There is unprecedented potential for AI to raise standards in average productivity and quality for knowledge workers. However, this comes with caveats.
A Working Paper just published by Harvard Business School resulting from a research collaboration between Boston Consulting Group and academic experts, highlights the potential of AI to raise standards in average productivity and quality for knowledge workers.
Although, this comes with caveats.
The results from an experiment involving 758 of BCG’s consultants found that augmentation with AI benefited consultants across skill levels, particularly those whose baseline was below average performance.
However, for tasks that were deliberately designed to be difficult for the AI to produce correct results, those consultants that used AI were less likely to produce the correct solution than those that did not use AI because those using AI assumed it to be correct. Relying on these tools risks ‘falling asleep at the wheel.’
The utilisation of AI to supplement or delegate tasks needs to be coupled with awareness of the current limitations of AI.
The ‘jagged frontier’ of AI capabilities means that these systems are suitable for some tasks, yet inadequate for other tasks of equivalent or less complexity but in different domains.
When it does make sense to utilise a large language model such as ChatGPT as a sparring partner or collaborator, we need to stay diligent to ensure it’s a true collaboration, with continuous scrutiny and peer review as part of the process.
Future ready?
Comments