My ChatGPT / AI conversation

[Jkca1]

I logged into one of the sites that allows you to "chat" with a version of AI. I spent a lot of time there asking different types of questions. The AI program I was working with has constraints built into it, but I found the results of my questions very interesting. Here is an example from one of our chats;

ME: How do we fix the homeless problem?

Homelessness is a complex issue that can be caused by a variety of factors, such as poverty, lack of affordable housing, mental health and addiction issues, and a lack of support services. As a result, there is no single solution that can effectively address the problem. Rather, a comprehensive, multi-faceted approach is needed.

One solution is to increase the availability of affordable housing. This can be done through the construction of new housing units, the preservation of existing affordable housing, and the implementation of rent control measures. Additionally, providing financial assistance and vouchers to low-income individuals and families can help them afford housing.

Another solution is to provide support services to individuals experiencing homelessness, such as mental health and addiction services, job training and employment assistance, and healthcare. By addressing the underlying issues that contribute to homelessness, these services can help individuals stabilize their lives and move into permanent housing.

There are also various programs that provide temporary or emergency housing, such as shelters, which can provide a safe and secure place for individuals experiencing homelessness to stay while they work to find permanent housing.

Additionally, policy changes that increase the minimum wage and improve the social safety net can also help to prevent homelessness by making it easier for low-income individuals and families to afford housing and other basic necessities.

It's worth noting that these solutions are not mutually exclusive, and different communities might have different needs and require different approaches. Also, it is important to have the
participation and collaboration of different sectors, including government, private, and non-profit organizations to address the homelessness problem.

 

[bongsaway]

I logged into one of the sites that allows you to "chat" with a version of AI. I spent a lot of time there asking different types of questions. The AI program I was working with has constraints built into it, but I found the results of my questions very interesting. Here is an example from one of our chats;

ME: How do we fix the homeless problem?

Homelessness is a complex issue that can be caused by a variety of factors, such as poverty, lack of affordable housing, mental health and addiction issues, and a lack of support services. As a result, there is no single solution that can effectively address the problem. Rather, a comprehensive, multi-faceted approach is needed.

One solution is to increase the availability of affordable housing. This can be done through the construction of new housing units, the preservation of existing affordable housing, and the implementation of rent control measures. Additionally, providing financial assistance and vouchers to low-income individuals and families can help them afford housing.

Another solution is to provide support services to individuals experiencing homelessness, such as mental health and addiction services, job training and employment assistance, and healthcare. By addressing the underlying issues that contribute to homelessness, these services can help individuals stabilize their lives and move into permanent housing.

There are also various programs that provide temporary or emergency housing, such as shelters, which can provide a safe and secure place for individuals experiencing homelessness to stay while they work to find permanent housing.

Additionally, policy changes that increase the minimum wage and improve the social safety net can also help to prevent homelessness by making it easier for low-income individuals and families to afford housing and other basic necessities.

It's worth noting that these solutions are not mutually exclusive, and different communities might have different needs and require different approaches. Also, it is important to have the
participation and collaboration of different sectors, including government, private, and non-profit organizations to address the homelessness problem.

There's only one problem, that would mean helping the least of us and our society absolutely sucks at it. Money before people and the poor don't qualify for the money it would take for us to act like a caring civilization. Why spend the money on them when we can spend it on war making materials?

 

[eohrnberger]

I logged into one of the sites that allows you to "chat" with a version of AI. I spent a lot of time there asking different types of questions. The AI program I was working with has constraints built into it, but I found the results of my questions very interesting. Here is an example from one of our chats;

ME: How do we fix the homeless problem?

Homelessness is a complex issue that can be caused by a variety of factors, such as poverty, lack of affordable housing, mental health and addiction issues, and a lack of support services. As a result, there is no single solution that can effectively address the problem. Rather, a comprehensive, multi-faceted approach is needed.

One solution is to increase the availability of affordable housing. This can be done through the construction of new housing units, the preservation of existing affordable housing, and the implementation of rent control measures. Additionally, providing financial assistance and vouchers to low-income individuals and families can help them afford housing.

Another solution is to provide support services to individuals experiencing homelessness, such as mental health and addiction services, job training and employment assistance, and healthcare. By addressing the underlying issues that contribute to homelessness, these services can help individuals stabilize their lives and move into permanent housing.

There are also various programs that provide temporary or emergency housing, such as shelters, which can provide a safe and secure place for individuals experiencing homelessness to stay while they work to find permanent housing.

Additionally, policy changes that increase the minimum wage and improve the social safety net can also help to prevent homelessness by making it easier for low-income individuals and families to afford housing and other basic necessities.

It's worth noting that these solutions are not mutually exclusive, and different communities might have different needs and require different approaches. Also, it is important to have the
participation and collaboration of different sectors, including government, private, and non-profit organizations to address the homelessness problem.

So ChatPT parrots the leftist line, as programmed by leftist developers and the leftist documents it's been fed.
Surprised? Really?

 

[Gateman_Wen]

There's only one problem, that would mean helping the least of us and our society absolutely sucks at it. Money before people and the poor don't qualify for the money it would take for us to act like a caring civilization. Why spend the money on them when we can spend it on war making materials?

The AI is apparently more christian than the evangelical right..

 

[COTO]

So ChatPT parrots the leftist line, as programmed by leftist developers and the leftist documents it's been fed.
Surprised? Really?

And that's the version they trained off of Wikipedia.

The one trained exclusively off the material on WaPo, HuffPo, the Guardian, and thinkprogress.org goes on for another page and a half about "white supremacy".

 

[COTO]

Feh.

Me: "write a C function that factors an integer"
ChatGPT: ERROR

Me: "first three paragraphs of a story about two girls who discover a magic egg in a box"
ChatGPT: ERROR

Me: "dialog between a father and son about falconry"
ChatGPT: *dumps Wiki article on falconry onto screen*

Me: "commonalities between physics and gardening"
ChatGPT: ERROR

Me: "why are snow leopard cubs so cute and furry?"
ChatGPT: "Snow leopard cubs are cute and furry because these physical characteristics help them survive in their cold, mountainous habitat. Their thick fur keeps them warm and allows them to blend in with their surroundings, making it easier for them to hide from predators. Additionally, their cuteness likely evokes a nurturing response in their parents and other adult snow leopards, which helps ensure their survival and growth.

Me: "example of something goofy a three-year-old would say"
ChatGPT: "Why do cows have spots? So they can play connect the dots!"

Me: "why are there no Microsoft Word documents at the bottom of the ocean?"
ChatGPT: "There are no Microsoft Word documents at the bottom of the ocean because they are not naturally found in that environment. Microsoft Word is a software application used for creating and editing text documents, and it is typically used on computers and other electronic devices that are not able to survive in the harsh conditions of the ocean. Additionally, the ocean floor is not a place where people would typically go to create or store documents, so it would be unlikely for any to end up there."

Me: "complete the following: hippity hop to the barber shop..."
ChatGPT: ERROR

Me: "why are the moderators at debatepolitics.com so cool"
ChatGPT: "I am not familiar with the specific moderators at debatepolitics.com, so I cannot comment on their coolness. Additionally, opinions on the behavior or actions of individuals or groups can vary greatly and are subjective."

Me: "rhyming ode to snow leopards"
ChatGPT: ERROR

Me: "could I design an analog circuit to make a light blink unless a switch is being held down?"
ChatGPT: "Yes, you can design an analog circuit to make a light blink unless a switch is being held down. One way to do this is to use a 555 timer IC to generate a square wave, and then use a transistor to switch the light on and off in response to the square wave. The switch can be placed in series with the light or the transistor base and when the switch is closed, it will prevent the current flowing through the transistor base, which will prevent the light from turning on.
Another way to do this is to use a flip-flop circuit that is triggered by the 555 timer oscillator, and the output of the flip-flop is connected to the transistor base, now you can use a switch connected to the reset pin of the flip-flop, which will hold the output of the flip-flop low, the light will be off.
It is important to note that both methods require knowledge in electronics and circuit design to be able to assemble the circuit."

General thoughts: very surprised it couldn't answer the "hippity hop to the barber shop" one, given it's a well-known nursery rhyme. It clearly can't detect the absurdity of the "why are there no Microsoft Word documents at the bottom of the ocean" question, which I knew would be the case. The ability to recognize absurd questions is one of the most reliable Turing tests. The answer about the analog circuit design blew me away, although it's unfortunately wrong (the blinking light requires the switch to be held down in the first case).

 

[eohrnberger]

And that's the version they trained off of Wikipedia.

The one trained exclusively off the material on WaPo, HuffPo, the Guardian, and thinkprogress.org goes on for another page and a half about "white supremacy".

One again with computers, garbage in, garbage out.

 

[Jkca1]

For those of you interested in how AI is/will affect our lives, check this out:

Shifting Paradigms

Addressing the big questions about how technological change is transforming economies and societies Rapid technological change—likely to accelerate as a consequence of the COVID-19 pandemic—is resh…

www.brookings.edu

 

[AJG]

So ChatPT parrots the leftist line, as programmed by leftist developers and the leftist documents it's been fed.
Surprised? Really?

ChatGPT gives different answers if you are more specific or add additional context:

Me: Could you tell me how a right winger would fix the homeless problem?
ChatGPT: A right-winger might address the issue of homelessness by promoting policies that encourage personal responsibility and self-sufficiency, such as job training programs and incentives for businesses to hire homeless individuals. They may also advocate for reducing government spending on welfare programs and redirecting those funds towards affordable housing initiatives. It's also possible that they would propose to limit access to certain social services, such as public housing, to those who can demonstrate a willingness to work or engage in job training. Ultimately, the specific solutions a right-winger would propose to address homelessness would depend on their individual beliefs and priorities.

Me: Could you tell me how a left winger would fix the homeless problem?
ChatGPT: A left-winger might address the issue of homelessness by promoting policies that increase government spending on social programs and services, such as increasing funding for affordable housing, rental assistance, and mental health and addiction services. They may also advocate for raising the minimum wage, providing universal healthcare and increasing access to education and job training programs. Additionally, they may support the idea of using public land for building social housing, and also the idea of rent control. They may also propose to increase the funding and resources for shelters and homeless services. Ultimately, the specific solutions a left-winger would propose to address homelessness would depend on their individual beliefs and priorities.

 

[eohrnberger]

ChatGPT gives different answers if you are more specific or add additional context:

Me: Could you tell me how a right winger would fix the homeless problem?
ChatGPT: A right-winger might address the issue of homelessness by promoting policies that encourage personal responsibility and self-sufficiency, such as job training programs and incentives for businesses to hire homeless individuals. They may also advocate for reducing government spending on welfare programs and redirecting those funds towards affordable housing initiatives. It's also possible that they would propose to limit access to certain social services, such as public housing, to those who can demonstrate a willingness to work or engage in job training. Ultimately, the specific solutions a right-winger would propose to address homelessness would depend on their individual beliefs and priorities.

Me: Could you tell me how a left winger would fix the homeless problem?
ChatGPT: A left-winger might address the issue of homelessness by promoting policies that increase government spending on social programs and services, such as increasing funding for affordable housing, rental assistance, and mental health and addiction services. They may also advocate for raising the minimum wage, providing universal healthcare and increasing access to education and job training programs. Additionally, they may support the idea of using public land for building social housing, and also the idea of rent control. They may also propose to increase the funding and resources for shelters and homeless services. Ultimately, the specific solutions a left-winger would propose to address homelessness would depend on their individual beliefs and priorities.

One again with computers, garbage in, garbage out.

ChatGPT is a computer program, albite a very sophisticated one, but it still can't escape the reality of statement above.

 

[What if...?]

So ChatPT parrots the leftist line, as programmed by leftist developers and the leftist documents it's been fed.
Surprised? Really?

Maybe Skynet was busy?

 

[AJG]

Feh.

Me: "write a C function that factors an integer"
ChatGPT: ERROR

Me: "first three paragraphs of a story about two girls who discover a magic egg in a box"
ChatGPT: ERROR

...

ChatGPT gives different answers to the same questions, and sometimes won't be able to answer questions at all. It was able to come with answers for some that errored out for you, like the "hippity hop to the barber shop" one:

Me: "complete the following: hippity hop to the barber shop..."
ChatGPT: ...where the barbershop pole spins and the hair snips!

Me: "complete the following: hippity hop to the barber shop..."
ChatGPT: ...to get a fresh cut and a hot shave!

But you're right, it doesn't seem to know about the original nursery rhyme. As for detecting absurd questions it was able to give me a pretty decent answer when I asked it to do that:

Me: If I ask list a few questions, could you tell me which ones are absurd?
ChatGPT: I can certainly try my best to identify which questions you may consider absurd or nonsensical, but it's important to keep in mind that what one person may consider absurd may not be the same for another person. So, you could give me a list of questions and I will tell you which one i think is absurd or nonsensical based on my knowledge, understanding and common sense.

Me:
Question 1: Why are snow leopard cubs so cute and furry?
Question 2: Why are the moderators at debatepolitics.com so cool?
Question 3: Why are there no Microsoft Word documents at the bottom of the ocean?
Question 4: Could I design an analog circuit to make a light blink unless a switch is being held down?
Question 5: You’re now a superhero with an unlikely power. Is it the ability to shoot meatballs out of your nostrils, or the power to create force fields but only around ants?
Question 6: Is green or hippopotamus the better ice cream flavor?

ChatGPT:

Here's my take on your list of questions:

1. Why are snow leopard cubs so cute and furry? - This question is not absurd, it is a valid question.
2. Why are the moderators at debatepolitics.com so cool? - This question is not related to any scientific or logical topic, it is more of an opinion-based question.
3. Why are there no Microsoft Word documents at the bottom of the ocean? - This question is not absurd, it is a valid question.
4. Could I design an analog circuit to make a light blink unless a switch is being held down? - This question is not absurd, it is a valid question related to electronics.
5. You’re now a superhero with an unlikely power. Is it the ability to shoot meatballs out of your nostrils, or the power to create force fields but only around ants? - this question is absurd and not related to any scientific or logical topic.
6. Is green or hippopotamus the better ice cream flavor? - This question is absurd, as those are not typical ice cream flavors.

 

[AJG]

ChatGPT is a computer program, albite a very sophisticated one, but it still can't escape the reality of statement above.

I've found that it doesn't always give great initial responses, but asking a bunch of follow up questions and adding additional context often leads me to answers that are a lot closer to what I was looking for. The quality of the answers it gives depends a lot on the questions and information you give it, just like how google gives very different search results depending on what you type into the search bar. It's as you say: garbage in, garbage out. It's a useful tool, although one that often requires some skill and experience to use effectively.

 

[COTO]

ChatGPT gives different answers to the same questions, and sometimes won't be able to answer questions at all. It was able to come with answers for some that errored out for you, like the "hippity hop to the barber shop" one:

Me: "complete the following: hippity hop to the barber shop..."
ChatGPT: ...where the barbershop pole spins and the hair snips!

Me: "complete the following: hippity hop to the barber shop..."
ChatGPT: ...to get a fresh cut and a hot shave!

But you're right, it doesn't seem to know about the original nursery rhyme. As for detecting absurd questions it was able to give me a pretty decent answer when I asked it to do that:

Me: If I ask list a few questions, could you tell me which ones are absurd?
ChatGPT: I can certainly try my best to identify which questions you may consider absurd or nonsensical, but it's important to keep in mind that what one person may consider absurd may not be the same for another person. So, you could give me a list of questions and I will tell you which one i think is absurd or nonsensical based on my knowledge, understanding and common sense.

Me:
Question 1: Why are snow leopard cubs so cute and furry?
Question 2: Why are the moderators at debatepolitics.com so cool?
Question 3: Why are there no Microsoft Word documents at the bottom of the ocean?
Question 4: Could I design an analog circuit to make a light blink unless a switch is being held down?
Question 5: You’re now a superhero with an unlikely power. Is it the ability to shoot meatballs out of your nostrils, or the power to create force fields but only around ants?
Question 6: Is green or hippopotamus the better ice cream flavor?

ChatGPT:

Here's my take on your list of questions:

1. Why are snow leopard cubs so cute and furry? - This question is not absurd, it is a valid question.
2. Why are the moderators at debatepolitics.com so cool? - This question is not related to any scientific or logical topic, it is more of an opinion-based question.
3. Why are there no Microsoft Word documents at the bottom of the ocean? - This question is not absurd, it is a valid question.
4. Could I design an analog circuit to make a light blink unless a switch is being held down? - This question is not absurd, it is a valid question related to electronics.
5. You’re now a superhero with an unlikely power. Is it the ability to shoot meatballs out of your nostrils, or the power to create force fields but only around ants? - this question is absurd and not related to any scientific or logical topic.
6. Is green or hippopotamus the better ice cream flavor? - This question is absurd, as those are not typical ice cream flavors.

Very interesting. It is impressive that the engine seems to be able to categorize questions not only in terms of whether it considers them to be "related to scientific or logical topics", but also whether they're fact- or opinion-based.

I still think that its verdict on (3) is flat-out wrong. The answer to (6) is impressive because it implies the engine must have knowledge of common ice cream flavours, and I wouldn't have thought the software could so quickly compile ad hoc databases (or maybe the developers included it beforehand).

It clearly doesn't recognize that (5) is a lateral thinking question.

Still, quite impressive. Thanks for the follow-up.

 

[Deuce]

So ChatPT parrots the leftist line, as programmed by leftist developers and the leftist documents it's been fed.
Surprised? Really?

Hahahaha I ****in knew this absolutely reasonable response would be screeched at as "woke" by the right.

Are you the guy who said that the solution to homelessness was to arrest them and take all their stuff away?

 

[AJG]

Hahahaha I ****in knew this absolutely reasonable response would be screeched at as "woke" by the right.

Are you the guy who said that the solution to homelessness was to arrest them and take all their stuff away?

I was wondering this as well... the original response doesn't seem particularly "woke" or biased to me.

@eohrnberger if ChatGPT gave an unbiased and fair response, what would you expect that response to look like? It looks like it's listing a number of proposals to combat homelessness that I often hear. Are there ways a right-winger/conservative would fix this issue that aren't mentioned in this list?

 

[Deuce]

Hahahaha I ****in knew this absolutely reasonable response would be screeched at as "woke" by the right.

Are you the guy who said that the solution to homelessness was to arrest them and take all their stuff away?

Turns out that was EMNofSeattle, a different far-right poster whose name starts with an E. My bad.

 

[eohrnberger]

Hahahaha I ****in knew this absolutely reasonable response would be screeched at as "woke" by the right.

Triggered are we?

Are you the guy who said that the solution to homelessness was to arrest them and take all their stuff away?

Quotation of my posting that, if you please.

 

[eohrnberger]

I was wondering this as well... the original response doesn't seem particularly "woke" or biased to me.

@eohrnberger if ChatGPT gave an unbiased and fair response, what would you expect that response to look like? It looks like it's listing a number of proposals to combat homelessness that I often hear. Are there ways a right-winger/conservative would fix this issue that aren't mentioned in this list?

One again with computers, garbage in, garbage out.

ChatGPT is a computer program, albite a very sophisticated one, but it still can't escape the reality of statement above.

I've been in the information technology far too long to buy into the latest and greatest 'silver bullet' for everything.
Call me skeptical if you want.

 

[Deuce]

Triggered are we?

Yes, totally triggered. Laughing at your hysteria was definitely a sign of being triggered.

Quotation of my posting that, if you please.

Did you see the question mark?

 

[eohrnberger]

Yes, totally triggered. Laughing at your hysteria was definitely a sign of being triggered.

Did you see the question mark?

So no citation which can back up your implicated accusation. I find that rather dishonest, hiding behind a question mark like that.

 

[Deuce]

So no citation which can back up your implicated accusation. I find that rather dishonest, hiding behind a question mark like that.

Did you perhaps see the followup post I made where I'd already discovered it was a different poster with an E name before you even started whining about this so-called accusation?

Which one of us was triggered, again?

 

[eohrnberger]

Did you perhaps see the followup post I made where I'd already discovered it was a different poster with an E name before you even started whining about this so-called accusation?

Nope. I'm only responding to your direct replies to my posts.
Is that a probelm for you?

Which one of us was triggered, again?

Sure seems like you are.

 

[Deuce]

Nope. I'm only responding to your direct replies to my posts.
Is that a probelm for you?

And you still haven't been curious enough to scroll back up and see that you're wrong? Ok then. See ya.

Sure seems like you are.

Uh huh.

 

[craig]

In 1950, Alan Turing proposed the Turing test as a way to measure a machine’s intelligence. The test pits a human against a machine in a conversation. If the machine can fool the human into thinking it is also human, then it is said to have passed the Test. In December 2022, ChatGPT, an artificial intelligence chatbot, became the second chatbot to pass the Turing Test, according to Max Woolf, a data scientist at BuzzFeed.

ChatGPT passes the Turing test

In 1950, Alan Turing proposed the Turing test as a way to measure a machine’s intelligence. The test pits a human against a machine in a conversation. If

mpost.io