This robot is playing the piano. I must say that I cannot really hear what it is playing. But don't you think that this is already a very amazing thing that we can find out.
6.30.2007
6.29.2007
Domo Robot Helping with Chores
The MIT CSAIL Humanoid Robot Domo helps a person with everyday chores. Although what it can done is just a simple things, we can see a bright future into the robotics area.
6.28.2007
Baby Robot learn like a human
This robot is just created without giving any intelligence to it except learning technics. From the video clip we can see that this robot is trying to learn by looking at and touching, just like the human baby acting.
6.22.2007
Microsoft Updates Robotics Dev Kit
This new has said that Bill Gates and Microsoft are starting to get involved in robotics in 5th April 2007. Microsoft has offered an upgrade to its recently released Microsoft Robotics Studio (MRS) development kit.
Video of the Microsoft Robotics Studio in Action
Video of the Microsoft Robotics Studio in Action
6.20.2007
Alyuda-Neural Network Software
Alyuda Forecaster is a neural network forecasting software for Windows 98/Me/2000/XP/Vista. It is an extremely easy- to-use and yet powerful tool that allows you to make business and financial forecasts and data analysis using neural network technology.
Its benefits is:
1) save your time - no need to spend your time on learning the neural network theory;
2) automatically prepare your data - your data will be analyzed and converted to make it suitable for a neural network;
3) automatically accomplish the best neural network.
You can download it from http://www.alyuda.com/forecasting-software.htm
It has some features:
1)Reliable forecasting - High quality forecasting and classification due to employment of the latest achievements in artificial neural networks.
2)Exceptional ease of use with Wizard-like interface - Comparing to other neural network packages, Alyuda Forecaster is one step ahead in giving its users comfort and freeing them from necessity to learn details of neural network theory.
3)Forecasting in several clicks
4)Automated network selection - Alyuda Forecaster will help you in selecting the most appropriate model for you forecasting.
5)A lot of available training algorithms and parameters - The most efficient algorithms, like Conjugate Gradient Descent, Levenberg-Marquardt, Quick-Propagation, variations of Back-Propagation, are available for neural network training.
6)Great capabilities of automatic and manual data preparation - You data will be automatically analyzed and transformed to be suitable for neural network.
7)Comprehensive Reporting - Alyuda Forecaster will allow you to customize your reports as well as save them as Excel, HTML or TXT.
Its benefits is:
1) save your time - no need to spend your time on learning the neural network theory;
2) automatically prepare your data - your data will be analyzed and converted to make it suitable for a neural network;
3) automatically accomplish the best neural network.
You can download it from http://www.alyuda.com/forecasting-software.htm
It has some features:
1)Reliable forecasting - High quality forecasting and classification due to employment of the latest achievements in artificial neural networks.
2)Exceptional ease of use with Wizard-like interface - Comparing to other neural network packages, Alyuda Forecaster is one step ahead in giving its users comfort and freeing them from necessity to learn details of neural network theory.
3)Forecasting in several clicks
4)Automated network selection - Alyuda Forecaster will help you in selecting the most appropriate model for you forecasting.
5)A lot of available training algorithms and parameters - The most efficient algorithms, like Conjugate Gradient Descent, Levenberg-Marquardt, Quick-Propagation, variations of Back-Propagation, are available for neural network training.
6)Great capabilities of automatic and manual data preparation - You data will be automatically analyzed and transformed to be suitable for neural network.
7)Comprehensive Reporting - Alyuda Forecaster will allow you to customize your reports as well as save them as Excel, HTML or TXT.
6.18.2007
Fuzzy Logic's advantages
This is fuzzy logic's advantages, I find it from a website.
1. Formalise and simulate the expertise of an operator or designer in process control and tuning
2. Provide a simple answer for processes which are difficult to model
3. Continually take into account cases or exceptions of different kinds, and progressively incorporate them into the expertise
4. Take into account several variables and perform “weighted merging” of influencing into variables
For other Fuzzy Logic questions, you can try this web, http://www.schneider-electric.com/cahier_technique/en/abstracts/fuzzy_logic.htm
1. Formalise and simulate the expertise of an operator or designer in process control and tuning
2. Provide a simple answer for processes which are difficult to model
3. Continually take into account cases or exceptions of different kinds, and progressively incorporate them into the expertise
4. Take into account several variables and perform “weighted merging” of influencing into variables
For other Fuzzy Logic questions, you can try this web, http://www.schneider-electric.com/cahier_technique/en/abstracts/fuzzy_logic.htm
6.16.2007
AI Algorithm Steps
This is a very long article, so I will not talk it in detail. For full article viewing, visit http://mind.sourceforge.net/aisteps.html
But this is a good article, it show us the steps to develop a artificial intelligence algorithm. Or I shall say that the part you need in the artificial intelligence algorithm.
This website also show you some web that contain the code to develop artificial intelligence algorithm.
But this is a good article, it show us the steps to develop a artificial intelligence algorithm. Or I shall say that the part you need in the artificial intelligence algorithm.
This website also show you some web that contain the code to develop artificial intelligence algorithm.
6.14.2007
A Multitasking Machine
STAIR - an artificial intelligence robot, is learning to unload.
This robot that has big eyes and swinging appendages is created by Andrew Ng and the nine other professors and 10 graduate students. He says that creating STAIR is a simple task, the difficult task is let STAIR to learn things by itself. They were delighted when STAIR use only eight months to learn "grasping task", instead of the projected three years.
Now STAIR is learning to unload now. But refer to Ng, it still need some times.
Full articles on http://www.stanfordalumni.org/news/magazine/2007/marapr/farm/news/stair.html
Here is the picture of STAIR
This robot that has big eyes and swinging appendages is created by Andrew Ng and the nine other professors and 10 graduate students. He says that creating STAIR is a simple task, the difficult task is let STAIR to learn things by itself. They were delighted when STAIR use only eight months to learn "grasping task", instead of the projected three years.
Now STAIR is learning to unload now. But refer to Ng, it still need some times.
Full articles on http://www.stanfordalumni.org/news/magazine/2007/marapr/farm/news/stair.html
Here is the picture of STAIR
6.12.2007
Artificial Intelligence in our lives and sex lives
This article is talking about if the robots become too intelligence one day, then we, human will face a lot of problems.
The main problem to think is: if the robots truly have intelligence such as human, can think, can feel, can love, and others, then, can we give them less right just because they are robots? Can human accept it? It took a long timefor mankind to treat people of different colours as equals. Will the history repeat again?
This article also mentione that Gianmarco Verruggio. He is a roboticist at the Institute of Intelligent Systems for Automation in Genoa, northern Italy. He believe that three main rules must be set which artificial intelligence needed to be controlled. First, the society must prevent any kind of illegal use; second, the protection of data in which the intelligence acquires; third, all artificial intelligence must have clear identification and must be traceable anywhere.
After that, the article talks about the sex problem the comes out of the intelligence robots. The author ask if having sex with the robots are better than with human being, what will happen? Should there be a limit as to how good sex with a robot could feel?
Although I don't think all things mentioned above will happens so fast, but yet it still a problem need to be concern. Some firms or stories already try to give answers of the above questions, such as "I, Robot". What do you think about it? In your opinion, if a robot can think, feel, has emotion just like the human, can it be count as a new kind of human? Why yes and why no please?
Source from: http://www.techkills.com/272-the-merging-of-man-and-machine-robots-in-our-lives-and-sex-lives.html
The main problem to think is: if the robots truly have intelligence such as human, can think, can feel, can love, and others, then, can we give them less right just because they are robots? Can human accept it? It took a long timefor mankind to treat people of different colours as equals. Will the history repeat again?
This article also mentione that Gianmarco Verruggio. He is a roboticist at the Institute of Intelligent Systems for Automation in Genoa, northern Italy. He believe that three main rules must be set which artificial intelligence needed to be controlled. First, the society must prevent any kind of illegal use; second, the protection of data in which the intelligence acquires; third, all artificial intelligence must have clear identification and must be traceable anywhere.
After that, the article talks about the sex problem the comes out of the intelligence robots. The author ask if having sex with the robots are better than with human being, what will happen? Should there be a limit as to how good sex with a robot could feel?
Although I don't think all things mentioned above will happens so fast, but yet it still a problem need to be concern. Some firms or stories already try to give answers of the above questions, such as "I, Robot". What do you think about it? In your opinion, if a robot can think, feel, has emotion just like the human, can it be count as a new kind of human? Why yes and why no please?
Source from: http://www.techkills.com/272-the-merging-of-man-and-machine-robots-in-our-lives-and-sex-lives.html
6.09.2007
Introduction of Neural Network
Neural Network, another type of artificial intelligence. I have never have any of this subject before. But today I find an interesting article which introduce the Neural Network.
This is the definition of Neural Network provided by the web: Neural Networks imitate the brain’s ability to sort out patterns and learn from trial and error, discerning and extracting the relationships that underlie the data with which it is presented. The web also say that the Neural Networks is just software that run inside Neural Computer. Each neurons has a few of inputs and produce one output. Every output of the neurons become another input of the other input until the final output has reached.
Then, this article also say about the benifits of the Neural Network. It say that Neural Network is good at providing very fast, very clso approximations of the correct answer. Although they are not as good as conventional computers for performing mathematical calculations or moving the comparing characters, but neural network are expert at recognizing and comparing shapes or patterns, learning of the experience, or sorting relevant data from irrelevant.
To read the whole article, you can visit http://www.bartleby.com/65/ne/neuralnet.html.
This is the definition of Neural Network provided by the web: Neural Networks imitate the brain’s ability to sort out patterns and learn from trial and error, discerning and extracting the relationships that underlie the data with which it is presented. The web also say that the Neural Networks is just software that run inside Neural Computer. Each neurons has a few of inputs and produce one output. Every output of the neurons become another input of the other input until the final output has reached.
Then, this article also say about the benifits of the Neural Network. It say that Neural Network is good at providing very fast, very clso approximations of the correct answer. Although they are not as good as conventional computers for performing mathematical calculations or moving the comparing characters, but neural network are expert at recognizing and comparing shapes or patterns, learning of the experience, or sorting relevant data from irrelevant.
To read the whole article, you can visit http://www.bartleby.com/65/ne/neuralnet.html.
6.08.2007
AI Games On Web
This is just a very simply AI Games Website, somehow it is funny. You can try it in http://www.20q.net/
Before you start the game, you need to choose the language. I try the English and Chinese. The English still got some sub topics. You need to assume yourself as a things refer to the topics, such as "Harry Potter", you need to assume yourself is a place or thing or person in the Harry Potter series. Then, the system will start to ask you question, you shall answer it correctly. The system will ask about 20 questions. Finally the system will guess what you assuming. It will learn from the answer you give and improve its database.
It still not a truly intelligence system, but I think that it can already help us in many place such as data management. What is your opinion?
Before you start the game, you need to choose the language. I try the English and Chinese. The English still got some sub topics. You need to assume yourself as a things refer to the topics, such as "Harry Potter", you need to assume yourself is a place or thing or person in the Harry Potter series. Then, the system will start to ask you question, you shall answer it correctly. The system will ask about 20 questions. Finally the system will guess what you assuming. It will learn from the answer you give and improve its database.
It still not a truly intelligence system, but I think that it can already help us in many place such as data management. What is your opinion?
6.05.2007
Robotics Evolution
I have read this from a book which written in chinese, the book name is the "100 Technologies Which Change The Future". In one sub chapter, some scientist say about the evolution of the robots.
They say that the robots will evolve from one stage to another stage, just like human. In year 2010, the robots will have the intelligence of the lizard. They will have two to six legs, and mostly their job will be cleaning the windows of the tall buildings.
Then, in year 2020, the robots will have the intelligence of the cat or dog, and they can act just like cat or dog. In year 2030, the robots will have the intelligence of the monkey. Besides that, they can understand some conversation of the human but not like now, just act according the database.
Finally, the last stage of the robots will start at year 2040. In this stage, the robots will just like humans. They can understand what human talk, and can think and learn by themselves. They may become stronger than human.
Now is 2007, just 33 more years the world that human live with robots will arrive, what do you think about this?
They say that the robots will evolve from one stage to another stage, just like human. In year 2010, the robots will have the intelligence of the lizard. They will have two to six legs, and mostly their job will be cleaning the windows of the tall buildings.
Then, in year 2020, the robots will have the intelligence of the cat or dog, and they can act just like cat or dog. In year 2030, the robots will have the intelligence of the monkey. Besides that, they can understand some conversation of the human but not like now, just act according the database.
Finally, the last stage of the robots will start at year 2040. In this stage, the robots will just like humans. They can understand what human talk, and can think and learn by themselves. They may become stronger than human.
Now is 2007, just 33 more years the world that human live with robots will arrive, what do you think about this?
6.02.2007
Food Collection Genetic Program
This is a artificial intelligence program. This program will mimic the food collection behaviour of ants. The teamwork aspect of this program comes into play when the ants must work together to solve a problem. The Genetic Program evolver that will create the ant programs, then evolve the programs. It allows problems to be solved without explicitly programming the solution.
Here is the image of the program when the program is running.
To download or get the softcode of this program, visit http://www.lalena.com/AI/Ant/
5.31.2007
Accoona Artificial Intelligence Search Engine
Accoona Artificial Intelligence is a search engine that understands the meaning of search queries beyond the conventional method of matching keywords. The users can either search by using the meaning of words, highlight the keyword, or get information from both the web and the Accona Business Information Database in real time.
I have used this search engine. This is a quite good search engine. I think that guys need to find information can try this website. The only problem is, I cannot adjust the articles' date that I want.
This search engine website is http://www.accoona.com/
5.29.2007
Self-Learning: Just Another Way to Say 'AI'
This article is saying about a self-learning software. It is IBM's new self-learning software. It will automatically categorizes large volumes of information, making it easier to find, access, and use.
Beside that, this self-learning software can automatically learns and interprets the meaning of unstructured text in real-time. This will make content classification more precise. This software understand words, linguistics, semantics, and context of the language as well as associated metadata. It become more accurate over time without requiring human adjustment.
This software can also analyze unstructured and semistrutured content and metadata, allowing previously untapped information to be quickly and precisely classified so it can be more effectively found and reused. This is making the content easier to manage and retrieve, and also reducing the need for manual categorization.
What do you think about this self-learning software? Did you use other self-learning software? What is its function?
More information on http://www.ddj.com/blog/portal/archives/2007/05/selflearning_ju.html
5.25.2007
AI Chatroom Software
Moji Intelligent Messenger (IM), a program features 3D virtual pets that have the ability to understand the context of speech and hold conversations on its own has now come to the market. These pets are essentially chatterbots or chat bots are programs designed to simulate human-like conversations.
The objective that Moji has created is to help the parents to keep their children safe from dangerous characters prowling the Net.
Although there is already certain child-safety programs can tract someone is being outright obscene in chatrooms, but most sexual predators are cunning enough to disguise their advances so an not to raise alarm. However, Moji is different. Mor(f) Dynamics Pty Ltd that create Moji said that Moji pets can 'understand' conversations so that it can detect other people's intentions and determine where the discussion is heading.
When the Moji pets find out the conversation reaches a threatening stage, it immediately sends a warning by e-mail or, in the future, via SMS to the child's parents, the perpetrator will also be blocked as well.
At the end of day, Moji can also e-mail the parents a summary of what their children discuss online without reveling every single detail. Therefore, the parents can know what their children discuss online without spying on their children.
Another good feature of Moji is it able to simultaneously connect to all major instant message services, such as MSn, Yahoo!, Google Talk, ICQ, and AIM. Then the users only need one chat program even if they have multiple accounts.
Moji pets also will take on the characteristics of its user. That means Moji pets will seem like their users. This is similar to the way real-life pets tend to adopt their owner's traits.
Moji pets also will literally live off its owners' words, which means the more you use, the more the pet will feel cared for. If the users stop interacting with it for a few month, it will begin to deteriorate and eventually die. This might teach children to be more responsible for their pets.
This software needs a subscription. Users can download and use Moji for free for the first three month, but will subsequently be charged RM3 monthly if they want to continue using it.
This is a good chatroom software. The parents no need to headache protecting their children from the not suitable conversation in the chatroom. This is also a new step to the AI technologies.
What do you think about this Moji software? Do you think it is good or bad, why?
More information about the company and its products is available at www.morfdynamics.com.
The objective that Moji has created is to help the parents to keep their children safe from dangerous characters prowling the Net.
Although there is already certain child-safety programs can tract someone is being outright obscene in chatrooms, but most sexual predators are cunning enough to disguise their advances so an not to raise alarm. However, Moji is different. Mor(f) Dynamics Pty Ltd that create Moji said that Moji pets can 'understand' conversations so that it can detect other people's intentions and determine where the discussion is heading.
When the Moji pets find out the conversation reaches a threatening stage, it immediately sends a warning by e-mail or, in the future, via SMS to the child's parents, the perpetrator will also be blocked as well.
At the end of day, Moji can also e-mail the parents a summary of what their children discuss online without reveling every single detail. Therefore, the parents can know what their children discuss online without spying on their children.
Another good feature of Moji is it able to simultaneously connect to all major instant message services, such as MSn, Yahoo!, Google Talk, ICQ, and AIM. Then the users only need one chat program even if they have multiple accounts.
Moji pets also will take on the characteristics of its user. That means Moji pets will seem like their users. This is similar to the way real-life pets tend to adopt their owner's traits.
Moji pets also will literally live off its owners' words, which means the more you use, the more the pet will feel cared for. If the users stop interacting with it for a few month, it will begin to deteriorate and eventually die. This might teach children to be more responsible for their pets.
This software needs a subscription. Users can download and use Moji for free for the first three month, but will subsequently be charged RM3 monthly if they want to continue using it.
This is a good chatroom software. The parents no need to headache protecting their children from the not suitable conversation in the chatroom. This is also a new step to the AI technologies.
What do you think about this Moji software? Do you think it is good or bad, why?
More information about the company and its products is available at www.morfdynamics.com.
5.24.2007
I Chat, Therefore I Am......
This artical is trying to figure out if the chatbot, robot created to chat to others, can fool another chatbot, or chat to each others in private. So they arranged the conversation between two chatbots. In this conversation, these two chatbots, ALICE and Jabberwacky asking each others question and they try to answer the question. This is quite a long conversation, and both of these chatbots seem can conversation mostly like a human. You can read their conversation in http://discovermagazine.com/2007/brain/i-chat-therefore-i-am.
So, do you belive that one day in the future, the robots are just like our human, and also be admired as a new kind of human, why yes and why no?
So, do you belive that one day in the future, the robots are just like our human, and also be admired as a new kind of human, why yes and why no?
5.21.2007
Prize announced to prove Turing Machine
This article said that Wolfram Research and Stephen Wolfram announced US $25,000 prize for the first person or group to prove or disprove that the Turing Machine can art as a universal computer. This article also introduce the Turing Machine. It is from http://ai-depot.com/news/story/Wolfram-Announces-Prize-Turing-Machine-Proof/
I check from http://en.wikipedia.org/wiki/Universal_computer, the universal computer is a Turing Machine that able to simulate to another Turing Machine. What I keep thinking is, is this really so important that they want to give US $25,000 to whoever that prove this? What do you think?
I check from http://en.wikipedia.org/wiki/Universal_computer, the universal computer is a Turing Machine that able to simulate to another Turing Machine. What I keep thinking is, is this really so important that they want to give US $25,000 to whoever that prove this? What do you think?
Subscribe to:
Comments (Atom)
