Hi Hi Hi,
Di postingan ini gw akan nge-share artikel yang dibuat oleh salah seorang mahasiswa/i gw mengenai Artificial Intelligence. Postingan ini salah satu bentuk apresiasi gw atas pemikiran mereka. Nama pembuat ada di judul postingan.
What is your first thought when you hear the word “Artificial Intelligence”? Maybe what you have in your mind is about robots. It’s not completely wrong, robots are one of Artificial Intelligence (AI). But, did you know that AI is simpler than that?
AI is already old, about 60 years old. Maybe you are wondering, does that mean that 60 years ago there were AI robots like now? The answer is certainly not, Turing Test or Turing’s Imitation Game is the first artificial intelligence program was created by Alan Turing. Shortly, the test was set up to examine machine thinking (Warwick, 2015). That program then inspired the next AI discovery. Then, the first digitally operated and a programmable robot called the Unimate was invented by George Devol in 1954 (Bellis, 2018). Well, the simple robot is only consisting of an arm for industry automated diecasting. But at that time, the term “Artificial Intelligence” had not been used. Around 5 years later, the term “Artificial intelligence” was coined by John McCarthy (Childs, 2011). After that, the first ‘intelligent’ humanoid robot, WABOT-1 was built in Japan in 1972 (Ray, 2018).
Picture 1: From left to right, Unimate and WABOT-1 (medium.com, 2018)
After all that big discoveries, AI development was slowed down due to inadequate computers back then. But now, they are powerful enough to really emulate kind of specialized human thinking. And then the computers take advantage of the fact that they can look at much more data than people can (Thrun & Anderson, 2017). People then begin to look back on AI and develop it again, especially in Machine Learning (ML) methods.
Nowadays, we always hear machine learning and AI on the same occasion. Actually, what is machine learning? Is “Machine Learning” another term for AI? Well, AI is an umbrella that holds other fields like image processing, cognitive science, neural networks and much more. Machine Learning is also a part of this umbrella. It is a subset of AI (Toshniwal, 2017). So next question will come to your mind, why machine learning is more popular than other AI subfields?
To answer that question, think about how you raise a child. You don’t spend the first 18 years giving kids a rule for every contingency and set them free and they have this big program. They stumble, fall, get up, and they have a positive experience, a good grade in school, and they figure it out on their own. That’s happening with computers now, which makes computer programming so much easier suddenly. They figure it out on their own. Now we don’t have to think anymore. We just give them lots of data (Thrun & Anderson, 2017).
That visualization also answers what machine learning really do. Machine learning is actually a system that learn directly from examples, data, and experience. If the broad field of AI is the science of making machines smart, then machine learning is a technology that allows computers to perform specific tasks intelligently, by learning from examples (The Royal Society, 2017). For example, we make a machine learning program to tell us what people do from a picture. Then, we give the program millions of example pictures test cases about everything people do; sitting, standing, smiling, basically every common thing we do.
Picture 2: Birthday Party (off-glass.com, 2018)
After the program is complete, we give the program a practice case which is the Picture 2, the program easily recognizes it and say that “there are 5 children are smiling.” Of course, it’s an incredible accomplishment because the program correctly recognizes the picture. So far, we have just taught the computer to see objects (Li, 2015). If we give it other type of picture, what will happen? For example, there are only a cat in the picture. What will the program do? Of course, the program will recognize it wrong because it never learns it beforehand.
There are many amazing discoveries on AI, especially on machine learning, for example the “5 children are smiling” case above. But actually there are so much more in the Picture 2 rather than just “there are 5 children are smiling.” What computer doesn’t see is that the children are friends and they are really happy celebrating a birthday party together. There are much more other things to discover and this is still far from perfect because computer is only memorizing, not understanding. Machine learning basically only remembers what it had learned, but we humans can understand the meaning. That is something which is very, very lacking in AI (Arai, 2017).
- Arai, N. (2017). Can a robot pass a university entrance exam?. TED2017. Retrieved 5 December 2018, from https://www.ted.com/talks/noriko_arai_can_a_robot_pass_a_university_entrance_exam
- Bellis, M. (2018). Who Pioneered Robotics?. Retrieved 5 December 2018, from https://www.thoughtco.com/timeline-of-robots-1992363
- Childs, M. (2011). John McCarthy: Computer scientist known as the father of AI. Retrieved 5 December 2018, from https://www.independent.co.uk/news/obituaries/john-mccarthy-computer-scientist-known-as-the-father-of-ai-6255307.html
- Li, F. F. (2015). How we’re teaching computers to understand pictures. TED2015. Retrieved 5 December 2018, from https://www.ted.com/talks/fei_fei_li_how_we_re_teaching_computers_to_understand_ pictures
- Ray, S. (2018). Artificial Intelligence in Depth. Retrieved 5 December 2018, from http://www.rayadvisors.com/blog/artificial-intelligence-in-depth
- The Royal Society (2017). Machine learning: the power and promise of computers that learn by example. Creative Commons Attribution License.
- Thrun, S. & Anderson, C. (2017). What AI is — and isn’t. TED2017. Retrieved 5 December 2018, from https://www.ted.com/talks/sebastian_thrun_and_chris_anderson_the_new_generation_of_computers_is_programming_itself
- Toshniwal, A. (2017). Quora answer on What is the difference between AI and machine learning?. Retrieved 5 December 2018, from https://www.quora.com/How-is-AI-different-from-Machine-Learning
- Warwick, K. & Shah, H. (2015). Can machines think? A report on Turing test experiments at the Royal Society. Journal of Experimental & Theoretical Artificial Intelligence 28(6): 989-1007.
Nice article Albertus. Keep the good work! Check his blog at https://aheronius.wordpress.com/
See you di postingan selanjutnya.