Over the past few years, there have been a number of proclamations on the rise of machines and artificial intelligence in the workplace. We’ve had Google testing driverless cars, IBM’s Watson beating the grand master of Jeopardy and beating a path to a hospital near you.
Probably the finest chronicle of this revolution has been The Second Machine Age by MIT academics Andrew McAfee and Erik Brynjolfsson, which suggests that there will soon come a time when machines can easily replicate everything that a human can do.
All of which is rather worrying for those looking to enter the jobs market anytime soon. One might assume that cooking will remain a task best suited to human beings however and that it’s impossible for a machine to replicate the kind of artistry that a good chef can bring to a meal.
Meet Robot Chef
Well, a recent project suggests you may well be wrong to think that. The American military agency DARPA is currently developing a robot that can learn how to put a recipe together simply by watching a human cooking on YouTube.
Of course, that’s something that many a student has done down the years, but for it to be achieved by a robot is quite something. The machine is capable of understanding what kind of utensils are required to make their dish, and can replicate their use without any human having to get involved.
Now, you may be wondering quite why the military is looking at robot chefs? The logic is that cooking is a skill that requires a number of skills that are quite difficult to learn, and that may be useful in other fields of robotics.
For instance, robots are currently quite good at recognizing patterns or objects, but they’re not so good at interpreting visual information and then performing actions based on this knowledge.
Alas, DARPA believe that their latest breed of robots may have overcome this hurdle and can trnslate visual information into valid activity. The robot was trained using a series of cooking videos hosted on YouTube.
The machine was capable of teaching itself to cook after the following the videos, with meals produced with a high level of accuracy. After a short period of time, it could recognise objects in the video with 79 percent accuracy, while also understanding how the objects were held by the human chef with a 91 percent accuracy level. It could also predict the appropriate action 83 percent of the time.
Interestingly, the machine was also capable of remembering the things it had learned, such as how to hold a spoon correctly. DARPA hope that they could easily share this new learning with other robot chefs.
"This learning-based approach is a significant step towards developing technologies that could have benefits in areas such as military repair and logistics," DARPA say.
It’s fair to say that a fully functioning robot chef is perhaps a little way off yet, but this latest experiment suggests that one might not be that far away.
Do you think that we will soon have robot chefs in our restaurants? Or do you think this is all speculation? Your thoughts and comments below please...