[vision] Google is not smart, so that's why we need to find the problem!

Updated: Feb 20




Hi, This is Takeshi CEO of su-re.co!!


I am teaching online Salone and high school students sustainability. But what I am really doing is teaching making vision and find a program. To find a problem, you need to understand the meaning behind the problem using multiple sensors and imagination.


Why do we need to find the problem? It is because Google does not tell us. Google may not be as smart as you think and definitely not smarter than you in term of finding a problem. For example, Artificial Intelligence (AI) only speculates on what tears mean based on how everyone else feels about them in the data, not on the real meaning of what they are. An AI that makes decisions based on words alone may not even have seen the image of a tear if the machine does not have a visual sensing system. In other words, this AI judges tears as a good or bad thing by referring to examples in the data. The same can be said for everything. If there are no examples to refer to, AI technology cannot judge a thing. Current AI technology can't find a whole new problem. I started using the word "AI technology", as the current AI is not really intelligent, and it is more like a technology looks like artificially intelligence.


To understand the meaning of a person's tears, a complex perception is needed. What is happening to the face of the person who is crying? Is there a groan? What is the situation around him or her? Is there anyone else with them? What colour of shirt are they wearing, what shoes are they wearing? All of these things help us to understand the tears.


Maybe, you are crying because your girlfriend has dumped you. Perhaps, you're crying because you've won tickets to a concert you've dreamed of going to. Your three-year-old is crying because he fell. Without a holistic view of the various perceptions, it is impossible to understand the meaning of the tears and what is at stake.


I give you another example!! If a computer cracker infiltrates the server room of Google, the world's best artificial intelligence technology, and raises a huge hammer, what will happen to the AI? If the server room cannot perceive a person with the huge hammer, and if it does not have the data to refer to the meaning of the huge hammer raised, then it will not be possible for the AI to understand the problem that is happening to it.


There are services claiming to be AI translation, but they are not really AI, but an artificial intelligence technology. These days, you can use online dictation, which is then translated your language into another language. Before Corona, once you had a smartphone and an internet connection, you could travel anywhere in the world by using it as an automatic translator. Whether in Russia, China, Egypt, Dubai or any other Asian or European country, I never had a problem to travel with the Google Translate Conversation function on my smartphone. The AI is not (yet) able to understand the meaning of what I am trying to say here. There is no way to make a computer understand this meaning.


The difference between science fiction and fantasy is that science fiction is a story that is an extension of current technological innovation. In contrast, fantasy is something that goes beyond it. In this context, artificial intelligence which people believed to exist may actually be closer to fantasy.


As for why I think this, I think that humans, with their comprehensive perceptual skills and ability to fantasise, are superior to AI. We need to expand our education in this area.



21 views0 comments
 

Subscribe Form

+62 812-3831-727

  • Google Places
  • Facebook
  • Twitter
  • LinkedIn
  • Instagram

©2020 by su-re.co (Sustainability & Resilience) Gift-maker to the Earth. Proudly created with Wix.com