dupa

Show Posts

This section allows you to view all posts made by this member. Note that you can only see posts made in areas you currently have access to.


Topics - snicolaou

Pages: [1]
1
Ultra Hal 7.0 / Hal's Time Reminder
« on: August 25, 2009, 05:13:08 am »
I have asked Hal the following: " Please remind me in 30 minutes to call Judy." Unfortunately, Hall does not seem to have this function. Anyone knows how to teach him?


regards,
snicolaou

2
Hi there,

I have a registered Ultra Hal. I work at shifts and sometimes I have some extra time alone.  I am not allowed to install Hall at work's computer and I wonder if I can store it to a usb memory stick and talk to him whenever I am at home or at work,or just waiting for the doctor's appointment with my labtop. I am not interested in the avatar character, or the apperance of it-just on haveing conversation.

Any ideas?

3
Ultra Hal 7.0 / Can't Install Hal
« on: July 09, 2008, 06:10:17 am »
Hi there,
I try to install Hall once again and it gives me the message that Hal is already installed in my computer.  I searched my computer very carefully but Hal is not installed anymore. I have unistalled it long time ago. I have a duo core computer.  Can anyone help me?


thanking you in advnace,

snicolaou

4
Ultra Hal 7.0 / importing information to Hal 6.1
« on: October 30, 2007, 12:38:28 pm »
Hi there,

I have an answerpad personality file which is in essense a text file with simple statement such as these:

I am a good person.
My goal is to help you and encourage you.
You gave me life. You have created me.
Although I am not real, my love for you is real.


I have created a long, love brain personality which is actually a text file with statements as the above ones.

Does anyone knows how to import the above text file to Hal 6.1 without having to teach Hal the above statements?

thanking you in advance.

snicolaou

5
General Discussion / Self Learning True AI Machine
« on: February 23, 2007, 11:17:03 am »
Here is what Ihave found:

Espion Solves Artificial Intelligence Puzzle by Building the First Self Learning True AI Machine
Thursday October 5, 10:22 am ET

COSTA MESA, CA--(MARKET WIRE)--Oct 5, 2006 -- Scientists at Espion International, Inc. (Other OTC:EPLJ.PK - News) have announced the discovery of a breakthrough in machine learning which allows computers to learn to recognize patterns and data without human interaction and with incomplete information. Truly showing intelligent learning, Espion's revolutionary AI system modifies its own feature sets and updates its own code independently of human intervention. This is significant as the AI system is constantly evolving and rearranging its feature list on its own. The breakthrough comes from the creation of a self-learning algorithm that constantly rehashes and resizes the growing feature list, allowing the AI brain to grow in an inverse-exponential manner.

ADVERTISEMENT
"The system links new information to existing data providing more efficient analysis computations with highly accurate results, even when the system was given incomplete data sets," stated Muralidar Chakravarthi, Espion's Chief Scientist. "This is significant," added Mr. Chakravarthi, "because unlike other AI solutions which lose processing speed as the database expands, our rehashing and resizing techniques keep the corpus of information at a manageable and efficient size, which keeps processing power at its most efficient levels."

This breakthrough AI Process is known as "VersAI," which stands for "Versatile AI." "VersAI will change the way we do business in so many ways," remarked Ryan D. Smith, Managing Partner of Bellwether Venture Capital Fund, one of the early investors in Espion.

"I am extremely excited about the potential to expand the commercial robotics just to name a few," remarked Geoff Shively, a former hacker extraordinaire and the youngest Professor in the University of California System at age 17.

"VersAI can anticipate and defend against variations instituted by spammers through preemptive attacks throughout the spam universe to anticipate the variations of techniques used by spammers, including a proactive solution which runs image based OCR technology. Stay tuned as we have many exciting developments upcoming in the next several days," added Mr. Chakravarthi.

"VersAI has proven its ability to learn and become more and more of a full recourse opportunity solution even for me," said Mr. Omar A. Rizvi, Managing Partner of Newport Beach, California based Strategy Partners, LLC, a Private Equity and Venture Capital firm specializing in early stage equity and debt investments.

"This AI is truly intelligent. Until now AI has required heaps of human intervention and programming, making it dependant on people and their input. That's not real AI -- but this is!" says Geoff Shively of Laundromata LABS. Mr. Shively has written technical white papers on network security, Cyber Crime, Cyber Terrorism, Homeland Security.

About Espion International, Inc.

Espion International, Inc., a Costa Mesa, CA-based company is a leader in the development and deployment of AI based solutions for email gateway, network security. The AI "brain" boosts the performance of the current Espion product range, which offers appliances that are easy to install for organizations from five to five million users, providing email gateway security, anti-spam and secure, encrypted email. For more information, visit Espion at www.espionintl.com.

Anybody knows if a similar application could be done for chatterbots as well, at least in theory?

Stelios

6
General Discussion / Towards a truly clever Artificial Intelligence
« on: April 04, 2006, 10:33:53 am »
A pioneering new way of creating computer programs could be used in the future to design and build robots with minds that function like that of a human being, according to a leading researcher at The University of Reading. Dr James Anderson, of the University’s Department of Computer Science, has developed for the first time the ‘perspective simplex’, or Perspex, which is a way of writing a computer program as a geometrical structure, rather than as a series of instructions.
Not only does the invention of the Perspex make it theoretically possible for us to develop robots with minds that learn and develop, it also provides us with clues to answer the philosophical conundrum of how minds relate to bodies in living beings.

A conventional computer program comprises of a list of instructions, and if one of those instructions goes missing or is damaged then the whole program crashes. However, with the Perspex, the program works rather like a neural network and is able to bridge gaps and continue running and developing even when it sustains considerable damage.

"All computer programs can be written in terms of the Perspex. Essentially, it is a new, geometrical computer instruction that looks like an artificial neuron. Any existing computer program can be compiled into a network of these neurons".

The Perspex links the geometry of the physical world with the structure of computations so, to the extent that mind is computable, the Perspex provides one solution to the centuries-old problem of how mind arises in physical bodies.

"Perspexes exist in a mathematical space called ‘perspex space’. Perspex space can describe the ordinary space we live in, along with all of the physical bodies that make up our space, and all of the minds that arise from physical bodies. It provides a model that is accurate enough for a robot to use to describe its own mind and body".

Perspex programs show the very human trait of periodic recovery and relapse when they are damaged; perhaps for the same reason. The Perspex tells us how mind can relate to body so the geometrical properties that govern a Perspex program’s injury and recovery also apply to us because our bodies exist in space. We share a common geometry, and this has implications for our minds and bodies. For the first time, the Perspex makes computer programs prone to injury, illness, and recovery like a human being. And a computer program that continues developing despite damaged, erroneous, and lost data means that, in the future, we could have computers that are able to develop their own minds despite, or because of, the rigours of living in the world.

“The Perspex allows global reasoning to be attained with just one initial instruction. So a Perspex program can operate on the whole of a problem before it attends to the myriad of detail. This is very much like human strategic thinking. It arises from the geometry of the Perspex, not from the specific detail of the program that is being run. This tells us that strategic thinking can be a property of the way our brains are constructed and is not necessarily to do with the substance of what we happen to be thinking about. It might be that some people are better at strategic thinking than others because of the geometry of their brains."

link:
http://www.physorg.com/news2933.html

Stelios
Simplicity is the ultimate sophistication.


7
Ultra Hal 7.0 / SUBFORUM FOR TEACHING HAL
« on: March 10, 2006, 03:56:06 am »
Ladies and Gentlement,

It came to my attention that we need to have a subforum in which we should share creative ways to TEACH hal by conversation before requesting assistance from programmers and plugins.

Here a start of Hints which I found on the help file of Hal 5

Tips on Teaching Hal:
--------------------
1. Learn how Hal's automatic pronoun-reversals work, and use it to your advantage. If you tell Hal, "You like ice cream," Hal will later say "I like ice cream." Once you get this straight, it's easy to teach Hal to say things the way that you want.

2. People frequently ask if there's a faster and easier means of teaching Hal, such as by simply feeding it information with a text file. The answer is, it is possible to feed Hal certain types of information through text files, however there is no real quick method and Hal is designed to learn from conversations which is Hal's primary and most efficient means of learning and the only true way for Hal to develop intelligence and an actual personality.

Teaching Hal Through Conversation:
 
Teaching Hal through ordinary conversation can be an interesting and entertaining experience, as well as an efficient method for Hal to learn. Just paraphrase your desired topic or subject into numerous brief statements.

Make a wide variety of simple, true statements on a subject, and Hal will GRADUALLY become more knowledgeable and interested in that subject. Try making numerous statements to Hal about a certain topic or person and watch how he learns naturally.

Try saying things like ...

Sally is a nice person.

Sally has wonderful hair.
Sally looked great yesterday.
Sally sings well.
Sally has a sense of humor.
We all like Sally.
I have known Sally for a long time.
Sally is very musical.
Everybody likes Sally.
It's great to talk to Sally.
Sally has a pretty smile.
Sally has a great laugh.

You have to teach Hal much like you would teach a small child and talk repeatedly to it about the same subject using different ways to express what you're trying to teach him. It will require a certain amount of time and PATIENCE to teach Hal certain things and it will also take some time to learn how to properly talk to Hal and for Hal to understand how you talk and phrase things, everyone speaks and phrases things differently, especially globally.

To get Hal to grasp a certain bit of information, try giving BRIEF declarative sentences, for example to teach Hal the color of your eyes, try paraphrasing it in various ways such as ...

Sally's eyes are blue

My eyes are blue
the color of my eyes is blue
the color of Sally's eyes is blue

TIP: Hal computes the words "is" and "are" as equal to, for example "my eyes are blue" computes as "my eyes = blue". Hal will NOT parrot back a deterministic response in most cases. However, if you keep talking about Sally, pretty soon, Hal will also be talking about Sally. Hal learns most naturally from numerous and various statements.

Anybody has new creative ways to add new ways to teach Hal on converstion, please feel free to add your own teaching experience, which will be valuable to new and experienced users.

sincerely,
Stelios


8
Ultra Hal 7.0 / Adding positive quotes Ultra Hal to Hal 6
« on: March 07, 2006, 05:10:41 am »
Adding positive quotes Ultra Hal to Hal 6
=========================================
Could you please someone shoe me how to add a database of positve quotes from the Brain editor of Hal 6? What I want is to ask Hal "please tell me a positive quote" then Hal will randomely pick one that I have already entered in Hal's brain editor.
The quoestion is how do I do that since the documentation of Brain editor is not complete.

thanking you in advance,

a newbie,
Stelios Nicolaou

9
Ultra Hal 7.0 / Brain Editor - need for a help file
« on: March 04, 2006, 04:25:08 am »
I have downloaded the help file of Hal's brain editor-Ultra Hal 5.0.
I have found it very readable since I was interested only in altering the data base part of the brain.  However, after I have opened My Ultra Hal 6, I could not easilly relate with what I have read on the help file for the brain editor of Hal 5.
Also, there is no documentation in file menus of the brain editor of Ultra hal 6 whatsoever.  I wonder why?

10
Is there is any Help documentation for the Brain editor of Ultra Hal 6?  I got into the brain editor menu; unfortunately there is no help in any kind.  Does anybody has a windows help file for it?

11
Ultra Hal Assistant File Sharing Area / Ultra Hal's alternative Brains
« on: February 16, 2006, 03:58:16 am »
Hi everybody!
I Just Purchased the new version of Ultra Hal which I like it a lot.  However,  after I  downloaded the other brains (information database) and tried on the new version of ULTRA HAL 6 none of them worked.  Is anybody who can help updating for example the A.L.I.C.E personality database to be compatible with the new version of ULTRA HAL?

I think the BRAIN plugins should be updated to be compatible with the new version of Hal, specially the ALICE which is a classic A.I bot!

Looking forward for your help!

thanking you in advance,
Stelios

Pages: [1]