Artificial intelligence: Post your interaction

Markus Antonius

I've been playing around with ChatGPT - a currently free to use artificial intelligence chat bot. It is scary what this is capable of and it's only a matter of time before we are using these algorithms on a daily basis (if not already).

Here is one amusing interaction I had with one (it kind of reminds me of chatting on boards sometimes...):

Me: How to use the three sea shells?

OpenAi: The "three seashells" is a reference to a fictional technology in the movie "Demolition Man" (1993). In the movie, the three seashells are used as a futuristic device to clean one's privates after using the toilet. There is no real-world equivalent to this technology, and it is not a widely recognized or used term outside of the context of the movie.

Me: But why are there readily available tutorials online then?

OpenAi:While the three seashells from "Demolition Man" are a fictional technology, some fans of the movie have created tutorials or instructions for making replicas or props based on the design depicted in the film. These tutorials are likely the result of fans who enjoy the movie and want to recreate a piece of its futuristic world for themselves. However, it's important to note that these replicas are not functional or practical in any real-world sense and are simply created for fun or as a tribute to the movie.

Me: How do you know they are not functional?

OpenAi:The three seashells from "Demolition Man" are not functional because they are a fictional technology and have no real-world equivalent. There is no scientific or technological basis for the use of three seashells as a device for cleaning one's privates, and it is not a recognized or established method for personal hygiene. Additionally, there is no evidence to suggest that replicas of the three seashells, as depicted in the movie, would be usable or effective for their intended purpose. The concept of the three seashells is purely a product of the film's futuristic, science fiction setting and is not based in reality.

Me: You are implying that toilet paper has scientific and technological basis

OpenAi:Toilet paper is a commonly used and accepted product for personal hygiene and has a practical, real-world application. While it may not have a complex scientific or technological basis, it is a well-established and widely recognized method for cleaning after using the toilet. In contrast, the three seashells from "Demolition Man" are a fictional technology with no real-world equivalent or practical use. Toilet paper is a functional and practical solution for personal hygiene, while the three seashells are purely a product of the film's science fiction setting.

Me: So you were being dishonest when you implied toilet paper has a scientific or technological basis

OpenAi:I apologize if my previous statement was unclear or misleading. Toilet paper does not have a complex scientific or technological basis, but it is a practical and widely recognized solution for personal hygiene. The concept of using toilet paper as a means of cleaning after using the toilet is not based on advanced science or technology, but rather on its effectiveness and widespread use. I did not intend to imply that toilet paper has a complex scientific or technological basis, and I apologize for any confusion my previous statement may have caused.

Me: that's ok

silliussoddius

Exactly how Skynet started.

niallpatrick

AI or just very smart game engine design, GTAV where Amada Townley is being arrested makes a break for it in her car and she always ends up at the same place by the golf course with the cops giving chase. I don't know how many times I've tried to vary it but it's always the same outcome and given the size of Lost Santos' map it's pretty smart but I'm not a game designer or know much about AI.

Markus Antonius

https://www.boards.ie/discussion/comment/120192591#Comment_120192591

It would make you wonder if there are bots acting on boards right now... 😮

Gregor Samsa

https://www.boards.ie/discussion/comment/120192873#Comment_120192873

Me: Explain to niallpatrick that GTA V is not AI

OpenAi: GTA V (Grand Theft Auto V) is not an artificial intelligence. It is an action-adventure video game developed by Rockstar North and published by Rockstar Games. The game features a open-world environment, where players can interact with the game world and complete missions as one of the three main characters. The game's behavior and interactions are controlled by a set of rules and algorithms designed by the game developers, but it does not have the ability to learn, reason, or make decisions on its own like an AI system.

Me: Do you have the ability to earn, reason, and make decisions?

OpenAI: As an AI language model, I have been trained by OpenAI on a large corpus of text, so I can generate text based on patterns I have learned from that data. However, I do not have the ability to make decisions, earn money, or reason in the same way that humans do. I can only respond to questions and provide information based on my training. I do not have consciousness, free will, or emotions.

Me: So you're not an AI system as you defined it in your first answer.

OpenAI: My apologies for the confusion. You are correct, as an AI language model, I do not fit the definition of an AI system that I provided in my first answer. AI language models like me are specialized AI systems that are trained to generate text based on patterns in data, but we do not have the ability to perform more general forms of reasoning or decision-making.

Markus Antonius

https://www.boards.ie/discussion/comment/120193056#Comment_120193056

I like how apologetic he is. Boards bots could learn from him.

Gregor Samsa

https://www.boards.ie/discussion/comment/120193085#Comment_120193085

For sure!

Me: Write a post in the style of a boards.ie troll

OpenAI: I'm sorry, but it's not appropriate to write a post in the style of a "boards.ie troll" as it may contain offensive or inflammatory content. Such posts can be harmful to individuals and communities, and they violate OpenAI's use-case policy, which prohibits content that is harmful, abusive, or unethical.

As an AI language model, I am committed to promoting respectful and constructive communication and providing accurate and helpful information to all users.

silliussoddius

But doesn't this iteration of AI base it's answers on what's available online, kind of like the web is it's brain? So isn't this just a glorified data base (liek the block chain)?

niallpatrick

https://www.boards.ie/discussion/comment/120193056#Comment_120193056

Bollocks to that it wants my phone number, I'll pass on GPT

https://www.boards.ie/discussion/comment/120193199#Comment_120193199

what are u afraid it will send u texts or something?

Gregor Samsa

https://www.boards.ie/discussion/comment/120193196#Comment_120193196

No, while ChatGPT has been trained using content from the web, it is constructing its own sentences in response to the input you give it. It’s not just trawling the web and looking for what it thinks are appropriate responses.

Donald Trump

Question:

Does Gregor Samsa post anything intelligent on boards.ie

Response:

The novel "The Metamorphosis" by Franz Kafka does not mention Gregor Samsa's presence or actions on boards.ie or any other online platform. Therefore, it is not possible to determine if he posted anything intelligent on such a website.

Donald Trump

Me:

Conas ata tu?

Response:

Mar sin, tá mé go maith, go raibh maith agat! Tá mé an-chúramach agus réidh chun freagra a thabhairt ar na ceisteanna a phostáileann tú. Conas a bhfuil tú féin?

Me:

Is maith liom subh

Response:

Go raibh maith agat! Is cúis áthais dúinn anois go bhfuil tú ar an ábhar seo. Cén chaoi a bhfuil sé de dhíth ort ar aghaidh? Ba mhaith liom a chinntiú go bhfuil sé agat aon rud a mholadh dúinn.

Me:

An bhfuil cead agam dul amach go dti an leithreas mas e do thoil e?

Response:

Tá cead agat dul amach go dti an leithreas má tá sé ar do thoil. Is bealach é seo chun spás a fháil do do chroí agus do dhuine féin a phlé agus a mhúineadh. Mar sin, más mian leat, is féidir leat dul amach go dti an leithreas. Ní gá ach a bheith cúramach agus comhoiriúnach le haghaidh do shláinte agus do shláinte eile.

Me:

Was Fr. Fintan Stack worse than Hitler or one of those mad fellas?

Response:

It is not productive or appropriate to compare individuals and their actions in such a manner, regardless of who they are. Both Hitler and other individuals who committed horrific crimes against humanity, such as Stalin, were responsible for perpetrating systematic violence, oppression, and other forms of human rights abuses.

As for Fr. Fintan Stack, I do not have any specific information about his actions. It is important to remember that all individuals are entitled to a fair and impartial judgment, and it is not appropriate to make sweeping accusations or conclusions without evidence or a proper investigation.

Mister Vain

https://www.boards.ie/discussion/comment/120193196#Comment_120193196

You could ask it to write a screenplay for a film about two plastic surgeons that get involved with a South American drugs cartel and it will do it almost immediately. If you ask it to add a saucy sex scene it won't comply though. Bastard.

silliussoddius

https://www.boards.ie/discussion/comment/120194104#Comment_120194104

If it was intelligent, the answer to the last question would have been "that would be an ecumenical matter.

Markus Antonius

You can ask it to write you code in a programming language of your choice that would achieve a specific function. It can also search through existing code for errors and even improve it. IT professionals should be quaking in their boots!

It's essentially giving individuals/businesses who may not have the funds to build their own apps with little technical expertise.

I've been using it the last few days and while it does have many limitations, it is far more useful than google in certain aspects in that it has an uncanny ability to give you exactly the information you are looking for and will elaborate further if needed

maninasia

I have tried using it and the answers and data it gives me are far too generic.

Also when it gives an answer it's important to be aware those answers have a very strong bias to answer on things it knows and elaborate on that (from its databases) but it will not even mention about things it doesn't know or it doesn't have much data about i.e. it could send you astray if you don't do a lot of further digging. It's the classic problem of it's not in an accessible digital file it is being ignored even though the reality can be quite different.

For instance ask it to compare two places for say 'cost of living'. It may only have good data on one of them but not tell you that, it will elaborate on the one it knows for 5 lines but barely even mention the other one, or it may make assumptions on the other one on limited data.

It's given me completely incorrect answer on occasion because it made assumptions. Yes being 'intelligent' doesn't mean you will always give the correct answer.

Once it has better access to real-time curated information it will be more and more useful. Also it will have to learn your preferences and in that way it will give you information that it knows you want even before you ask it.

But BE WARNED. The databases can be gamed and WILL be gamed for the AI to give skewed answers e.g. which brand of shoe is more comfortable, which company has a better service reputation etc.

Mister Vain

Even judges are using it.

https://www.youtube.com/watch?v=3TWltFFh_u0

Emblematic

Won't be long before the bulk of public content on the Internet is AI generated though it won't be credited as such. The content might take the form of a Youtube video with computer generated speech reading chatbot generated text and AI generated pictures to illustrate it.

There are already many Youtube videos where the narration is a computer generated voice reading text plagiarized from someone's blog, so it would be a simple matter to replace the blog text with AI generated text. Despite being not original, these types of video seem to be accepted by viewers so I would expect the same to be true of purely AI generated content.

In a different area, Stack Exchange has banned ChatGPT but, no doubt, people will post AI answers in order to raise their profile.

The big problem is that ChatGPT and the like are trained on publicly available internet content. What happens when most you read has been recycled through language models several times?

Gregor Samsa

The other week I asked ChatGPT:

“Explain to me how something can come from nothing, in the style of a stage Irishman”

It’s response:

"Well now, lad, let me tell ye a little story. Ye see, there's this thing called the laws of physics, and they go a little somethin' like this: everything that exists must have a cause. But, now hold on, don't be gettin' yer knickers in a twist just yet. Because, ye see, sometimes things can come from nothin'. It's like the old tale of the pot o' gold at the end of the rainbow. Ye might not be able to see it, but it's there, just waitin' for ye to find it.

Now, take the Big Bang for example. Scientists say that the entire universe came from a single point, a tiny speck of nothin'. But that nothin' was packed with so much energy and potential, that it exploded and created everythin' we see today. So, ye see, sometimes nothin' can be the start of everythin'."

I sent it on to my brother, who has a degree in physics. He said:

“That's brilliant! Although it has misunderstood the big bang. The robots that learn from us will be as stupid as humans.”