This might sound weird, but now researchers are developing robots with artificial intelligence (AI) that can learn by simply observing, and watching YouTube videos.
Latest attempt to develop artificial intelligence in robots
University of Maryland’s researchers recently made it in headlines as they are working on certain robots that can learn and respond to visual stimuli. The project is funded by DARPA, and its aim is to create robots that can collect information and act accordingly. Currently, researchers just want their robots to cook!
The team consists of researcher Yiannis Aloimonos, Cornelia Fermuller, Yezhou Yang and NICTA Australia’s Yi Li.
DARPA’s program manager Reza Ghanadan recently interacted with journalists and shared some information about this project. Reza said that researchers programmed these robots to carefully watch YouTube videos, and process the knowledge about various objects used in concerned cooking videos. Later, these robots applied all the acquired knowledge on objects in real world.
Fortunately, these machines successfully used the collected data to identify and handle certain kitchen objects, just like they were handled by humans in the YouTube video.
Reza Ghanadan also said that their team of researchers is working on technology that can help robots to retain the data that they acquire after watching videos and share it with other robots as well.
If this project succeeds, researchers won’t have to use time-consuming and expensive programming codes in machines that perform various tasks. Instead, they can start implementing this new technology that promotes learning and knowledge sharing approach in machines.
Vicarious’s project to develop artificial intelligence
Last year, Ashton Kutcher and Mark Zuckerberg attracted a lot of attention when they declared their investment in artificial intelligence developer Vicarious FPC. The company’s co-founder, Scott Phoenix interacted with reporters and said that their company is working on developing a part that will work just like human brain’s “neocortex.” If the company successfully develops this part, it would be able to help any robot to see, control its body parts, learn languages, perform calculations, and do almost everything that humans do.
Robots can never learn and behave like humans
PayPal’s co-founder Elon Musk recently shared his opinion about artificial intelligence with one of the top newspapers in Britain. He is the co-founder of rocket technology firm called- Space X, and electric car manufacturing start-up called- Tesla. Co-incidentally, Elon is also one of the main investors in Vicarious.
Elon Musk believes that artificial intelligence can prove to be the biggest threat for humanity. He also suggested that all the countries should come together and form regulatory unit to make sure that companies and researchers do not create evil robots like those in Terminator, Robocop and I-Robot.
Artificial intelligence expert and Microsoft’s researcher Stephen Hawking also shared similar opinion during one of his recent interviews. He believes, AI can cause several psychological, economic, ethical and legal issues in society.
Machines might learn faster than humans in the future. But, they may not act and implement their knowledge like humans.