Results 1 to 4 of 4

Thread: Human Level Artificial Intelligence, Mitchell Kwok

  1. #1

  2. #2

    Cooking multiple dishes (1/3) using Human Level Artificial Intelligence

    Published on Jun 2, 2013

    This video shows a robot chef cooking multiple dishes in a restaurant. When customers order dishes in a restaurant, the robot chef has to cook these dishes in the fastest time possible. This includes multi-tasking and making 2 or more dishes simultaneously. This video has no sound because i wanted to show the viewers what the robot is thinking while making multiple dishes in a restaurant. The text flashing on the screen are internal thoughts (sound data) of the robot and are not instruction text for viewers.

    Human beings have the ability to do multiple tasks at the same time. Some people can drive a car, text a friend, eat breakfast and play chess at the same time. However, the more tasks a human has to juggle, the less accurate each task will be accomplished. When working in a restaurant, the robot chef must make multiple meals and he has to do it in an optimal way. When meat has to cook for 4 minutes, the robot won't stand around and wait 4 minutes staring at the meat. He takes that 4 minutes and do other tasks, like making a second dish, or preparing ingredients for the second dish.

    At the beginning of the video, the robot was given an order from a customer sitting at tableC. The order included: (1) a bowl of rice for a family of 4, (2) Chinese lobster dish, and (3) bitter melon dish. The robot's conscious has to first plan out, in his mind, what tasks to do. He determines that the lobster dish is the main course and everyone likes lobsters, so he wanted to serve that dish last. The bitter melon and rice should be served first as an appetizer.

    This video shows the robot managing the cooking of 2 dishes simultaneously. He cooks the bitter melon dish first and when he has free time, he uses that to prepare ingredients for the lobster dish. After the bitter melon dish has been cooked, he devotes all his time to making the lobster dish. This shows that the robot chef is working optimally and he isn't wasting any time doing nothing.

    In real world examples, chefs have to make 3-4 meals at any given moment. As the orders come in the chef is busy every second preparing these dishes. In a very large restaurant with 30-40 chefs, a manager is hired to coordinate their activities and to run the restaurant in an optimal manner.

    Knowledge from cooking school and team work determine what chefs in a restaurant have to do at any given moment. For example, in a Chinese restaurant, some popular dishes are prepared long before a customer makes an order.

    How exactly does the robot learn to multi-task? The robot learns to multi-task through teachers in school. Teachers teach the robot to multi-task. I'm not using any modern day artificial intelligence method, like machine learning or decision trees, This robot learns "everything" from teachers in school.

  3. #3

    Super intelligent robot patent (building the smartest robot in the world)

    Published on Feb 27, 2015

    This invention is very simple. A super intelligent robot has a built in virtual world and he does work inside said virtual world to save time.

    This is how the process works. The super intelligent robot determines a problem it wants to solve and enters the virtual world to do work. as soon as he solves the problem he exits the virtual world and apply that knowledge in the real world. It took human programmers at microsoft 30 years to write an operating system. This super intelligent robot can write an operating system in 1 second inside a virtual world -- He sets up a software company inside the virtual world to do work and in less than 1 second the task is accomplished (read my patent applications for specific details).

    the patent claims include important aspects of the invention such as the universal work algorithm and a system of doing work in both the real world and virtual world. The idea is for these robots to accomplish any "human work" in the "fastest time possible". If it takes human beings 30 years to write an operating system, these robots can accomplish that task in 1 second. If it takes human beings 1 year to build a house, these robots can accomplish that task in 1 week. If it takes human beings 20 years to find a permanent cure for cancer, these robots can accomplish that task in 1 second. Question: why does it take these robots 1 week to build a house in the real world? Answer: because the robots need to work in the real world to build the house. Writing software takes only 1 second because the entire process can be done in the virtual world.

    In reality these super intelligent robots aren't smarter than humans. In fact, a super intelligent robot has equal intelligence to a human. The only difference is that a super intelligent robot has the ability to transport his mind into a virtual world to do work. Humans do not have this ability. Humans can only work in the real world.

    As of its filng, starting on 2007, the patent office determined that my super intelligent robot is new novel and non-obvious. which means there are no pre-existing patents or copyrighted material that have striking similarities to my invention.

    Why is this technology so important? The smart phones today (2015) can't do first grade worksheets yet. In the future, not only can they do elementary school worksheets, but they can do any college assignment. What I present in this video is essentially a computer that can accomplish "any human work" in the "fastest time possible". In the future, this technology will be used on every smartphone, personal computer, electronic devices, or robot. If there are any technology companies out there who are interested in this invention, please email me.

    Background of invention:
    "when i was in college, right before an exam, i would dream about solving math problems. In the dream i would be given a complex problem to solve. At the end of the dream i solved the problem and i wake up." -- Mitchell Kwok

    This idea inspired me to design a super intelligent robot that can predict the future inside a virtual world, and extract specific information from that prediction. In order to predict the future, we have to trick the robots mind in a virtual world to make him believe events are actually happening and work is being done.

    Also, when the robot works in the real world, he can speed up his mind so that objects in the real world are moving slowly. If you combine this with super mobility, the robot can move and think really fast in the real world (kind of like superman).

  4. #4

    Advance Artificial Intelligence (pt 1) with commentary by Mitchell Kwok

    Advance Artificial Intellligence (pt 2) with commentary by Mitchell Kwok

    Published on Apr 8, 2016

    In 2016, no Artificial Intelligence software can play the call of duty. The best so far is Google's artificial intelligence called Deep Q Network. It can only play simple atari games with simple rules and goals.

    I have determined that a robot must have at least a 6th grade level intelligence in order to play the call of duty. All skills in elementary school must be learned by the robot, which includes:

    1. understanding natural language.
    2. doing recursive tasks.
    3. generating common sense knowledge.
    4. making complex decisions.
    5. the ability to do complex math equations.
    6. etc. etc.

    all knowledge must be learned in order for the robot to play the call of duty. for example, if the robot has 2 bullets and surrounded by 5 enemy soldiers, the robot has to do addition and subtraction to generate logical facts. 5 enemy soldiers require 5 bullets. therefore, 5 - 2 equals 3. The robot needs 3 more bullets in order to take out 5 enemy soldiers. The robot must learn facts like: jumping off a building is a bad thing because of gravity. This will allow the robot to make better decisions in the game. All knowledge and skills from elementary school must be learned before the robot can do something complex like play the call of duty.

    My robot, filed in 2006 with the patent office, can play the call of duty. This robot learns information like a human. It uses a bootstrapping process in order to learn information from elementary school to high school and finally college. It can also do complex recursive tasks, which is vital to playing games like the call of duty or Zelda.

Similar Threads

  1. Replies: 4
    Last Post: 3rd December 2016, 09:47
  2. Replies: 5
    Last Post: 8th June 2016, 18:05
  3. Replies: 3
    Last Post: 25th May 2016, 00:19
  4. Mitchell Kwok
    By Airicist in forum Personalities
    Replies: 0
    Last Post: 8th April 2016, 21:52
  5. Replies: 0
    Last Post: 28th July 2014, 12:52

Социальные закладки

Социальные закладки

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts