Two teams of researchers said on Thursday they had created machines that could reason, formulate theories and discover scientific knowledge on their own, marking a major advance in the field of artificial intelligence.
Such robo-scientists could be put to work unraveling complex biological systems, designing new drugs, modeling the world's climate or understanding the cosmos.
For the moment, though, they are performing more humble tasks.
At Aberystwyth University in Wales, Ross King and colleagues have created a robot called Adam that can not only carry out experiments on yeast metabolism but also reason about the results and plan the next experiment.
It is the world's first example of a machine that has made an independent scientific discovery -- in this case, new facts about the genetic make-up of baker's yeast.
"On its own it can think of hypotheses and then do the experiments, and we've checked that it's got the results correct," King said in an interview.
"People have been working on this since the 1960s. When we first sent robots to Mars, they really dreamt of the robots doing their own experiments on Mars. After 40 or 50 years, we've now got the capability to do that."
Their next robot, Eve, will have much more brain power and will be put to work searching for new medicines.
King hopes the application of intelligent robotic thinking to the process of sifting tens of thousands of compounds for potential new drugs will be particularly valuable in the hunt for treatments for neglected tropical diseases like malaria.
King published his findings in the journal Science, alongside a second paper from Hod Lipson and Michael Schmidt of Cornell University in New York, who have developed a computer program capable of working out the fundamental physical laws behind a swinging double pendulum.
Just by crunching the numbers -- and without any prior instruction in physics -- the Cornell machine was able to decipher Isaac Newton's laws of motion and other properties.
Lipson does not think robots will make scientists obsolete any day soon, but believes they could take over much of the routine work in research laboratories.
"One of the biggest problems in science today is finding the underlying principles in areas where there are lots and lots of data," he told reporters in a conference call. "This can help in accelerating the rate at which we can discover scientific principles behind the data."