Hard data that ChatGPT helps with work.
From the study:
The tasks demanded clear, persuasive, relatively generic writing, which are arguably ChatGPT’s central strengths. They did not require context-specific knowledge or precise factual accuracy.
And:
We required short tasks that could be explicitly described for and performed by a range of anonymous workers online
The graphs also show greater improvement for the lowest performers than for the high performers.
Definitely an encouraging result, but in line with anecdotes that, currently, LLMs are only useful for genetic and low complexity tasks, and are most helpful for low performers.
I think this is probably true. In my own use case, I primarily use it to quickly create summaries, comparisons, and contrasts of things I am searching for, such as:
Give me a summary of using Adobe Experience versus Madcap Flare to create a websiteWhat are the pros and cons of using madcap flare to build a website
