AI

The Robot Blog

     What do you think of when someone says "robot?"

     I think of the Jetsons. Diligent and sassy Rosie, always cleaning something and making quirky, self-aware remarks about the things going on around her. And then I think of I, Robot; and of Isaac Asimov's three laws of robotics. Those, should anyone need a refresher, are verbatim as follows:

1. A robot may not injure a human being or, through inaction, allow a human being to come to harm.
2. A robot must obey orders given it by human beings except where such orders would conflict with the first law.
3. A robot must protect its own existence as long as such protection does not conflict with the first or second law. 
                                                            I also think of Bender from Futurama.

                                                            I also think of Bender from Futurama.

     Now, those were from Asimov's science fiction, and were largely a literary device he used to create the plot of the story around his laws of robotics posing an unfair paradoxical conundrum to a self aware robot. BUT, the world is making some huge leaps and bounds in the development of robotics and artificial intelligence these days. Asimov would probably be both amazed and terrified of the advances we've made since 1992. We do not have similar laws of robotics in the real world today. Where we stand right now, robots only do what they are programmed to do. But programming is becoming more and more intricate, and we are starting to see programming being written that can continue to adjust itself as the AI "learns" from its environment. There are some fantastic examples of this, and you don't have to look very far to learn about them. 


     Let's start with a fun one! Here:  Meet Cleverbot. Go ahead, take a quick field trip to talk to Cleverbot, I'll wait. :) 

                                                   The sassiest chatting computer in the world.

                                                   The sassiest chatting computer in the world.

     Cleverbot is AI developed to continuously learn how to interact with humans by talking to them in a chat format. Cleverbot was created by British AI scientist Rollo Carpenter. The AI was introduced to the world at large (online) in 2006, but Carpenter has been teaching and developing this little jewel since 1988. In 2010, Cleverbot was rated 42.1% human in a Turing Test in which an audience watched volunteers interact with it. That same year, this clever little bot won the BCS Machine Intelligence Competition. Cleverbot is a fantastic example of AI with programming that learns as it goes. It's also a fun friend to talk to when you're feeling lonely.


 

     British AI researcher Simon Colton  is developing and teaching AI called The Painting Fool  to create art. The Painting Fool self identifies as an artist seeking recognition, and wants to be taken seriously as a creative artist. It has been programmed with components that give it moods, an imagination and an ability to create what it perceives as art based on the things it sees, reads, and interacts with; such as news articles, its teachers, movies, etc. Its skill set and the traits the creators are focusing on imbuing it with are being endearingly referred to as Artificial Imagination.  Have a look at the galleries of its paintings. They're nothing short of amazing when you stop to think that someone spent hours and hours writing code for this machine to be able to make its own decisions about what to paint! The Painting Fool has had its work featured in physical and online galleries--most prominently in an exhibit at the Galerie Oberkampf called You Can't Know My Mind, where the AI spent time learning more about painting by painting portraits of visitors. The Painting Fool is well on its way to becoming a recognized artist.

This art was imagined and created by an AI program called The Painting Fool. This is the result of  AI/computer programs being given the ability to decide what art is, and it is simply fascinating. 

This art was imagined and created by an AI program called The Painting Fool. This is the result of  AI/computer programs being given the ability to decide what art is, and it is simply fascinating. 


 

In a similar vein of AI and art, Google is developing its own AI that, among other things, can have artificial "dreams." Not aspiration dreams, (perhaps they will incorporate that or are already working on giving the AI an ability to choose what it would want to do on a life-long basis) but sleep dreams. The AI has been producing stunning surreal images that are a sort of regurgitation of original images that it is given to process. The researchers working with this AI are trying to gauge how their AI is learning by watching it process images. This type of AI is called an Artificial Neural Network, or ANN, and they are modeled after the way that animal brains work. This research is stunning, and worth so much more blog space than I have time or space or knowledge of to fit into this blog. I will have to move on because there is much more I want to cover, and I've spent far too long gushing about learning AIs.

 

One last honorary fun/fictional mention for AI that learns as it goes: the 2015 film Chappie is an absolutely heart melting film set in the near future, about a scrapped police unit android who is given sentient learning AI by his creator, but then tossed unexpectedly into the cruel streets of Johannesburg, South Africa where he must learn quickly to adapt to his new life. If you have the chance to watch it, it is a fantastic action packed science fiction film about AI adapting to its environment. (I think I'll go ahead and use this opportunity as a segway as well.) Speaking of heartwarming robot movies, Wall-e is a classic favorite for many. Wall-e represents a lot of the typical things that people think of when they think of robots. He's a little subservient dude who is made of metal parts to fulfill a set of specific purposes. The next set of robots and robot related things I want to show you are going to be more like Wall-e than the previous AIs we covered that not many people can picture as a physical robot. I'd like to lead with this thematic quote by Social Robotics and Human Robot interaction mogul Dr. Cynthia Breazeal:

 

"If you look at the field of robotics today, you can say robots have been in the deepest oceans, they've been to Mars, you know? They've been all these places, but they're just now starting to come into your living room. Your living room is the final frontier for robots." -- Dr. Cynthia Breazeal

 

Let's talk about space robots for a minute, shall we? Just a brief minute, I promise. There are so many that I don't even know where to begin, so let's take a side road here. We all know about NASA's great probes and rovers from Luna 9 all the way to the Mars Curiosity Rover, but I want to talk more about where this technology is going more than where it's been, mostly for the sake of keeping this blog shorter than a text book. So I want to show you where little space robots are starting to go, in a more relatable sense to the common everyday person. Did you know that you can have your own satellite? That's right.

NASACubesats.jpg

These little guys are called Cubesats. They're being marketed as personal micro-satellites that can be used for schools, small governments, etc to help with projects involving pictures from space or communication via satellite, or even to help with R&D for future technology. Currently, they're pretty unaffordable to purchase completely out of pocket, but they're quoted at under $50,000, which is remarkably cheap as far as space technology goes--and they are simple enough to assemble yourself! Isn't that fascinating? They're so versatile that NASA is even beginning to use them as less expensive deep space probes

 

 

Alright, let's float gently back down to Earth now, and talk about those living room robots. Going back to the work of Dr Cynthia Breazeal, I'd like to introduce you to Jibo, the world's first social personal robot. He is designed to be a big help around the house. He can order food for you, tell your children stories and entertain them, talk to you, sync with your phone and read you messages, make and remind you of appointments or events, help you with recipes while you cook, and let you video chat with your friends and family. He is built and programmed to learn your and your households' names and faces and to be able to interact with you and learn your preferences. He is a personal assistant with a personal touch. Jibo is a brilliant idea in the making, and the first batch of Jibos, funded by Indegogo and on a pre-order basis, is set to ship out around April 2016. Here is Jibo's Indegogo page, with some FAQ, but the campaign has met its funding needs and was closed in June earlier this year. You can currently be put on a waiting list for news on new pre-order opportunities for Jibo. 

 

 

And let's switch gears one more time before winding down this awesome robotic smorgasbord: Robotics are currently dominating the world of medical prosthetics. Here is an MIT technology review on brain-controlled prosthetic limbs being developed to give paralyzed patients the ability to move and use robotic limbs that can be controlled by thought. This is amazing stuff that people said was impossible less than 50 years ago. The article also chronicles some major milestones in the research it took to develop this technology to this point. It touches on the advanced research on the human motor cortex and the work that went into developing programming able to recognize and utilize neurological brain activity from the motor cortex. Exoskeleton robotics are also being developed for the lower half of the body, to help paralyzed people walk. Here is a brief article on the work of Vanderbilt University's Center for Intelligent Mechatronics, detailing some of their work on advanced exoskeleton machinery for mobility and independence for victims of paralyzation. So, if you ever find yourself in need of a new arm or leg, optimally, you could end up with some really awesome robot limbs.

 

 

And finally, can I get a robotic drumroll, please? I saved the most useful link for last. Would you, reader, like to learn the basics of building your own robot? Would you be even more interested if you could learn for free? MIT is offering an open online Introduction to Robotics course. The course's prerequisites are also available in MIT's open courses. The course comes with lectures, lecture notes, assignments, exams, and projects. It's an intro course, so you're not likely to be learning to make a robot that can paint or learn or read your brain's signals, but you could still build something helpful, or something just for giggles. The project posted as the last class' featured project was a robot that could "rescue" a doll. Perhaps you could build a semi-small robot that can retrieve things, or one that can do a little dance. With everything that we've looked at and learned about today, I'd be willing to bet that with enough time and experience, if you can dream it, you can program a robot to do it.