I like it! Seems somewhat neglected. Econ researchers have worked on modeling broad economic impacts, but I haven’t seen as much done that tries to model specific impacts on specific jobs. It would be especially helpful if you’re able to extrapolate benchmark results to job automation. Like finding some data on SWE work that’s been automated in the past few years and correlate it to coding benchmark performance, then see if that approach can generalize to other professions.
Thats a good thought. I also have been playing with using aggregated research bottlenecks like gap-map.com and automated research to identify progress towards or completion of those tasks.
I like it! Seems somewhat neglected. Econ researchers have worked on modeling broad economic impacts, but I haven’t seen as much done that tries to model specific impacts on specific jobs. It would be especially helpful if you’re able to extrapolate benchmark results to job automation. Like finding some data on SWE work that’s been automated in the past few years and correlate it to coding benchmark performance, then see if that approach can generalize to other professions.
Thats a good thought. I also have been playing with using aggregated research bottlenecks like gap-map.com and automated research to identify progress towards or completion of those tasks.