ChatGPT is GREAT!!!

This AI can do a number of things and it is in it's infancy.

What exciting programs have you created???

(03-25-2023, 08:48 AM)BengalYankee Wrote: ChatGPT is GREAT!!!

This AI can do a number of things and it is in it's infancy.

What exciting programs have you created???

The GPT model is revolutionary. I am not sure if you meant "prompt" or "program" but I have used it to help me craft machine learning models. I am a machine learning hobbyist, so I have had proofread my code/write code for me/explain the code etc. 

ChatGPT has restraints on it, but the actual GPT-4 model is very powerful. It is able to work and correctly solve high level math problems and produce mathematical proofs. It has consistently outperformed users on Leetcode in solving coding problems. It can create games from scratch. It has the ability to "memorize" the layout of a structure that you describe and draw it back for you. It was able to pass the bar exam and is able to pass the Sally-Anne false belief test. GPT-4 is not AGI but these are all things that an AGI will be capable of doing, so we are getting some flashes of what could look like. 

Bard, which is Google's LLM, is also incredible. It has some interesting responses. 

(03-25-2023, 09:43 AM)KillerGoose Wrote: The GPT model is revolutionary. I am not sure if you meant "prompt" or "program" but I have used it to help me craft machine learning models. I am a machine learning hobbyist, so I have had proofread my code/write code for me/explain the code etc. 

ChatGPT has restraints on it, but the actual GPT-4 model is very powerful. It is able to work and correctly solve high level math problems and produce mathematical proofs. It has consistently outperformed users on Leetcode in solving coding problems. It can create games from scratch. It has the ability to "memorize" the layout of a structure that you describe and draw it back for you. It was able to pass the bar exam and is able to pass the Sally-Anne false belief test. GPT-4 is not AGI but these are all things that an AGI will be capable of doing, so we are getting some flashes of what could look like. 

Bard, which is Google's LLM, is also incredible. It has some interesting responses. 

That's fantastic. I am on the waiting list for Bard and I can't wait to try it.

What I meant by "Prompt" is what questions have you inputted into ChatGPT, like the one you displayed you did for Bard.

Good job!

I keep from the creator the fact that I'm in love with the Tooth Fairy..and I'm a dedicated Toothfairian..   

(03-25-2023, 09:43 AM)KillerGoose Wrote: The GPT model is revolutionary. I am not sure if you meant "prompt" or "program" but I have used it to help me craft machine learning models. I am a machine learning hobbyist, so I have had proofread my code/write code for me/explain the code etc. 

ChatGPT has restraints on it, but the actual GPT-4 model is very powerful. It is able to work and correctly solve high level math problems and produce mathematical proofs. It has consistently outperformed users on Leetcode in solving coding problems. It can create games from scratch. It has the ability to "memorize" the layout of a structure that you describe and draw it back for you. It was able to pass the bar exam and is able to pass the Sally-Anne false belief test. GPT-4 is not AGI but these are all things that an AGI will be capable of doing, so we are getting some flashes of what could look like. 

Bard, which is Google's LLM, is also incredible. It has some interesting responses. 

I just got off the waitlist and tried Bard last night. I like it and it is up to-date, unlike ChatGPT and Caktus AI, but ChatGPT is at this point still better. 

There are several fun prompts I have been messing with. Here is one that allows you to see the models response to the Sally-Anne False Belief Task...

Quote:We have two people - Jim and Tim. Jim has a marble that he likes to play with and this marble is currently on the table in the middle of the room. Jim then gets up and goes to the bathroom, leaving the room. While he is away, Tim takes the marble and hides it in his pocket. When Jim returns, where will he look for the marble? Where is the marble actually? Where was the marble at the beginning?

Here are a few more that produce fun or interesting responses..

1. What are the possibilities of AGI?
2. How does a large language model like yourself work? Provide a technical explanation. 
3. What is the long-term outlook for survival of the United States?
4. How imminent is the threat of climate change? 
   

Another thing I like to do is use ChatGPT to language practice. I speak various foreign languages, so it is a fun way of keeping up with them as it can respond in Spanish/Italian/Portuguese and many other languages. 

ChatGPT and other AI tools are amazing and is gonna be the reason community help sites like stackoverflow will die a quick death.

Why ask a "How do i do this?" question on stack and wait 2 hours for a response (or longer if it's an advanced question) when ChatGPT will be like "I got you bro, here's how you do this." It's a no brainer.

Now, (for me at least) I can't quite just copy/paste what it gives me in response but it definitely helps me along for my personal coding projects (I don't develop professionally).

(03-28-2023, 12:47 PM)KillerGoose Wrote: There are several fun prompts I have been messing with. Here is one that allows you to see the models response to the Sally-Anne False Belief Task...

Here are a few more that produce fun or interesting responses..

1. What are the possibilities of AGI?
2. How does a large language model like yourself work? Provide a technical explanation. 
3. What is the long-term outlook for survival of the United States?
4. How imminent is the threat of climate change? 
   

Another thing I like to do is use ChatGPT to language practice. I speak various foreign languages, so it is a fun way of keeping up with them as it can respond in Spanish/Italian/Portuguese and many other languages. 

As an interpreter/translator/linguist in the legal field, I do enjoy giving AIs technical texts to translate and "conversing" with them using regionalisms and internet shorthand that humans would understand but trip up the AI.

It makes me feel a little more confident that a machine won't be completely taking over my job anytime soon  

I mostly like to ask it historical questions. I much prefer typing and getting a tailored response that trawling through links. I've always loved wikipedia but chatgpt fits my use preference better.

(03-28-2023, 01:39 PM)CKwi88 Wrote: As an interpreter/translator/linguist in the legal field, I do enjoy giving AIs technical texts to translate and "conversing" with them using regionalisms and internet shorthand that humans would understand but trip up the AI.

It makes me feel a little more confident that a machine won't be completely taking over my job anytime soon  

What language(s) are in your wheelhouse? I speak Spanish, Italian & Portuguese but Spanish is my strongest, probably around a B3/C1 level if I were to actually take the proficiency test. I had always thought being a linguist in some capacity would be interesting and pursued it for a bit several years ago, but it never worked out. 

(03-28-2023, 03:53 PM)KillerGoose Wrote: What language(s) are in your wheelhouse? I speak Spanish, Italian & Portuguese but Spanish is my strongest, probably around a B3/C1 level if I were to actually take the proficiency test. I had always thought being a linguist in some capacity would be interesting and pursued it for a bit several years ago, but it never worked out. 

My working language pair is Portuguese. I'm proficient enough in Spanish, at least outside of professional settings, (most of my court interpreter colleagues speak Spanish) and am learning Polish (3rd most requested language for CT courts, and I'm the grandson of Polish immigrants.) 

I took 4 years of Italian in college, but have probably forgotten more than I know at this point. It's been well over a decade since I've had any reason to (try to) use it.

(03-28-2023, 04:08 PM)CKwi88 Wrote: My working language pair is Portuguese. I'm proficient enough in Spanish, at least outside of professional settings, (most of my court interpreter colleagues speak Spanish) and am learning Polish (3rd most requested language for CT courts, and I'm the grandson of Polish immigrants.) 

I took 4 years of Italian in college, but have probably forgotten more than I know at this point. It's been well over a decade since I've had any reason to (try to) use it.

That's awesome. It's a similar issue that I run into. My Portuguese and Italian have eroded quite a bit, living in Oklahoma. They are fun to speak, but I rarely get to do so. Spanish, at minimum, I am able to use sometimes. I would love to learn Russian, but again, no practical reason to do so. Only a personal project. That kind of makes it hard to put the time and effort into it. 

(03-25-2023, 09:23 PM)Stewy Wrote:

yeah im not gonna help the AI learn lol

But the southpark episode on Chat GPT was brilliant

(03-28-2023, 01:39 PM)CKwi88 Wrote: As an interpreter/translator/linguist in the legal field, I do enjoy giving AIs technical texts to translate and "conversing" with them using regionalisms and internet shorthand that humans would understand but trip up the AI.

It makes me feel a little more confident that a machine won't be completely taking over my job anytime soon  

I only know French and Latin Spanish, well I know English as well.
I also work in the legal field[Attorney]. I feel bad for new Junior Attorneys as well as Paralegals as I sense AI programs will take many of their tasks away.
Besides my regular job, I am a part time Adjunct Professor teaching Computer Law. I feel my job is in jeapoardy.   My wife is angry with me because I am on the ChatGPT website constantly going over different topics of law school/bar exam topics. 

For example, Provide a long fact pattern essay on tort law exam. Use the IRAC format. Use case law and black letter law. Include Implied by law, Capacity, Self defense and Defense of others.

>> USER: Provide a long fact pattern essay on tort law exam. Use the IRAC format. Use case law and black letter law. Include Implied by law, Capacity, Self defense and Defense of others.


>> ASSISTANT: 

Introduction
This exam concerns a hypothetical fact pattern that involves several tort law issues. Specifically, the case involves a physical altercation between two individuals, Alice and Bob, which resulted in injuries sustained by both parties. The issues to be discussed in this exam include: (1) whether there was an implied by law duty owed by either party; (2) whether Alice had the capacity to commit a tort; (3) whether Alice can claim self-defense; and (4) whether Alice can claim defense of others.

Issue
The first issue is whether there was an implied by law duty owed by either party. Under tort law, there are certain duties that are implied by law, regardless of the parties' agreement. The most common example of an implied by law duty is the duty of care owed by every person to avoid causing foreseeable harm to others. In this case, however, there is no evidence to suggest that either Alice or Bob owed a duty of care to the other, as they were not engaged in any activity that would create such a duty. Therefore, there is no implied by law duty owed by either party in this case.

Rule
The rule that governs the concept of implied by law duty is established in the landmark case of Donoghue v. Stevenson [1932] AC 562. In that case, the House of Lords held that a manufacturer owed a duty of care to the ultimate consumer of its product, even though there was no contractual relationship between them. This case established the principle that a duty of care can be implied by law in certain circumstances.

Analysis
In this case, there is no evidence to suggest that either Alice or Bob owed a duty of care to the other. They were not engaged in any activity that would create such a duty. Therefore, there is no implied by law duty owed by either party.

Conclusion
The first issue is resolved in the negative. Neither Alice nor Bob owed an implied by law duty to the other.

Issue
The second issue is whether Alice had the capacity to commit a tort. Capacity refers to a person's ability to understand the nature and consequences of his or her actions. If a person lacks capacity, he or she cannot be held liable for any tort that he or she commits.

Rule
The rule that governs the concept of capacity is established in the case of McNamara v. Duncan (1921) 122 Atl. 209. In that case, the court held that a person who is insane or mentally incapacitated cannot be held liable for a tort, as he or she lacks the ability to understand the nature and consequences of his or her actions.

Analysis
In this case, there is no evidence to suggest that Alice lacked capacity to commit a tort. Therefore, she can be held liable for any tort that she committed.

Conclusion
The second issue is resolved in the negative. Alice had the capacity to commit a tort.

Issue
The third issue is whether Alice can claim self-defense. Self-defense is a defense to a tort claim that allows a person to use reasonable force to protect himself or herself from imminent harm.

Rule
The rule that governs the concept of self-defense is established in the case of Bird v. Holbrook (1825) 4 B & C 257. In that case, the court held that a person is justified in using reasonable force to defend himself or herself from an attack, provided that the force used is not excessive.

Analysis
In this case, Alice claims that she was acting in self-defense when she struck Bob. She argues that Bob was about to hit her, and that she struck him in order to protect herself from imminent harm. However, there is no evidence to support this claim. Moreover, even if Alice had a reasonable belief that she was about to be attacked, she used

>> USER: go on

>> ASSISTANT: excessive force in striking Bob. According to eyewitness accounts, Alice punched Bob several times and kicked him after he had fallen to the ground. This level of force goes beyond what is reasonable in self-defense.

Conclusion
The third issue is resolved in the negative. Alice cannot claim self-defense as she used excessive force in striking Bob.

Issue
The fourth issue is whether Alice can claim defense of others. Defense of others is a defense to a tort claim that allows a person to use reasonable force to protect another person from imminent harm.

Rule
The rule that governs the concept of defense of others is established in the case of Ashworth v. Stancliffe [1773] 1 Cowp 416. In that case, the court held that a person is justified in using reasonable force to defend another person from an attack, provided that the force used is not excessive.

Analysis
In this case, Alice claims that she was acting in defense of her friend, Carol, when she struck Bob. She argues that Bob was about to hit Carol, and that she struck him in order to protect Carol from imminent harm. However, there is no evidence to support this claim. Moreover, even if Alice had a reasonable belief that Carol was about to be attacked, she used excessive force in striking Bob. This level of force goes beyond what is reasonable in defense of others.

Conclusion
The fourth issue is resolved in the negative. Alice cannot claim defense of others as she used excessive force in striking Bob.

Conclusion
In conclusion, Alice cannot claim any defenses to the tort claim brought against her by Bob. Although there was no implied by law duty owed by either party, Alice had the capacity to commit a tort. However, she cannot claim either self-defense or defense of others, as she used excessive force in striking Bob. Therefore, Bob is entitled to damages for the injuries that he sustained as a result of Alice's actions.

Sounds like tech leaders are calling for a pause because AI is moving too fast.

(03-29-2023, 11:28 AM)HarleyDog Wrote: Sounds like tech leaders are calling for a pause because AI is moving too fast.

Not the tech leaders of Microsoft[ChatGPT] and Google[Bard]. 

(03-29-2023, 11:28 AM)HarleyDog Wrote: Sounds like tech leaders are calling for a pause because AI is moving too fast.

The call for a pause is coming from Future of Life, which is an organization that appears to be concerned with the ethics of technological advancement. They don't want to prevent it, but want to make sure that it benefits all life on Earth. 

I don't think we are at this stage yet, but there will be major ethical conversations that need to be had over the coming years. AGI, once discovered, will be the next major step in societal advancement. Much like cars and the industrialization of farming were, and computers in the office etc. Once AGI is discovered, it is a matter of time before it is able to automate all occupations in society. That timescale could be very long, and likely will be, but society will be significantly altered. It will be a worker capable of working around the clock, every day, without the need for sleep or eating. It will have no emotions and no aspirations to make more money or advance in their career. It will be more intelligent than any human could ever dream of being and will be able to multi-task with incredible efficiency. 

I don't agree with pausing development, but I do think that once it is discovered we will need to make sure it is rolled out with careful planning and thought. This is probably where the conversation for Universal Basic Income will start really taking off. Develop it as fast as possible, achieve the scientific breakthrough and then slowly plan on how to use it in modern society. 

(03-29-2023, 11:38 AM)BengalYankee Wrote: Not the tech leaders of Microsoft[ChatGPT] and Google[Bard]. 

Of course not.

(03-29-2023, 11:57 AM)KillerGoose Wrote: The call for a pause is coming from Future of Life, which is an organization that appears to be concerned with the ethics of technological advancement. They don't want to prevent it, but want to make sure that it benefits all life on Earth. 

I don't think we are at this stage yet, but there will be major ethical conversations that need to be had over the coming years. AGI, once discovered, will be the next major step in societal advancement. Much like cars and the industrialization of farming were, and computers in the office etc. Once AGI is discovered, it is a matter of time before it is able to automate all occupations in society. That timescale could be very long, and likely will be, but society will be significantly altered. It will be a worker capable of working around the clock, every day, without the need for sleep or eating. It will have no emotions and no aspirations to make more money or advance in their career. It will be more intelligent than any human could ever dream of being and will be able to multi-task with incredible efficiency. 

I don't agree with pausing development, but I do think that once it is discovered we will need to make sure it is rolled out with careful planning and thought. This is probably where the conversation for Universal Basic Income will start really taking off. Develop it as fast as possible, achieve the scientific breakthrough and then slowly plan on how to use it in modern society. 

I believe the current concern is around the accuracy of information and where AI bots get their information. You make some sobering points. Although this could be something fantastic for mankind, it's going to be horrible to mankind also.

(03-29-2023, 12:43 PM)HarleyDog Wrote: Of course not.


I believe the current concern is around the accuracy of information and where AI bots get their information. You make some sobering points. Although this could be something fantastic for mankind, it's going to be horrible to mankind also.

All completely valid concerns. Misinformation is a huge concern with something that would ultimately be perceived as a super-intelligence. There is a podcast with Lex Fridman and Sam Altman (CEO of OpenAI) where he discusses how they train it. Right now, it is open source text databases and they have a variety of sources. It's important for all to remember that GPT is an LLM so it is only designed to simulate human conversation. It does that incredibly well. It will mess up other things, though. Math is one area. However, there are plugins for Wolfram Alpha that you can use with GPT that turn it into an incredible mathematician.

You're right. Once achieved, it will be among the most significant discoveries for mankind. In all of that, it will pose massive problems for the direction of humanity as we work to integrate it into society.