I was thinking of an objective measure of the impact of GenAI on society - labor productivity [1] seemed like possibly one decent measure.
Gen AI and esp. ChatGPT has been mainstream for around a year and very popular for at least 6 months (reaching almost 2 billion visits in March '24, business leaders having rolled out GenAi-mandates), probably hundreds of billions USD invested, and reports of 2x-10x individual productivity gains, when would you expect to see an impact on some broad economic measures? Which ones?
Any better datasets? The general non-farm labor productivity is clearly not ideal as it encompasses so many jobs that would not be impacted, but manufacturing and lodging+dining (hospitality) seem to account for only 20% of US GDP. More sector-specific datasets are more out of date or more sporadically collected.
[1] https://fred.stlouisfed.org/series/PRS85006092
Which is accurate if you are a demented workaholic or compulsive ladder climber, but hilariously inaccurate if we look at the average office worker. Most office workers barely reach 50% potential productivity by simple virtue that going much higher than this requires well-defined tasks to be delegated effectively amongst a well-managed and motivated workforce that has been sized for the appropriate capacity instead of as an outcome of intracompany politics or as a response to concurrency issues (ergo, badly managed workloads and poorly set agendas sporadically overloading the workforce. This is never, EVER fixed by hiring more people).
Read that again and you'll realize that out of the 5 or so factors I named for good productivity most corporate settings have about 1, maybe 2.
The result of this is that most workplaces are hamstrung fully by Parkingson's law brought about by 2 or more of the above factors being present. They are already capped out on productivity. All involving AI would ever do is excasserbate it, because they will have even less workload to spread over more hours.
Another thing is that by the time you've collected and thought through all the context, direction, nuance and caveats required to make the AI do your knowledge labor for you with consistency....You'll already have done 90% of the actual heavy lifting yourself, and you'll now come to realize that prior to this you just paced your typing to match your thinking, while you know think and have the AI do the typing. The eventual outcome most find is about a 10-20% boost in actual delivery speed, more consistent writing and code style, a decrease in the Stack Overflow visits, and a significant increase in junior errors who prior to this would find out the nuances and caveats while they were bashing away at the issue.