Using AI at work means that the best workers will increasingly outperform average workers, right?
Some fascinating new research suggests the opposite. The use of large language models (LLMs) for customer support resulted in large improvements in productivity for new and low-skilled workers, whilst top performers saw little improvement.
One of the reasons for this is that the AI models learnt from the best workers and then transferred that learning to help new and low-skilled workers perform better.
What might be the long-term implications of this?
Maybe top performers will stop trying so hard realising that they can take it easy and perform at the same level as before with the help of AI?
Or maybe top performers will demand additional pay, realising that their work is being used to train models that increases the productivity of an entire team?
What do you think?