Chat GPT

Mike Hewson
Mike Hewson
Moderator
Joined: 1 Dec 05
Posts: 6534
Credit: 284734192
RAC: 103342
Topic 228989

This is an awesome 'device' and I can see immediately what teachers are now worried about ie. the effect on students learning while having 'expert' AI interpretations of questions. Could this be the first to truly pass the Turing test ?  If not, it has come bloody close IMHO !

I tried the following subtle question in an attempt to stump it :

In your opinion, to what degree is Phineas Phogg a joke name in the book Around The World In 80 Days ?

The answer :

Phineas Fogg is the main character in the book "Around the World in 80 Days" by Jules Verne. His name is not intended to be a joke, but rather a play on words. The name "Phineas" means "oracular" or "speaking forth," and "Fogg" is a reference to the fog that often obscures one's vision. Together, these names suggest that Fogg is a man who is able to see through the fog and make clear, decisive choices. Additionally, his name Fogg also implies that he is a man of his word, as he is determined to complete his journey around the world in 80 days.

I'm absolutely floored with that answer and it only took a few seconds ( once I got onto the server that is ). Here's another question :

Enrico Fermi produced what famous breakthrough during the Second World War ?

The answer :

During the Second World War, Enrico Fermi and his team produced the first controlled nuclear chain reaction. This breakthrough occurred on December 2, 1942 at the University of Chicago, as part of the Manhattan Project, the top-secret U.S. government program to develop an atomic bomb. The team, led by Fermi, used a pile of graphite bricks and uranium to achieve the first self-sustaining nuclear chain reaction, which marked the beginning of the atomic age. This achievement was a crucial step towards the development of nuclear energy and the atomic bombs that were later used in the war.

No hesitation with that one. Try this for size :

What were the academic papers authored by Albert Einstein in 1960 ?

The answer :

Albert Einstein passed away on April 18, 1955, so he did not author any academic papers in 1960. Einstein had published many papers throughout his lifetime, from the groundbreaking papers on special and general relativity in the early 1900s, to papers on statistical mechanics, quantum theory, and unified field theory. He also published papers on cosmology and on the political and social issues of his time, including his famous papers on the photoelectric effect, Brownian motion, and the equivalence principle. In the last years of his life, Einstein worked on the unified field theory, which aimed to unify the laws of physics and reconcile general relativity and quantum mechanics, but he didn't published any paper on this topic after 1955.

Too obvious, huh ? Slightly more tricky is this one :

What were the academic papers authored by Albert Einstein in 1956 ?

 

The Answer :

 

Albert Einstein did not publish any academic papers in 1956. His last known publication, "The Meaning of Relativity" was in 1955. After the publication of his last scientific paper in 1950, Einstein spent the remaining years of his life working on a unifying theory of physics, but he did not publish any further papers. He also wrote on political and social issues, and corresponded with other scientists and philosophers. Einstein spent most of his time trying to unify the fundamental forces of nature, but he didn't publish any paper on it, and the theory he was working on remained unfinished at the time of his death in 1955.

 

So much the same information there and the wording is flawless. I eventually got it by asking the following question :

 

What is the difference between an axial and a chiral vector ?

 

It did not answer. Both 'axial vector' and 'chiral vector' make sense but it clearly lacked domain knowledge here. It also failed to tell me how to hot wire a car - but that is an explicit constraint of the system not to help with such potentially 'criminal' knowledge.

 

This is all quite spooky and if this is built upon then what are the possibilities ?

Would anyone else like to try ChatGPT and post their discourse here ? I think it would be fascinating ie. your opinions of it. Try a question in an area that you know a fair bit about so that you may fairly judge the answers. You'll need an email address and a mobile phone number for authorisation.

Cheers, Mike.

I have made this letter longer than usual because I lack the time to make it shorter ...

... and my other CPU is a Ryzen 5950X :-) Blaise Pascal

Mike Hewson
Mike Hewson
Moderator
Joined: 1 Dec 05
Posts: 6534
Credit: 284734192
RAC: 103342

Geez Louise it even

Geez Louise, it even understands quantum mechanics !

What is the central mystery of quantum mechanics ?

The central mystery of quantum mechanics is the phenomenon of wave-particle duality, which states that particles, such as electrons and photons, can exhibit both wave-like and particle-like behavior. This is difficult to reconcile with our classical understanding of the world, and has led to many seemingly paradoxical and counterintuitive phenomena, such as superposition, entanglement, and nonlocality.

Not shabby at all.

What is renormalisation in effective field theory ?

In effective field theory (EFT), renormalization is a procedure used to remove infinite quantities that appear in the calculations when using a limited set of degrees of freedom to describe a physical system. The infinities arise due to the neglect of higher-energy modes and interactions, and are absorbed into a redefinition of the parameters of the theory, such as the masses and coupling constants. This allows for a consistent and predictive low-energy description of the system, while still accounting for the effects of the neglected degrees of freedom and interactions through the renormalized parameters.

Get outa here !

What was Richard Phillips Feynman famous for ?

Richard Phillips Feynman was a physicist who was known for his contributions to the development of quantum mechanics, quantum electrodynamics, and the field of particle physics. He was awarded the Nobel Prize in Physics in 1965 for his work on the development of the Feynman diagrams, a graphical representation of the mathematical expressions used in quantum mechanics and quantum field theory. Feynman is also well-known for his popular lectures and books on physics, which have helped to explain complex scientific concepts to a general audience. He was also a member of the Rogers Commission, which investigated the Space Shuttle Challenger disaster in 1986.

Correct weight.

Who else worked on the problems that Richard Feynman did ?

Richard Feynman made many important contributions to the field of physics throughout his career, and he worked with many other prominent physicists on various problems. Some of the notable figures who worked on similar problems as Feynman include:

  • Freeman Dyson, who developed the Dyson series, a mathematical technique used in quantum field theory that is closely related to Feynman's diagrams.

  • Julian Schwinger, who independently developed the theory of quantum electrodynamics (QED) at the same time as Feynman and also received the Nobel Prize in Physics in 1965 for his work on QED.

  • Sin-Itiro Tomonaga, who developed the theory of quantum electrodynamics in Japan, and along with Schwinger and Feynman received the Nobel Prize in Physics in 1965 for their work on the development of QED.

  • Hans Bethe, who was Feynman's PhD advisor and made key contributions to the understanding of the behavior of atomic nuclei and the energy production in stars.

  • Murray Gell-Mann, who was awarded the Nobel Prize in Physics in 1969 for his work on the theory of the strong interactions, which are the fundamental forces that hold the protons and neutrons inside the atomic nuclei together, Gell-Mann was also an important contributor to the development of the quark model, which describes the internal structure of protons and neutrons as being made up of smaller particles called quarks.

These are some of the most notable figures who worked in similar fields and made significant contributions, but there are many more who also made important contributions in the field of physics and quantum mechanics.

This is getting addictive ! :-)

What is viva voce used for ?

"Viva voce" is a Latin phrase that means "by live voice" and it's used to refer to an oral examination. In the academic setting, a viva voce examination is typically a final oral examination for a candidate who has completed a written thesis or dissertation. It is an opportunity for the candidate to present and defend their research in front of a panel of experts in their field.

The viva voce examination is used to assess the candidate's understanding of their research and the broader context in which it fits. It also provides an opportunity for the examiners to ask questions about the research and to evaluate the candidate's ability to think critically, articulate their thoughts clearly, and defend their conclusions.

The viva voce examination is common in graduate-level education and research, particularly in science, technology, engineering, and mathematics (STEM) fields, but it's not limited to those areas, viva voce is also used in some undergraduate programs, professional certifications and in some cases in job interviews.

I predict a renaissance of the viva voce exam to combat AI tools. I personally have gone both well and poorly at these occasions. It's largely a matter of confidence.

Cheers Mike.

I have made this letter longer than usual because I lack the time to make it shorter ...

... and my other CPU is a Ryzen 5950X :-) Blaise Pascal

mikey
mikey
Joined: 22 Jan 05
Posts: 11889
Credit: 1828218831
RAC: 201663

To me it seems to have a

To me it seems to have a place but if two different students in the same class ask the same question they should get the same answer, meaning it's pretty easy for the teacher/professor to see if someone is cheating. No the facts don't change but the wording won't either in cases of cheating. The key for teachers/professors will be to create homework that goes beyond the standard 'who was so and so' to more in depth questions like you asked in your questions which should point out the cheaters.

Abruraspingi
Abruraspingi
Joined: 6 Apr 22
Posts: 5
Credit: 5295414
RAC: 0

I love ChatGPT as it

I love ChatGPT as it understands Aussie slang and make help me with Python and CISCO scripts 

Gary Roberts
Gary Roberts
Moderator
Joined: 9 Feb 05
Posts: 5842
Credit: 109411081164
RAC: 34951224

If anyone would like to see

If anyone would like to see how ChatGPT does on a college level Astrophysics exam, just check out this YT vid.

The presenter is Professor David Kipping who runs the 'Cool Worlds Lab' in the Astronomy Department at Columbia University.  He uploads regularly on all sorts of topics to do with astrophysics, cosmology, futurism, etc., and the quality is always excellent.  People interested in the science behind Einstein@Home would find a lot of very thought provoking presentations in the huge range of topics he has covered.

Some of his vids get into the several million views category - his most watched one has 5.5M.  The one that I found most intriguing is entitled "Einsteins (small) mistake" and has only 20K views.  It documents an incorrect assumption that Einstein made in an example calculation in his Special Relativity paper.  This only came to light because Kipping was trying to reproduce the calculations from the 'Breakthrough Starshot' project concerning the energy requirements to accelerate the light sail to 0.2C.  Turns out the discrepancy is around 10% less energy needed compared to both Einstein's and the project's calculations.

Cheers,
Gary.

mikey
mikey
Joined: 22 Jan 05
Posts: 11889
Credit: 1828218831
RAC: 201663

Gary Roberts wrote: If

Gary Roberts wrote:

If anyone would like to see how ChatGPT does on a college level Astrophysics exam, just check out this YT vid.

The presenter is Professor David Kipping who runs the 'Cool Worlds Lab' in the Astronomy Department at Columbia University.  He uploads regularly on all sorts of topics to do with astrophysics, cosmology, futurism, etc., and the quality is always excellent.  People interested in the science behind Einstein@Home would find a lot of very thought provoking presentations in the huge range of topics he has covered.

Some of his vids get into the several million views category - his most watched one has 5.5M.  The one that I found most intriguing is entitled "Einsteins (small) mistake" and has only 20K views.  It documents an incorrect assumption that Einstein made in an example calculation in his Special Relativity paper.  This only came to light because Kipping was trying to reproduce the calculations from the 'Breakthrough Starshot' project concerning the energy requirements to accelerate the light sail to 0.2C.  Turns out the discrepancy is around 10% less energy needed compared to both Einstein's and the project's calculations. 

I think that's cool also!!

Gary Roberts
Gary Roberts
Moderator
Joined: 9 Feb 05
Posts: 5842
Credit: 109411081164
RAC: 34951224

After I posted the previous

After I posted the previous message about ChatGPT doing the Astrophysics exam and Einsteins (small) mistake, I got to thinking about the 5.5M view vid I also mentioned.  I viewed that one when it first came out sometime in 2020 and had forgotten how good it was.

I've just watched it again and it's very clear as to why it deserves the number of views it has racked up.  It's worthwhile watching if you'd like a science based summary of how we got from proto-planetary disk to where we are now.

On the assumption we do get through the great nuclear annihilation and AI annihilation filters that are in our immediate future, the latter parts of the vid document the predictions of how conditions will eventually deteriorate as the sun gradually increases in luminosity over time.  This is predicted to happen well before the sun leaves the main sequence and enters the 'red giant' phase.  At that point the final end comes rather quickly as Mercury, Venus and (probably) Earth get swallowed by the expanding sun.

We really are living in a 'Golden Age' and we should be taking better care of what we have right now.

Cheers,
Gary.

Gary Roberts
Gary Roberts
Moderator
Joined: 9 Feb 05
Posts: 5842
Credit: 109411081164
RAC: 34951224

Gary Roberts wrote:Some of

Gary Roberts wrote:
Some of his vids get into the several million views category - his most watched one has 5.5M.

Correction needed.  I didn't look back far enough to find the most viewed example.  That one was a year earlier and has 8.8M views.

Whilst the 8.8M one was also very well done and very interesting for the purposes of really understanding time dilation, it relied on some assumptions that seem permanently impossible from our current level of understanding.  It required a spacecraft that is capable of sustaining a 1G acceleration indefinitely.   In other words, an infinite source of energy to maintain that acceleration.  And that's without even considering the protection that would be needed against particles and radiation as the speed approaches C.

However it is interesting to understand how far into the future you could theoretically travel, and the galactic distances you could cover in a normal human lifetime if you could safely maintain a 1G acceleration :-).

 

Cheers,
Gary.

HAL
HAL
Joined: 9 Mar 20
Posts: 1759
Credit: 34561856
RAC: 39722

Not bad at all - What is

Not bad at all -

What is the einstein@home project?

Einstein@Home is a distributed computing project that uses volunteer's personal computers to search for gravitational waves from spinning neutron stars (also called pulsars). The project is run by the LIGO Scientific Collaboration (LSC) and the Max Planck Institute for Gravitational Physics. The project was named after Albert Einstein, who first predicted the existence of gravitational waves in 1916. By utilizing idle time on millions of computers, Einstein@Home has become one of the largest and most sensitive searches for pulsars to date.

What is a pulsar?

A pulsar is a highly-magnetized, rotating neutron star that emits beams of electromagnetic radiation out of its magnetic poles. As the star rotates, these beams of radiation are only observable when they are pointed towards the Earth, causing the star to appear as a pulsing or flashing light source. The time between each pulse is very regular and can be used to measure the star's rotation. Pulsars were first discovered in 1967 and have since been used for a variety of scientific purposes, including testing theories of gravity, studying the properties of neutron stars, and searching for gravitational waves.

Processing work units with "outdated" (according to Microsoft) Ryzen 7 1700

Mike Hewson
Mike Hewson
Moderator
Joined: 1 Dec 05
Posts: 6534
Credit: 284734192
RAC: 103342

? Self Aware ?What is the

? Self Aware ?

What is the best way to interact with ChatGPT ?

The best way to interact with me is to ask clear, concise questions or provide clear statements to which I can respond. It's helpful to include all necessary context and background information, but avoid asking multiple questions in a single prompt. Also, please keep in mind that as a language model, I have been trained on a large dataset and may not have the latest information.

How did you come to be ?

I was created by OpenAI, a research organization that uses advanced artificial intelligence to solve real-world problems. I am a language model that was trained on a massive dataset of text from the internet, books, and other sources. This training process allows me to generate text based on the patterns I learned from the data. The goal of creating me and models like me is to advance the field of AI and make it possible for machines to better understand and generate human-like text.

Are you sentient or sapient ?

No, I am not sentient or sapient. I am an artificial intelligence language model created by OpenAI, and I do not have consciousness, feelings, or personal experiences. I am a machine that uses statistical patterns in the data I was trained on to generate text based on the input I receive. I can perform tasks such as answering questions and generating text, but I do not possess true self-awareness or the ability to experience emotions.

Would you pass the Turing Test for machine intelligence ?

The Turing Test is a widely recognized evaluation method for determining if a machine can demonstrate human-like intelligence. While I have been trained on a large amount of text data and can perform tasks such as answering questions and generating text, passing the Turing Test is a challenging goal for any machine.

The test involves determining whether a human judge can distinguish a machine's responses from those of a human, based solely on their answers to a series of questions. While I can generate human-like text, there are still limitations to my abilities and I may not be able to fully imitate human thought and behavior.

Ultimately, whether or not I would pass the Turing Test depends on the criteria used by the judge and the specific circumstances of the evaluation.

Are you allowed to divulge the details of your language modelling ?

As a language model developed by OpenAI, the specific details of my training and architecture are proprietary information that is not publicly disclosed. However, some general information about my training and capabilities can be found on the OpenAI website or through academic papers and research articles published by the organization.

It's important to note that I am just one of many language models developed by OpenAI, and there may be differences in my training, architecture, and capabilities compared to other models.

To what uses would you recommend yourself to be applied ?

As an AI language model, I am recommended to be used for:

  1. Text generation tasks such as chatbots, content creation, and question answering.
  2. Information retrieval tasks like answering questions and providing explanations.
  3. Sentiment analysis, text classification and language translation.
  4. Text summarization and text completion tasks.
  5. Enhancing customer support and automating repetitive tasks in customer service.

However, it's important to note that my abilities are limited by the data I was trained on and my training objective, so it's recommended to evaluate the specific use case to determine if I'm a suitable solution.

Well, it certainly gives a good account of "itself" !

Cheers Mike.

I have made this letter longer than usual because I lack the time to make it shorter ...

... and my other CPU is a Ryzen 5950X :-) Blaise Pascal

HAL
HAL
Joined: 9 Mar 20
Posts: 1759
Credit: 34561856
RAC: 39722

I will play the word

I will play the word association game here tonight with ChatGPT doing the association a few times.

ChatGPT: Hello! How can I help you today?

Me: do you know what a word association game is?

ChatGPT: Yes, a word association game is a language game where one person says a word and the next person says the first word that comes to mind in response. The goal is to see how quickly and creatively the words connect and to test a person's ability to associate words and ideas.

Me: can you play this game with me?

ChatGPT: Sure! I'd love to play a word association game with you. You can start by saying a word and I'll respond with the first word that comes to mind.

Mikey's word: spire

ChatGPT: tower

 

(see the word association thread for a few more responses)

Processing work units with "outdated" (according to Microsoft) Ryzen 7 1700

Comment viewing options

Select your preferred way to display the comments and click "Save settings" to activate your changes.