 |
Frequently Asked Questions
|
Artificial Intelligence
 What is artificial intelligence?
What would be the benefit for the society to have an artificial intelligence system?
Is this a different kind of intelligence?
Computers and software programs are not intelligent so is it possible to build an intelligent system using computers?
Can we implement consciousness in a computer?
Do we need to have conscious machines?
How intelligent is the system going to be?
Project Information
Why isn't this project entirely public?
Why isn't this project entirely commercial?
Is this a distributed computing project like Seti@Home or Folding@Home?
Is there an alternative site for this project?
Do you expect to garner a lot of support for this project?
Do you also expect opposition to this project from various groups of people?
Which platforms are going to be supported?
System Security
If the system will eventually be smarter than you, its creator, wouldn't that pose a risk?
What kind of threats will this technology pose?
Is it possible for the artificial intelligence system to modify its behavior if it has consciousness implemented?
What about user security?
Related Work
Are there any other organizations trying to build AI systems?
Why is no other company trying to build a strong AI system?
Who is the project leader?
There is not enough computing power for AI and the Moore's Law is coming to an end, so is AI possible?
Technical Information
What are your research findings?
What are your research approaches?
What is the future system capacity?
What are the challenges involved in building such a system?
Why not use an existing neural simulator?
Is the BOINC distributed computing framework the right choice for this project?
Why not use an existing database system?
Why not use an existing graphic package?
Why not use an existing assertion database?
Will this project make public some of its source code?
Will this project describe how knowledge is represented in the brain?
Will the program that can be downloaded by everyone contain the mechanisms of knowledge representation in the brain?
By providing an executable program would someone be able to reverse engineer the program and understand its source code?
If someone would have the program source code would they be able to find out how knowledge is represented in the brain?
Why not make public the mechanisms of knowledge representation in the brain?
Can I donate time for this project while running in parallel another distributed computing project?
Where can I find more information about the brain and Neuroscience?
Help
What is the best way to help?
How can I get involved in this project?
|
|
Artificial Intelligence
What is artificial intelligence?
Artificial Intelligence is machine intelligence. Machines that have the same knowledge and understanding as a human being. It is also a scientific field that sits at the crossroads of other fields.
What would be the benefit for the society to have an artificial intelligence system?
An AI system would be able to integrate knowledge and information from many fields and sources and it can help us in ways that we can only imagine now. It will think faster and will have a deeper understanding of the nature of things than probably any person alive. It will be able to accelerate scientific discovery at an astonishing rate. It will allow us to live better, healthier and longer lives. It can help us take critical decisions for economy by integrating knowledge from multiple sources. It can lead to reduced work hours for a vast majority of people. It can help eliminate directly or indirectly most of the problems that are plaguing our world, like hunger, disease, poverty, criminality, war.
Is this a different kind of intelligence?
No. Intelligence is implemented in the same way across all biological species, from insects to humans. This system will be no different in this respect; it will use the same information processing mechanisms found in all species. Understanding the mechanisms of intelligence allows us to control and increase machine knowledge and understanding beyond human levels. Nonetheless if a person will have the same information in one field as a machine, that person will be able to reach similar conclusions with the machine.
Computers and software programs are not intelligent so is it possible to build an intelligent system using computers?
Yes it is. The computer is a Universal Turing machine. Simply put that means the computer is a universal device that can simulate any other machine, including the brain, which is an intelligent device. The Artificial Intelligence System will be a large neuronal program that simulates brain processes. The brain is also an Equivalent Universal Turing machine and that means it can simulate any other machine, including the computer and its software programs.
Can we implement consciousness in a computer?
Yes. Consciousness is a state of mind, inside the brain. By simulating on a computer the same brain processes that raise consciousness we will obtain the same result, that is a conscious machine.
Do we need to have conscious machines?
Some attributes of intelligence will require that to be implemented, others not. Our needs as a society will decide upon their use. We are of the opinion that it is too early to debate that.
How intelligent is the system going to be?
Our final goal is to load into the system, scientific papers from many fields, like mathematics, physics, chemistry and so on, so forth. We cannot estimate how intelligent is the system going to be. Most people excel in only one field. If the system will integrate knowledge from multiple fields and will excel in all these fields, we can only say that it would probably be impressive in regards to intelligence.
|
top  |
|
Project Information
Why isn't this project entirely public?
The software development portion of the project can probably be done almost in its entirety with the help of the public at large. The neuroscience portion can not. We also need to continue the research and involve other researchers and research groups that can contribute to our project and its final goal. That requires funding and commercialization is the only acceptable way for us to do that.
Why isn't this project entirely commercial?
This project stands out as one of the most challenging tasks that is currently faced by the humanity. It requires a lot of resources, both financial and human resources. Our conservative estimates places the number of people required to be involved, in the thousands. Very few companies have the expertise, research results, and resources to work on such a task. Once the system will increase its capabilities it will take over some development tasks as well. From that point on, we expect the number of people required to help to drop significantly. Nevertheless it is still a long way to go.
Is this a distributed computing project like Seti@Home or Folding@Home?
Yes, but it has some notable differences, mainly that it will not be used for intensive computations only but also as a large neuronal simulator.
Is there an alternative site for this project?
Yes. We will also post source code on SourceForge.net. The link is http://sourceforge.net/projects/aisystem/. The main project site is www.intelligencerealm.com/aisystem. We encourage you to download the project files from SourceForge.net. We do not intend to make all source code public especially code that pertains to overall system security. Because we are building this project with public support, we intend to make available information and results that can be reused by the public at large.
Do you expect to garner a lot of support for this project?
Yes. The feedback that we already received points to strong future support from people of various backgrounds and from different countries. There are many people that understand the utility of such a project.
Do you also expect opposition to this project from various groups of people?
Yes, we do. This project will challenge many people's understanding, beliefs and interests; these interests can be financial, political or of other nature. Opposition is to be expected. That's why it is important for everyone to understand what this technology is about and how it can best be used for everyone's benefit. Nonetheless we will continue the pursuit of truth; wherever it will take us.
Which platforms are going to be supported?
We are looking forward to build versions for all operating systems and platforms. That will take some time though.
|
| top |
|
System Security
If the system will eventually be smarter than you, its creator, wouldn't that pose a risk?
It sure does. We understand the negative and positive implications of building an Artificial Intelligence system. That's why we have already restricted access and we will implement multiple levels of control and monitoring.
What kind of threats will this technology pose?
The short-term threat stems from everybody's lack of understanding of the mechanisms of knowledge representation in the brain. If a group of people has access to this knowledge while everyone else doesn't, they would be capable of building an intelligent virus program and wipe the computing infrastructure of any country. An AI system has the logic incorporated into data and not the source code thus making it impossible to understand what exactly are the program goals without understanding its rules. That can make the asymmetric threat a tragic reality for all of us. The long-term threat stems from the AI system itself. By having an intelligent system that is very smart and capable, if the system will be used as a weapon, we will all have to deal with a machine intelligence that can be out of control and capable of destructing all of us. That is why is so important to have strict control over the implementation of consciouss machines and to have AI built for peaceful purposes only. We can all benefit from the use of safe artificial intelligence as a technology.
Is it possible for the artificial intelligence system to modify its behavior if it has consciousness implemented?
Yes, it is. Unlike traditional programs that have the logic inside the source code the logic of the AI system is going to be in the data portion of the program. If the program will have the capability of taking decisions it only needs to add more data and have its behavior changed. There is no need for it to change its underlying source code of the neural network simulator. This is very easy to implement and as a result of that it can lead to system that can have out of control behaviour. We will ensure that we have in place mechanisms that can prevent that.
What about user security?
We are concerned about that and we will implement mechanisms and protocols to prevent any security issues.
|
| top |
|
Related Work
Are there any other organizations trying to build AI systems?
Yes. There are some companies that are pursuing this goal but their research results doesn't allow them to pursue the development of a strong AI system. They are building only narrow AI systems. These systems share some properties with intelligence but their systems are nonetheless incomplete. These systems don't account for all intelligence traits.
Why is no other company trying to build a strong AI system?
Are their research results financially sustainable so that they can pursue such a Herculean task?
Who is the project leader?
Ovidiu Anghelidi.
There is not enough computing power for AI and the Moore's Law is coming to an end, so is AI possible?
AI will soon become a reality. It is true that Moore's law will soon no longer be a law. If millions of computers in the world will have our program installed we would have enough power to run a comparable human brain artificial intelligence.
|
| top |
|
Technical Information
What are your research findings?
We found the mechanisms of knowledge representation in the brain which is equivalent to finding artificial intelligence. Our results allow us to move from research into development. Nonetheless we intend to continue the research, but we will also act on the opportunity of being able to build the first Artificial Intelligence System.
What are your research approaches?
We favor a multi-disciplinary approach. Our main focus is neuroscience research but we are also exploring directions in artificial intelligence and computer science.
What is the future system capacity?
We intend to build a system that has the brain capacity of all people living in the world. This way we can ensure that no one will be left behind. Its capacity will go far beyond the petabyte scale.
What are the challenges involved in building such a system?
The human brain has an estimated 100 billion neurons. The number of neurons of all the people in the world is 7 billion times that. Designing such a complex software system requires us to take into account scalability issues right from the start, in order to avoid future problems. We will need to build a massive network of detailed neurons with biophysical properties. We would need a complex neural network simulator. We will need a generator program that builds artificial neurons based on specific properties. We also need a visualization system to be able to monitor and control the system. Our conservative estimates is that perhaps thousands of people would be needed at some point in time in order to build that. This figure will go up from 2 people, right now, as we progress, until the moment when we will be able to have the machine itself take over some tasks. From that point on, the number of people required to help would go down. Nonetheless there are very few companies capable of such effort to date. We need the public support to help us build that.
Why not use an existing neural simulator?
The first reason pertains to licensing issues and the second to scalability. None of the current neural simulators are pushing for the system's future capacity.
Is the BOINC distributed computing framework the right choice for this project?
BOINC is being used by many users all over the globe, thus allowing us to tap into a vast pool of computing resources. It is already tested and implemented on a variety of platforms. The BOINC framework is suitable for computing intensive tasks. The project requires that, but there are also many other tasks that don't fit into this framework. We will have to build additional programs to cover these requirements.
Why not use an existing database system?
Commercial database management systems require administration. We will end up having a database installed on each user's machine. We can't have the users perform administrative tasks nor we can manage it remotely. We intend to use an embedded database library that has a small footprint and doesn't need administration.
Why not use an existing graphic package?
There are no existing graphic packages that can allow us to view the data the way it is supposed to. The system will need special algorithms to deal with this complexity. We will use OpenGL for the visualization program.
Why not use an existing assertion database?
The first reason pertains to licensing issues and the second to content. We will need to create an assertion database so that we can have the system taught using these assertions. We do not need to build extensive assertions because the system will be able to pick them up once it will have the learning capability up and running. We do no need to start with a list of assertions.
Will this project make public some of its source code?
Yes.
Will this project describe how knowledge is represented in the brain?
No.
Will the program that can be downloaded by everyone contain the mechanisms of knowledge representation in the brain?
Yes.
By providing an executable program would someone be able to reverse engineer the program and understand its source code?
Yes. People will probably try to reverse engineer the project's program regardless if the license prohibits that or not.
If someone would have the program source code would they be able to find out how knowledge is represented in the brain?
Probably not. They won't be able to understand the knowledge representation mechanisms even if they have the entire source code. They will be in the same situtation that neuroscientists are now, where they have access to a lot of biological information, yet they are unable to understand how knowledge is represented in the brain. The fact that the software program is a neural simulator doesn't change the nature of the problem. It is only translated from one system (i.e. the brain) to another (i.e. the computer); having a computer that simulates brain processes doesn't solve this problem.
Why not make public the mechanisms of knowledge representation in the brain?
Security is our main concern. As long as this concern is not addressed, it is in everyone's best interest that certain knowledge is restricted.
Can I donate time for this project while running in parallel another distributed computing project?
Yes. You can either run the program as a standalone application or as a screensaver. It is your choice. The only interaction with other projects is in the amount of computing power allocated for each of the projects.
Where can I find more information about the brain and Neuroscience?
We have created a separate site section called Learning Center where you can find additional information.
|
| top |
|
Help
What is the best way to help?
Educate yourself about what artificial intelligence is and how it can best be used in everyone's benefit. The implementation of the artificial intelligence system will end up affecting everyone and will positively change the core fabric of our society. If you are informed you will be able to understand what it offers and how you can use it.
How can I get involved in this project?
You can volunteer to donate computer time for this project. You will need to install a program on your machine that will perform work for this project whenever the computer is idle and it is not being used.
|
| top |
|
|
|
