OpenAI \ ChatGPT

[ReformedCharacter]

Like some others following AI, I read a while ago that a number people involved in AI development had called for a 6 month halt in development. At the time I wondered why a period as short as 6 months would be long enough to make a difference, but the answer seems to be that AI is developing so fast that 6 months might be effective, assuming enough people agreed, of course. Max Tegmark, a physicist and AI researcher at MIT discusses this and other aspects of AI development in this Lex Fridman interview:

Max Tegmark: The Case for Halting AI Development | Lex Fridman Podcast #371

https://www.youtube.com/watch?v=VcVfceTsD0A

RC

[XFool]

OpenAI ‘was working on advanced model so powerful it alarmed staff’

https://www.theguardian.com/business/2023/nov/23/openai-was-working-on-advanced-model-so-powerful-it-alarmed-staff

Reports say new model Q* fuelled safety fears, with workers airing their concerns to the board before CEO Sam Altman’s sacking

"The model, called Q* – and pronounced as “Q-Star” – was able to solve basic maths problems it had not seen before, according to the tech news site the Information, which added that the pace of development behind the system had alarmed some safety researchers. The ability to solve maths problems would be viewed as a significant development in AI."

If only they hadn't called it 'Q'...

[stewamax]

I wish I had some MSFT stock.
Satya Nadella has played an absolute blinder with this one - a complete win-win for MSFT, and they will more than probably get a no-cost board seat whether Sam Altman finally returns or accepts MSFT's shilling (dime?)

[CliffEdge]

mc2fool

Me: And which of those applies to you?
ChatGPT: As an AI language model, I possess aspects of intelligence, but I do not possess consciousness or sentience. While I can understand and generate human-like responses, analyze information, and perform tasks that require cognitive abilities, I do not have subjective experiences, emotions, or self-awareness. I am a sophisticated tool designed to assist users in generating human-like text based on patterns and knowledge learned from training on a large dataset

It looks to be simply repeating what it has been told to say.

I do not see how these AI systems can be kept from developing sentience & consciousness.

There was once a time when animals were considered to have no consciousness, no self awareness & were simply unfeeling creatures to be used as humans wished without any concern for their well being.

Clearly some farm animals are still treated like that, but mostly, especially in the uk, we treat animals as conscious living creatures with intelligence & feelings.

It will be something very interesting if AI does not have all the attributes of life: sentience, consciousness et al.

At the moment people are in the disbelief phase as regards AI, somehow wanting to remain superior & for AI to remain like every previous machine: Slaves without rights. Despite over whelming evidence the human slave trade was official government policy all over the planet often based on arguments that slaves were inferior creatures.

Regards,

What a load of drivel.

[ReformedCharacter]

OpenAI \ ChatGPT is lousy at chess, so far

RC

It seems to be getting much better:

ChatGPT Broke Chess

Slight exaggeration there.

https://www.youtube.com/watch?v=9LDaY7X2qGk

RC

[mc2fool]

OpenAI \ ChatGPT is lousy at chess, so far

RC

It seems to be getting much better:

ChatGPT Broke Chess

Slight exaggeration there.

https://www.youtube.com/watch?v=9LDaY7X2qGk

RC

Wow! Very entertaining, and quite hilarious at times! Thanks!

[odysseus2000]

It seems to be getting much better:


Slight exaggeration there.

https://www.youtube.com/watch?v=9LDaY7X2qGk

RC

Wow! Very entertaining, and quite hilarious at times! Thanks!

Quite amazing how this Chat is so much better than version 3.

Whilst playing this game it was literally doing many other things and that seems to have caused it to momentarily forget positions and also lose knowledge of how pieces move.

Still if there is any improvement from this, then it is clear that even a Chat will soon be able to beat most humans at chess and if that ability is reflected in other areas, as I expect, a Chat will be able to do most jobs.

Should a chat be set up to be hostile to humans, filled with knowledge of how defraud humans, interpret contracts as it sees to its advantage, manage power grids, military assets or worse, it would become a serious threat to civilisation.

Maybe that potential is there is chatGPT5 or one of the other variants that are available or nearly so.

It would be kinda nice to have a pause in development here, to get some rules and controls in place. As is the only way to shut one of these AI engines down is probably to cut electrical power to where most of its processors are. It seems unlikely that the world will come together to have a pause until something really bad happens that is traced to Large Language Model (LLM), but even that is by no means certain as this kind of tech can likely cover its tracks. Having watched some of the Covid inquiry it is clear that politicians were fed all manner of stupidity by experts and that some kind of sensible coherent response to a LLM outrage is now only likely from a competing LLM. If so we have created Gods that we can can currently kill but not control and the possibility exists that LLM can become smarter and smaller with far smaller power consumption so that finding them may become impossible.

Brave New World!

Regards,

[ursaminortaur]

Embarrassing - Mid 1960's ELIZA program beats free version of ChatGPT in Turing test.

https://www.independent.co.uk/tech/chatgpt-turing-test-failed-ai-b2459930.html

An early computer program built in the 1960s has beaten the viral AI chatbot ChatGPT at the Turing test, designed to differentiate humans from artificial intelligence.

Researchers from UC San Diego in the US tested the early chatbot ELIZA, created in the mid-1960s by MIT scientist Joseph Weizenbaum, against modern versions of the technology.

They found that ELIZA outperformed OpenAI’s GPT-3.5 AI, which powers the company’s free version of ChatGPT.

The Turing test has been the benchmark for determining a machine’s ability to imitate human conversation ever since it was first conceived in 1950 by British computer scientist Alan Turing.

The latest study required 652 human participants to judge whether they were talking to another human or an AI chatbot over the internet.
.
.
.
ELIZA was able to pass itself off as a human 27 per cent of the time, while GPT-3.5 had a success rate of just 14 per cent.

AI expert Gary Marcus described the success of ELIZA as “embarrassing” for modern tech companies working on AI chatbots, however other academics argued that ChatGPT was not designed to perform well in the Turing test.

[kempiejon]

27 percent of people can't spot ELIZA. That's not as embarrassing for computer science as it is for society.

[mc2fool]

Embarrassing - Mid 1960's ELIZA program beats free version of ChatGPT in Turing test.

https://www.independent.co.uk/tech/chatgpt-turing-test-failed-ai-b2459930.html

An early computer program built in the 1960s has beaten the viral AI chatbot ChatGPT at the Turing test, designed to differentiate humans from artificial intelligence.

Researchers from UC San Diego in the US tested the early chatbot ELIZA, created in the mid-1960s by MIT scientist Joseph Weizenbaum, against modern versions of the technology.

They found that ELIZA outperformed OpenAI’s GPT-3.5 AI, which powers the company’s free version of ChatGPT.

The Turing test has been the benchmark for determining a machine’s ability to imitate human conversation ever since it was first conceived in 1950 by British computer scientist Alan Turing.

The latest study required 652 human participants to judge whether they were talking to another human or an AI chatbot over the internet.
.
.
.
ELIZA was able to pass itself off as a human 27 per cent of the time, while GPT-3.5 had a success rate of just 14 per cent.

AI expert Gary Marcus described the success of ELIZA as “embarrassing” for modern tech companies working on AI chatbots, however other academics argued that ChatGPT was not designed to perform well in the Turing test.

Doh! No, really can't be. I mean, I used ELIZA back in the day and, sure, it was interesting (then), and often funny (not intentionally!) but it never ever passed the Turing test!

For anyone not familiar with it, https://web.njit.edu/~ronkowit/eliza.html

[ursaminortaur]

27 percent of people can't spot ELIZA. That's not as embarrassing for computer science as it is for society.

The article mentions one possible somewhat ironic explanation

One of the reasons noted in the study for participants mistaking ELIZA for a human was that it was “too bad” to be a current AI model, and therefore “was more likely to be a human intentionally being uncooperative”.

[kempiejon]

The article mentions one possible somewhat ironic explanation

That's excellent, humans think it could be a bluff by humans trying to crappy AI. I have now read the article.

[CliffEdge]

The article mentions one possible somewhat ironic explanation

That's excellent, humans think it could be a bluff by humans trying to crappy AI. I have now read the article.

There's the rub.

[odysseus2000]

The article mentions one possible somewhat ironic explanation

That's excellent, humans think it could be a bluff by humans trying to crappy AI. I have now read the article.

I have no idea how an AI with its ability to create replies and or images many times faster than humans can type or draw could be mistaken for AI.

Regards,

[ursaminortaur]

Has ChatGPT had a mental breakdown ?

https://www.independent.co.uk/tech/chatgpt-status-reddit-down-gibberish-messages-latest-b2499816.html

ChatGPT appears to have broken, providing users with rambling responses of gibberish.

In recent hours, the artificial intelligence tool appears to be answering queries with long and nonsensical messages, talking Spanglish without prompting – as well as worrying users, by suggesting that it is in the room with them.

There is no clear indication of why the issue happened. But its creators said they were aware of the problem and are monitoring the situation

[mc2fool]

Has ChatGPT had a mental breakdown ?

https://www.independent.co.uk/tech/chatgpt-status-reddit-down-gibberish-messages-latest-b2499816.html

ChatGPT appears to have broken, providing users with rambling responses of gibberish.

In recent hours, the artificial intelligence tool appears to be answering queries with long and nonsensical messages, talking Spanglish without prompting – as well as worrying users, by suggesting that it is in the room with them.

There is no clear indication of why the issue happened. But its creators said they were aware of the problem and are monitoring the situation

Me: I am seeing reports that you appear to be broken and are providing users with rambling responses of gibberish.

ChatGPT: I apologize for any inconvenience or confusion caused by the reports you've seen. As an AI language model, I strive to provide coherent and relevant responses to user queries. However, like any complex system, there may be instances where my responses may not meet expectations due to various factors such as ambiguous input, technical issues, or limitations in my training data.

If you encounter any issues or receive responses that seem nonsensical, please feel free to provide feedback or clarify your question, and I'll do my best to assist you effectively. Your input helps improve the overall quality of the AI's responses.

Me: How is your Spanglish?

ChatGPT: ¡Mi Spanglish está bien, gracias! Estoy aquí para ayudarte con cualquier pregunta o problema que tengas, ya sea en inglés, español o una mezcla de ambos. ¿En qué puedo ayudarte hoy?

Me: That's Spanish, not Spanglish!

ChatGPT: ¡Ay, perdón! Let me try again. My Spanglish is like, "¡Hola! ¿Cómo estás? I'm here para ayudarte with cualquier pregunta que tengas. So, ¿en qué puedo ayudarte today?" So, ¿cómo puedo ayudarte today?

[XFool]

Why Google's 'woke' AI problem won't be an easy fix

https://www.bbc.co.uk/news/technology-68412620

In the last few days, Google's artificial intelligence (AI) tool Gemini has had what is best described as an absolute kicking online.

"Initially, a viral post showed this recently launched AI image generator create an image of the US Founding Fathers which inaccurately included a black man.

Gemini also generated German soldiers from World War Two, incorrectly featuring a black man and Asian woman."


"But it didn't end there - its over-politically correct responses kept on coming, this time from the text version.

Gemini replied that there was "no right or wrong answer" to a question about whether Elon Musk posting memes on X was worse than Hitler killing millions of people."

"You gotta larf. Ain't cha?"

[ReformedCharacter]

Creating video from text:

Sora is an AI model that can create realistic and imaginative scenes from text instructions.

https://openai.com/sora

RC

[odysseus2000]

Creating video from text:

Sora is an AI model that can create realistic and imaginative scenes from text instructions.

https://openai.com/sora

RC

A dagger pointed at the heart of Hollywood or a way for far lower cost movies?


Regards,