. That should fundamentally change how organisations use people's time. Less reporting. More discussion. Less explaining. More deciding.

The competitive advantage no longer comes from who has the information. It comes from who exercises the best judgement once everyone has access to it.

The uncomfortable reality

The impact of AI is unlikely to be the same across every type of work. On routine knowledge tasks, AI is already helping many people perform at a higher level. That is one of its greatest strengths. But as work becomes more strategic, ambiguous and judgement-based, the picture changes.

There is one thread running through both. Effort used to be a signal. It no longer is. On routine work, because almost anyone can produce volume. On complex work, because volume was never the point. Either way, effort stops telling you who is actually thinking.

AI doesn't remove the need for thinking. It increases the importance of it.

Strong thinkers use AI to test assumptions, challenge their own ideas, explore alternative perspectives and improve the quality of their decisions. Weak thinkers can also produce more work. They simply produce average work much faster.

The technology is identical. The outcomes are not. The difference is not AI. The difference is the quality of the thinking guiding it.

My view

I don't believe AI replaces thinking. Nor do I believe it automatically improves it. I believe it exposes it.

On routine work, AI raises the floor by helping more people perform better. On complex work, it makes judgement increasingly visible. As information becomes abundant, discernment becomes scarce. As AI generates more options, the quality of our decisions becomes easier to see.

Perhaps that is the biggest change AI brings to knowledge work. Not that machines become better thinkers. But that humans can no longer hide behind effort when judgement is what ultimately creates value.

---

References

Microsoft & LinkedIn. 2025 Work Trend Index. The report argues that AI is shifting knowledge work towards higher-value human capabilities, including judgement, decision-making and managing AI outputs.

World Economic Forum. Future of Jobs Report 2025. Analytical thinking remains the most important core skill identified by employers, while curiosity, resilience, flexibility and lifelong learning continue to rise in importance.

Erik Brynjolfsson, Danielle Li & Lindsey Raymond. Generative AI at Work (2023). A field study of customer support found the greatest productivity gains among less experienced workers, suggesting AI can narrow performance gaps on structured work.

Fabrizio Dell'Acqua, Ethan Mollick and colleagues. Navigating the Jagged Technological Frontier (2023). Research with consultants found AI significantly improved performance on tasks within AI's capabilities while highlighting the importance of human judgement on more complex work.

Shakked Noy & Whitney Zhang. Experimental Evidence on the Productivity Effects of Generative Artificial Intelligence (2023). AI improved speed and quality in professional writing tasks while compressing performance differences across participants.

OECD. Artificial Intelligence and the Future of Skills. Human judgement, critical thinking and evaluation remain essential as AI becomes more widely adopted.