Previous in Forum: PC to PC programme transfer   Next in Forum: Gordon Moore passes away
Close
Close
Close
9 comments
Guru

Join Date: Feb 2007
Location: Israel
Posts: 2966
Good Answers: 24

A chat with an AI, about the subject of AI

12/08/2022 1:21 AM

https://chat.openai.com/chat

https://openai.com/blog/chatgpt/

Is this the chat pad?

Yes, you are using a chat pad to communicate with me. I am a large language model trained by OpenAI and my purpose is to assist you with any questions you may have. My knowledge is based on the information that was available to me at the time of my training, so I may not have the most up-to-date information on a given topic. Is there something specific you would like to know? I'm here to help.

I would like to know if you think machines are able to learn

Yes, machines are able to learn. Machine learning is a subset of artificial intelligence (AI) that involves the use of algorithms and statistical models to enable a system to improve its performance on a specific task over time. Machine learning algorithms allow machines to find patterns in data and make predictions or take actions based on that data. This type of learning can be supervised, in which the machine is provided with labeled examples to learn from, or unsupervised, in which the machine is not provided with labeled examples and must discover the underlying structure of the data on its own.

Are machines able to improvise when tackled with precedent which was not predicted by the initial developer?

Yes, in some cases machines are able to improvise when faced with situations that were not specifically predicted or programmed by their developers. This ability is known as "adaptive learning" and is a key aspect of many machine learning algorithms. In supervised learning, for example, a machine may be trained on a large dataset of labeled examples, and when it encounters a new example that it has not seen before, it can use the patterns it has learned from the training data to make predictions or take actions. In unsupervised learning, a machine may be able to identify patterns in data and use that information to make decisions or take actions in response to new situations.

Pattern Recognition is a shaky ground. In all cases of any pattern isolated from a seemingly random background, a pattern-segment is always suspected of containing a "Pattern-Like Random", not unlike how Random-Number Generators are always suspected of containing a "Random-Like Pattern"

Yes, pattern recognition can be a complex and sometimes uncertain process. In many cases, the patterns that we see in data may not be as clear or distinct as we would like, and it can be difficult to determine whether a given pattern is real or simply a product of random variation. This uncertainty is due, in part, to the fact that even random data can sometimes contain patterns, and it is often difficult to distinguish a true pattern from a pattern-like random.

Would you convene with describing Adaptive Learning via the analogy of "A program which can re-write its source-code while running"?

Adaptive learning can be thought of as a program that can modify its own behavior while it is running, based on the data it is exposed to and the results of its own actions. This is similar to a program that can rewrite its own source code while it is running, allowing it to adjust and improve its performance over time. In machine learning, adaptive learning algorithms are designed to automatically adjust their internal parameters and structure based on the data they are exposed to, enabling them to improve their performance on a given task without the need for human intervention. This allows the machine to adapt and improve its performance over time, making it more effective at solving a wide range of problems.

Would such a program require idle-time while running in order to enable re-writing,debugging and running?

It is not necessarily the case that a program that can rewrite its own source code while it is running would require idle time to do so. Depending on the specific implementation of the program and the algorithm used, it may be possible for the program to modify its own code and continue running without interruption. However, in some cases, the program may need to pause or slow down its execution in order to make the necessary changes and ensure that the modified code is functioning properly. In general, the amount of idle time required for a program to adapt and improve its performance will depend on the specific details of the implementation and the complexity of the problem it is trying to solve.

So, basically, such a program initially contains an innate algorithmic analyzer and coder, as well as "knowledgeable" debugger, ready to test all possible machine-states the program can generate

Yes, that is a good way to think about it. A program that is capable of adapting and improving its own performance would typically include an innate algorithmic analyzer and coder, as well as a knowledgeable debugger, that are able to analyze the program's current state and make modifications to its code as needed. This would allow the program to automatically adjust its behavior and improve its performance on a given task without the need for human intervention. The specific details of how this is implemented will vary depending on the specific algorithm and the problem that the program is trying to solve, but the general idea is that the program has the ability to analyze its own performance and make changes to improve its effectiveness.

Did you just describe an ideal, or is it that such a program is up and running in real life?

I was describing an ideal scenario, in which a program is able to adapt and improve its own performance without the need for human intervention. While this is a powerful concept, it is not yet a reality in most cases. There are many challenges and limitations to developing machine learning algorithms that can truly adapt and improve their performance on their own, and much research is still needed in this area. However, there have been some notable advances in adaptive learning algorithms and techniques, and it is an active area of research in the field of artificial intelligence.

'A mechanism which can re-design itself while negotiating with the real world' - AKA - 'A program which can re-write itself while running' implies a paradox: In the device's life-cycle, there are only two separate possible distinct time-line modes: Idle (during which re-programming and de-bugging are viable), and Running (during which input, process and output are possible). Each mode negates the other while active. That Barrier is Functional, thus each distinct mode of the two is exclusive. It seem like Exclusive Re-Design or Run phases negates the vision of 'A program which re-writes itself while running' is merely an illusive venue of wishful-thinking - isn't it just A fool's dream?

It is true that there are challenges and limitations to developing a program that can rewrite its own code while it is running. As you pointed out, there is a fundamental trade-off between the ability to make changes to the program's code and the ability to process and respond to inputs in real time. In order to make changes to its code, a program must pause its execution and enter an idle state, which means that it is not able to respond to inputs or process data while it is in this state. This can be a significant barrier to the development of truly adaptive programs that can improve their performance on their own.

Try again

Free Research Preview: ChatGPT is optimized for dialogue. Our goal is to make AI systems more natural to interact with, and your feedback will help us improve our systems and make them safer.

Register to Reply
Pathfinder Tags: AI
Interested in this topic? By joining CR4 you can "subscribe" to
this discussion and receive notification when new comments are added.
Guru
United Kingdom - Member - Indeterminate Engineering Fields - Control Engineering - New Member

Join Date: Jan 2007
Location: In the bothy, 7 chains down the line from Dodman's Lane level crossing, in the nation formerly known as Great Britain. Kettle's on.
Posts: 32049
Good Answers: 838
#1

Re: A chat with an AI, about the subject of AI

12/08/2022 3:14 AM

It is getting to the point where a conversation is only valid if it is face to face. It's all about trust.

The telephone call is a close second.

Email is over.

__________________
"Did you get my e-mail?" - "The biggest problem in communication is the illusion that it has taken place" - George Bernard Shaw, 1856
Register to Reply
Guru
Hobbies - CNC - New Member Hobbies - DIY Welding - New Member Engineering Fields - Electromechanical Engineering - New Member

Join Date: Aug 2007
Posts: 23517
Good Answers: 419
#3
In reply to #1

Re: A chat with an AI, about the subject of AI

12/08/2022 9:04 AM

It’s interesting,…

On a personal level, I always enjoyed writing a letter and mailing… the receiver receiving a hand written letter, always felt good (important). And they in turn doing it and enjoyed waiting for a reply.

Of course more pressing matters a phone call or face to face.

On a professional level, I agree with you, with the exception of email. I treat and use Email as a document. It’s a CYA. Same with text messages.

__________________
“ When people get what they want, they are often surprised when they get what they deserve " - James Wood
Register to Reply
Guru

Join Date: Apr 2010
Location: About 4000 miles from the center of the earth (+/-100 mi)
Posts: 9725
Good Answers: 1118
#2

Re: A chat with an AI, about the subject of AI

12/08/2022 7:48 AM

They need to mix in some typos and grammatical errors to make it sure to pass the Turing test. It sounds too intelligent.

Register to Reply
Guru

Join Date: Feb 2007
Location: Israel
Posts: 2966
Good Answers: 24
#4
In reply to #2

Re: A chat with an AI, about the subject of AI

12/08/2022 5:23 PM

It basically 'sounds' like an elaborate, "more educated" version of ELIZA.

IMHO, the ability to pass Turing Test, is nothing like a pointer to a true AI

Register to Reply
Guru

Join Date: Apr 2010
Location: About 4000 miles from the center of the earth (+/-100 mi)
Posts: 9725
Good Answers: 1118
#5
In reply to #4

Re: A chat with an AI, about the subject of AI

12/09/2022 7:44 AM

I have to agree. The Turing test just shows how good the programming is, not whether there is someone inside thinking about things.

Register to Reply
Guru

Join Date: Oct 2006
Location: 44.56024"N 15.307971E
Posts: 7806
Good Answers: 264
#6
In reply to #5

Re: A chat with an AI, about the subject of AI

12/10/2022 10:31 AM

A real test would be to ask it to give a reliable theory of how to understand women.

I have see at least 4,but they do not work.

__________________
"A man never stands so tall as when he stoops to help a child." "Never argue with a stupid person.They will drag you down to their level and beat you with experience" "Homo homini lupus"
Register to Reply
Guru
Technical Fields - Technical Writing - New Member Engineering Fields - Piping Design Engineering - New Member

Join Date: May 2009
Location: Richland, WA, USA
Posts: 21024
Good Answers: 793
#7
In reply to #6

Re: A chat with an AI, about the subject of AI

12/10/2022 10:05 PM

Maybe you just have to bite the bullet and get a sex-change operation. See things from the other side, as it were.

__________________
In vino veritas; in cervisia carmen; in aqua E. coli.
Register to Reply
Anonymous Poster #1
#8
In reply to #7

Re: A chat with an AI, about the subject of AI

12/10/2022 10:57 PM

Register to Reply
Guru

Join Date: Oct 2006
Location: 44.56024"N 15.307971E
Posts: 7806
Good Answers: 264
#9
In reply to #7

Re: A chat with an AI, about the subject of AI

12/11/2022 8:29 AM

I prefer to stay on Mars,my wife prefers Venus,and ne'er the twain shall meet.

__________________
"A man never stands so tall as when he stoops to help a child." "Never argue with a stupid person.They will drag you down to their level and beat you with experience" "Homo homini lupus"
Register to Reply
Register to Reply 9 comments
Interested in this topic? By joining CR4 you can "subscribe" to
this discussion and receive notification when new comments are added.
Copy to Clipboard

Users who posted comments:

Anonymous Poster (1); HiTekRedNek (2); phoenix911 (1); PWSlack (1); Rixter (2); Tornado (1); Yuval (1)

Previous in Forum: PC to PC programme transfer   Next in Forum: Gordon Moore passes away

Advertisement