By Alyssa Shaffer
Preprogramming everything into a robot requires clairvoyance. By rather than becoming a psychic, software developers have decided to focus on creating learning algorithms. These kinds of algorithms give robots the ability to learn from demonstration and exposure to stimuli. Despite the beginning stages of progress in this young field, engineers also need to address how artificial intelligence will play a part in a robot’s intelligence. But for now, they need to figure out which learning methods make robots most efficient.
So what’s the big deal with artificial intelligence?
Even though our understanding of AI is still limited, many prominent scientists have been debating over the potential dangers of it. Some, like Rethink Robotics CEO Rodney Brooks, see AI as a potential to better technology and mankind. Brooks’ TED Talk in 2013 argues that AI and robots can assist the baby boomers as they age and outnumber the workforce over the next 40 years. Others, like Microsoft Founder Bill Gates, think the experimental technology can grow out of our control. He and Stephen Hawking have expressed that humans will not be able to control AI in the long-term.
The possibility of AI is also up for debate. Some people argue that humans are just not able to build a technology that surpasses their own understanding since, ironically, it would require a understanding beyond our own. But others say those cynical people underestimate our own potential. So could someone develop an artifical intelligence that surpasses human understanding? Is AI really the solution to unlocking a robot’s potential? Are learning algorithms like current research the better alternative? Everyone has an opinion, and the debate will continue as the research progresses.