Claude Fable 5 just posted a 16.1 percent success rate on the Remote Labor Index, a benchmark of 240 paid freelance projects that demand end-to-end professional output. The jump from 4.2 percent five months earlier has prompted Dan Hendrycks to fit a basic sigmoid curve and conclude that most remote tasks could clear the automation threshold within roughly twelve months.
A basic sigmoid extrapolation suggests that this measurement will be close to its performance ceiling in around a year---meaning most randomly sampled remote work projects would be highly automatable next year.
The automation rate of remote projects has increased ~4x in the past five months.
The projection rests on a single functional form applied to four data points from one benchmark. No one has yet published confidence intervals or tested alternative growth models against the same tasks.
Projects that currently pay a median of two hundred dollars and take half a day are the exact slice moving from 16 percent to near-ceiling automation. Workers in software, design, and data analysis will see demand shift first if the trend holds.
No Digg Deeper questions have been answered for this story yet.
The automation rate of remote projects has increased ~4x in the past five months.
New Remote Labor Index results: AI automation of real remote work is increasing fast. Claude Fable 5 now completes 16.1% of projects at a professional standard, roughly double the next model and up from Opus 4.6’s 4.2% automation rate.
@hendrycks for a sec. i was confused if this was the Dan i kindof know. apparently mans going thru some soul-upheaval. hang in there dan.
@hendrycks 🤔 But.. I work remote...
@hendrycks Admittedly I keep wondering what 'professional standard' actually means here. Some remote work is pretty formulaic anyway.
@hendrycks A basic sigmoid extrapolation also suggested we'd have full self-driving by 2018. S-curves look obvious in hindsight, but calling the inflection point in real time is another matter.
