AI won't replace good professionals - but it will expose bad habits

As routine tasks become increasingly automated, the value of professional work shifts. Less time is spent producing answers; more time is spent identifying the right questions, exercising judgment, and deciding how to act on the information available. In that sense, AI may not reduce the intellectual demands of professional work; it may increase them - and there is evidence for this.

The well-known "jagged technological frontier" study involving 758 Boston Consulting Group consultants found that, on tasks within AI's capabilities, those using GPT-4 completed 12.2% more tasks, worked 25.1% faster and produced higher-quality work. 

But there was a catch.

On a task deliberately selected to sit outside that technological frontier, those using AI were 19% less likely to reach the correct answer. 

For professionals, I think there are three important lessons.

1. Treat AI as a junior colleague, not an expert adviser.
   • AI can be remarkably good at drafting, summarising, brainstorming, and identifying issues; but confidence is not correctness. The professional remains responsible for the advice.
2. Asking the right question is becoming more valuable.
   • Prompting is not magic. It is structured thinking. Giving AI the legal, commercial, and risk context it needs is, in many respects, simply good professional briefing - subject, of course, to protecting confidential information.
3. Know the boundary, and verify the output.
   • The important question is not simply whether AI can produce an answer, but whether you could defend that answer if challenged by a client, regulator, court, or insurer.

And then there is governance.

While policymakers debate how AI should be regulated, a significant part of practical AI governance is already being written into contracts.

Commercial agreements increasingly need to address training-data rights, ownership, and licensing of AI-generated outputs, hallucination and accuracy risk, cybersecurity, audit and transparency rights, human oversight, liability, and indemnities.

For many businesses, those contractual questions may have a more immediate impact than the headline debate about future AI regulation.

AI will undoubtedly automate parts of professional work. But automation is not the same as replacing professional judgment.

The valuable professional of the AI age may not be the person who can generate the fastest answer. Perhaps the most valuable professionals in the AI age will be those who know which questions matter, understand where the technology can add value, and recognise where human judgment still needs to take over.