AI models can't handle real freelance work, study finds
A new study shows that even the best AI models are struggling with actual freelance work.
Researchers tested six top AIs on real projects from areas like game dev, design, and data analysis—using a benchmark called the Remote Labor Index (RLI), which tracked over 6,000 hours of paid human work.
Today's AIs struggle with complex jobs
The leading AI model, Manus, could only finish 2.5% of projects at paid quality—earning just $1,810 out of a possible $143,991.
No other model broke past 3%.
Essentially, today's AIs appear to struggle with complex jobs that require flexible thinking and problem-solving.
Robots won't be taking our jobs anytime soon
For anyone worried about robots taking all the remote gigs: the study's findings suggest we're not there yet.
Businesses eyeing full automation should know that replacing freelancers with AI does not appear realistic at this time.
For now, human creativity and adaptability are still tough to beat.