It’s amazing, and perhaps a little frightening: an MIT computer’s algorithms outperformed human intuition in a groundbreaking new test.

The Massachusetts Institute of Technology has been testing a computer system that is supposedly better at finding patterns in data sets than human beings. The Data Science Machine, as it is called, looked through a database of promotional sales dates and weekly profits for the test, according to a UPI report.

Computers are faster than humans at performing a lot of tasks, but one areas humans always seemed to best them is pattern recognition, which computers can do but usually with human help. Where computers usually fall down is in finding the meaning in those patterns rather than just the patterns, but MIT researchers think they may have cracked the code.

To find out if they had, the researchers put their Data Science Machine up against 906 human teams. They didn’t beat them all, but they beat a large junk of them — 615 to be exact.

The teams were tasked with working on n algorithm for months that would predict patterns, while the machine was able to compute these prediction in 12 hours at the most, and sometimes as little as two.

The computer tried to find correlations in the data using some numerical identifiers, according to the report, and then it updates the identifiers as it continues to crunch the numbers. Then, the machine refines its process and starts picking out trends in the data.

Hopefully, this is a good thing for mankind, in that it will be a huge boost to human intelligence and alleviate a lot of the work we do, particularly when it comes to analyizing data.

The findings will be presented this upcoming week at the IEEE International Conference on Data Science and Advanced Analytics.

There are a number of tasks the researchers believe the computer could be helpful at. One example listed in the report is looking at statistics and predict if a student is likely to drop out of an online course. It’s just one of many examples where the machine could immediately be put to work.