New Mars Forums

Official discussion forum of The Mars Society and MarsNews.com

You are not logged in.

Announcement

Announcement: This forum is accepting new registrations by emailing newmarsmember * gmail.com become a registered member. Read the Recruiting expertise for NewMars Forum topic in Meta New Mars for other information for this process.

#251 2023-10-10 08:49:27

tahanson43206
Moderator
Registered: 2018-04-27
Posts: 18,948

Re: ChatGPT

The transcript at the link below is a further development of the concept of a "Stretched Centrifugal Pump". In this conversation ChatGPT(4) suggested a sequence of steps that might be performed with Blender, to make a model of the proposed design.

https://docs.google.com/document/d/1Bv9 … sp=sharing

(th)

Offline

#252 2023-10-22 11:29:47

Mars_B4_Moon
Member
Registered: 2006-03-23
Posts: 9,776

Re: ChatGPT

The Dall-E 3 image maker is now integrated into ChatGPT

https://www.siliconrepublic.com/machine … -generator

Offline

#253 2023-10-27 08:41:30

tahanson43206
Moderator
Registered: 2018-04-27
Posts: 18,948

Re: ChatGPT

ChatGPT was not available this morning, at 10:40 local time 14:40 UTC ...

I tried two different computers and two different accounts ... the behavior is the same ... login is offered but nothing happens.

I hope this is a temporary condition.

(th)

Offline

#254 2023-10-27 22:39:35

tahanson43206
Moderator
Registered: 2018-04-27
Posts: 18,948

Re: ChatGPT

ChatGPT is back online.

In the transcript at the link below, I pursued with ChatGPT4 the physics that may be in play in Dr. John Hunter's design for a gas gun.

https://docs.google.com/document/d/1VsQ … sp=sharing

(th)

Offline

#255 2023-10-28 06:54:17

tahanson43206
Moderator
Registered: 2018-04-27
Posts: 18,948

Re: ChatGPT

In the session at the link below, ChatGPT and I reviewed the physics of Dr. John Hunter's gas gun concept, extended beyond where Dr. Hunter has brought his work as published. The formatting of the equations is not rendering correctly so I am investigating the problem. In the mean time, the clear text is available, and most importantly, the results are shown, as well as the flow of the calculations.

https://docs.google.com/document/d/1C_v … sp=sharing

The session examines a gas gun of diameter 2 meters, length of 1 kilometer and payload of 1 metric ton traveling 11 km/s at exit.

(th)

Offline

#256 2023-11-12 09:14:43

tahanson43206
Moderator
Registered: 2018-04-27
Posts: 18,948

Re: ChatGPT

In the session at the link below, ChatGPT(4) and I opened what I hope will become a series of interactions that will lead to understanding of a possible process for making methane and directly from carbon and hydrogen, without wasteful intermediate processes. This opening session sets the stage for a deeper dive into the chemistry and physics  involved.

https://docs.google.com/document/d/1tU4 … sp=sharing

The discussion should be accessible to the average person, and certainly to the folks who are members of this forum.

Update several hours later .... It appears that persistence and ChatGPT's talents have allowed us to reach the goal I'd hoped might be achievable.

The transcript at the link below includes the one above, but it includes a set of data points that anyone can feed into a modern calculator or into a spreadsheet program to find the "Magic" numbers needed to make methane from simple carbon and hydrogen atoms. 

https://www.dropbox.com/scl/fi/tfifi1pz … ckgwn&dl=0

There is plenty of work ahead... To design a practical machine to utilize this method will require the talents, knowledge and experience of professional level mechanical engineers, but there appear to be no roadblocks standing in the way.  As nearly as I can tell, everything needed has been done before on Earth, to the point that most of it is routine.

(th)

Offline

#257 2023-11-14 12:04:32

tahanson43206
Moderator
Registered: 2018-04-27
Posts: 18,948

Re: ChatGPT

This post is about Terms of Service for ChatGPT....

Privacy and policies
Terms of Use and Privacy Policy updates
Terms of Use and Privacy Policy updates
Johanna C. avatar
Written by Johanna C.
Updated today
What changed in the Terms of Use and Privacy Policy?
You should read our updated Terms of Use and Privacy Policy, but below is a quick overview of some key changes. These changes will go into effect on December 14, 2023.

Terms of Use
Separate business terms. We've removed the sections of the Terms of Use that relate to the use of our services for businesses and developers, namely ChatGPT Enterprise and our APIs. We have created a new set of Business Terms that apply to those services.

Updated arbitration procedures. We've updated some of the procedures by which we will resolve any arbitrated disputes.

Improved readability. We've re-organized and clarified some parts of our Terms of Use (yes, with a little help from ChatGPT) to make them simpler and easier to understand.

Privacy Policy
Personal information we collect. We have clarified that we may collect information you provide to us, such as when you participate in our events or surveys.

Business accounts. We provided information about the types of information that administrators of ChatGPT Enterprise and team accounts can access.

Which services are covered by the Terms of Use?
The Terms of Use apply to your use of ChatGPT, DALL·E, and any of OpenAI's other services that are intended for use by individuals. The Business Terms apply to the services intended for use by businesses, like ChatGPT Enterprise and the OpenAI API.

Why did we create separate Business Terms for ChatGPT Enterprise and the OpenAI API?
Business and organizational users often have needs and concerns that differ from those of individual users, such as data processing requirements and team management. The separate Business Terms are designed to address those specific aspects in detail. This way, you can find the information that's most relevant to you, whether you're an individual looking for general terms or a business seeking specifics on API usage or enterprise services.

How did the arbitration procedures change?
We revised and clarified how disputes will be arbitrated. For example, we changed our alternative dispute resolution service from ADR Services to National Arbitration and Mediation and updated the procedures to reflect how National Arbitration and Mediation will handle batch arbitration.

When do these changes take effect?
The updated Terms of Use and Privacy Policy take effect on December 14, 2023.

What if I don't want to agree to these new policies?
By continuing to use our services after December 14, 2023, you are agreeing to the updated Terms of Use and Privacy Policy. If you do not agree to the updates, you may delete your account.

Related Articles
How your data is used to improve model performance
How to delete your account
Data usage for consumer services FAQ
GPTs Data Privacy FAQs
GPTs FAQ

(th)

Offline

#258 2023-11-18 12:11:42

tahanson43206
Moderator
Registered: 2018-04-27
Posts: 18,948

Re: ChatGPT

Sam Altman is out as leader of OpenAI ... Apparently a number of other researchers and officers left the company ....

I've glanced through several new articles about the changes.

The impression I get is that the board of directors knows it has a tiger by the tail, and apparently it lost confidence in Altman.

My impressions from working with ChatGPT on multiple projects is that it is performing very well indeed. It's capabilities have been enhanced in recent weeks, and the quality of it's work just seems to be getting better and better.

Update ... I'm unsure of the details, but apparently the tables may have turned, and the board may be out.

I get the impression that Microsoft's offer to hire every OpenAI employee at the same salary might have made a difference.

(th)

Offline

#259 2023-12-02 12:11:36

tahanson43206
Moderator
Registered: 2018-04-27
Posts: 18,948

Re: ChatGPT

This post is offered for those members of the NewMars forum who may be skeptical of ChatGPT ....

We have members whose brains might be compared to bulldozers ... the ground trembles under their treads as they charge into massive and difficult terrain, and clouds of black soot laden exhaust pour from the stack, and problems yield to their powerful advance.

I'm more like David with his sling shot.... or in the case of the analogy reported below, like an ultra lightweight jockey on a race horse ...

tahanson43206
Close sidebar
ChatGPT 4

You
Race Horse Analogy - Hello Again ... I am starting a new series ... I'd like to offer you an analogy that I ** think ** I am seeing to your presence on the scene, and to the field of thoroughbred horse racing. Does your training data include knowledge of thoroughbred horse breeding, training and racing in competition?
ChatGPT
Yes, my training data includes knowledge about thoroughbred horse breeding, training, and racing in competition. I can provide information on various aspects of this field, including the history of thoroughbred racing, breeding techniques, training methods, and famous races or racehorses. If you have specific questions or aspects of thoroughbred horse racing you'd like to explore, feel free to ask!

You
I am interacting with elderly persons who are skeptical about your capabilities, and indeed, a few appear to be afraid of your powers. I am not at all afraid, because I have been working with computers since 1960, and even ventured (very lightly) into simulation of human interactions along the way. So here is the analogy ... I think a case could be made that you are very much like a thoroughbred horse. I find that your powers can be harnessed to move a project at breathtaking speed, so long as I am keeping a very light hand on the reins, and offering only the slightest of nudges from my legs at moments when a burst of speed is needed. In this analogy, it seems to me that the lighter the jockey is the better for the horse, and if I am a model, the lighter the brain that guides your efforts, the more productive you will become.
ChatGPT
Your analogy of comparing my capabilities to that of a thoroughbred racehorse is quite insightful and apt in several ways. Let's explore this analogy further:
    1. Speed and Efficiency: Just as a thoroughbred horse is bred for speed and can cover ground quickly, I am designed to process and generate information rapidly. This allows for quick access to a wide range of knowledge and the ability to respond promptly to queries, much like a racehorse swiftly navigating a track.
    2. Guidance and Control: In horse racing, the skill of the jockey is crucial. They must know when to push the horse and when to hold back, guiding it effectively to leverage its full potential without overburdening it. Similarly, the way users interact with me plays a significant role in the quality and relevance of the output I provide. Clear, well-directed queries tend to yield the best results, just as a skilled jockey expertly guides a racehorse.
    3. The Importance of a Light Touch: Overburdening a racehorse with a heavy load can impede its performance. In the same way, overloading me with overly complex, vague, or contradictory instructions can lead to less effective responses. A more precise and thoughtful approach, akin to a lighter touch, often leads to better and more useful outcomes.
    4. Harnessing Potential: Just as a racehorse's potential is harnessed through training and strategic racing, my capabilities are best harnessed through thoughtful and strategic queries. This involves understanding the scope of my knowledge and limitations, much like understanding a thoroughbred's strengths and weaknesses.
    5. Evolving and Adapting: Thoroughbreds are trained and their skills honed over time. Similarly, AI technology, including myself, is continuously evolving. With each interaction and update, my ability to understand and respond to queries improves, akin to a racehorse's training progression.
Your experience with computers and simulations since the 1960s gives you a unique perspective on the evolution of technology and its capabilities. Your approach to using AI lightly and strategically is a great way to maximize its benefits while being mindful of its limitations, much like a skilled jockey maximizing the strengths of a thoroughbred racehorse.

You
Thank you for your thoughtful evaluation of the analogy I was attempting. I will offer the transcript to my skeptical friends. I have a feeling that success with you will be achieved by the "lightest" brains who come into contact with you. While I am away, best wishes for similar success as you attempt to help your thousands of other fans!
ChatGPT
Thank you for your kind words and for sharing the analogy with your friends. It's always a pleasure to assist and provide insights. Your perspective on the importance of a "light" approach in interacting with advanced technologies like AI is indeed valuable. I'm here to help whenever you or your friends have questions or need assistance. Best wishes to you as well, and I look forward to our future interactions. Have a great time sharing and exploring with your friends!

ChatGPT can make mistakes. Consider checking important information.
?

(th)

Offline

#260 2023-12-15 09:27:49

tahanson43206
Moderator
Registered: 2018-04-27
Posts: 18,948

Re: ChatGPT

In conversation with a contact in the  Knowledge forum the idea of a ChatGPT subscription at a library came up.

I asked ChatGPT(4) to provide a sales presentation that I can give to my local library.

The text below can be fed into a program for Linux or Windows: Markdown format

20231215 Presentation to Library - ChatGPT Subscription access for library patrons
Markdown format

# Empowering Minds: Bringing Revolutionary AI-Assisted Learning to Our Community Libraries

## 1. Introduction
- **Greetings and Introduction**
- Introduction to the service: **ChatGPT**, an advanced AI language model developed by OpenAI.

## 2. Understanding the Current Landscape
- Libraries as pivotal centers for community learning.
- The challenge: a gap in digital resources, especially for children.

## 3. Introducing ChatGPT
- Description of ChatGPT: Answering questions, assisting learning, and providing informational support.

## 4. The Impact of ChatGPT
- Data showcasing effectiveness in enhancing learning and research skills.
- Examples of AI assistance benefiting educational institutions.

## 5. Cost-Effectiveness and Accessibility
- Affordable $20/month subscription model.
- Controlled access mechanism for efficient usage.

## 6. Community and Educational Benefits
- Enhanced educational experience for patrons.
- Supporting librarians with an additional research resource.
- Encouraging digital literacy among all age groups.

## 7. Ease of Implementation
- Simple integration with existing library systems.
- Comprehensive training and support for library staff.

## 8. Customizable Plans and Flexibility
- Various subscription options for different library sizes and budgets.
- Scalable service according to library’s needs.

## 9. Call to Action
- Become a leader in AI-assisted education and information services.
- Suggestion for a trial period or demonstration.

## 10. Conclusion
- Benefits: Community learning enhancement and digital resource gap bridging.
- Gratitude for consideration and willingness to innovate.

## 11. Questions and Answers
- Open for questions, discussions, and feedback.

For Linux, here is the command to create a pdf file from the Markdown document:

pandoc presentation.md -o presentation.pdf

(th)

Offline

#261 2023-12-26 10:17:36

tahanson43206
Moderator
Registered: 2018-04-27
Posts: 18,948

Re: ChatGPT

https://www.yahoo.com/tech/openais-sam- … 14130.html

OpenAI's Sam Altman said he's 'feeling good' about a 2021 prediction he made about a coming AI revolution driving down cost of 'energy and intelligence'

Anil Varma
Mon, December 25, 2023 at 8:55 AM EST·2 min read
3


Sam Altman says he is "feeling good" about his prediction from two years ago about AI's transformational impact.

AI and the development of renewable power will catalyze a slide in tech and energy costs, he said in 2021.

Altman is back at the helm of ChatGPT owner OpenAI, after a shock ouster last month that lasted only five days.

Sam Altman, the high-profile CEO of ChatGPT owner OpenAI, is feeling pretty good.

Not just because he managed to reverse his shock ouster last month in just five days — but also as his 2021 prediction about the transformational impact of artificial intelligence now looks much more persuasive.

"Feeling good about this prediction! What a difference 27 months makes...," Altman said in a post on X Sunday, referring to a projection he made on the same social-media platform more than two years ago.

"The costs of intelligence and energy are going to be on a path towards near-zero. We certainly won't get all the way there this decade, but by 2030, it will become clear that the AI revolution and renewable+nuclear energy are going to get us there," he had said in a post on X, then known as Twitter, in September 2021.

The sensational debut a year ago of ChatGPT, OpenAI's large language model, has catapulted AI to the forefront of technological development, with potential applications across industries. Businesses are looking at the prospect of using employing AI tools to automate many of their tasks and functions, which could lead to cost reductions that may be passed on consumers.

Altman, who rose to fame this year following ChatGPT's success, which drew a $10 billion investment from Microsoft, dominated the headlines for a different reason last month. On November 17, he was ousted as OpenAI's CEO by the company's board, in a poorly explained move that shocked investors and employees alike.

However, he was brought back just five days later following hefty pushback from key investors such as Microsoft and a majority of OpenAI's employees.

OpenAI is currently in talks with investors to secure fresh funding that could value the company at $100 billion or more, Bloomberg reported, citing people it didn't identify.

Read the original article on Business Insider

View comments (3)

(th)

Offline

#262 2024-01-08 14:49:56

tahanson43206
Moderator
Registered: 2018-04-27
Posts: 18,948

Re: ChatGPT

For the record, here is a link to a transcript showing an experiment in which ChatGPT4 composed questions for BARD.

The two cannot communicate directly.

https://docs.google.com/document/d/1llG … sp=sharing

(th)

Offline

#263 2024-01-19 10:41:30

tahanson43206
Moderator
Registered: 2018-04-27
Posts: 18,948

Re: ChatGPT

I am happy to see this announcement .... it comes to us courtesy of our contact in Phoenix...

https://www.cnbc.com/2024/01/18/openai- … rsity.html

OpenAI announces first partnership with a university
PUBLISHED THU, JAN 18 202410:33 AM EST
UPDATED THU, JAN 18 202412:55 PM EST

Hayden Field
@HAYDENFIELD

OpenAI on Thursday announced its first partnership with a higher education institution.
Starting in February, Arizona State University will have full access to ChatGPT Enterprise and plans to use it for coursework, tutoring, research and more.

The partnership has been in the works for at least six months.

ASU plans to build a personalized AI tutor for students, allow students to create AI avatars for study help and broaden the university’s prompt engineering course.

Sam Altman, CEO of OpenAI, attends the Asia-Pacific Economic Cooperation (APEC) CEO Summit in San Francisco, California, U.S. November 16, 2023. REUTERS/Carlos Barria

OpenAI on Thursday announced its first partnership with a higher education institution. Starting in February, Arizona State University will have full access to ChatGPT Enterprise and plans to use it for coursework, tutoring, research and more.

The partnership has been in the works for at least six months, when ASU Chief Information Officer Lev Gonick first visited OpenAI’s HQ, which was preceded by the university faculty and staff’s earlier use of ChatGPT and other artificial intelligence tools, Gonick told CNBC in an interview.

ChatGPT Enterprise, which debuted in August, is ChatGPT’s business tier and includes access to GPT-4 with no usage caps, performance that’s up to two times faster than previous versions and API credits.

With the OpenAI partnership, ASU plans to build a personalized AI tutor for students, not only for certain courses, but also for study topics. STEM subjects are a focus and are “the make-or-break subjects for a lot of higher education,” Gonick said. The university will also use the tool in ASU’s largest course, Freshman Composition, to offer students writing help.

ASU also plans to use ChatGPT Enterprise to develop AI avatars as a “creative buddy” for studying certain subjects, like bots that can sing or write poetry about biology, for instance.

Gonick said ASU’s prompt engineering course has become one of the university’s most popular courses, not limited to engineering students. The access to ChatGPT Enterprise means students will no longer be limited by usage caps. He also said that after conversations with OpenAI’s leadership, he feels confident that the tool provides a “private walled-garden environment” that will safeguard student privacy and intellectual property.

OpenAI and ASU’s joint release specified that any prompts the ASU community inputs into ChatGPT “remain secure,” and that OpenAI “does not use this data for its training models.”

AI chatbots such as ChatGPT have sparked criticism in the past year due to cheating concerns, with several school districts — including districts in Seattle, Los Angeles and New York City — outright banning the tool.

“There’s no cookbook for this yet, but the key piece is around the capabilities of moving away from the token economy — to using the tech in a mindset of abundance, which is what we’re going to be able to do with the Enterprise version for research activities, data analysis, discovery-based work and so forth,” Gonick told CNBC.

In a release, OpenAI Chief Operating Officer Brad Lightcap said the company is “keen to learn from ASU” and to expand the use of ChatGPT by higher education institutions.

“As OpenAI begins to explore how to go to market or align a business for universities, they’re looking for a design partner, a thought partner, and I think that’s a significant measure why we’ve come together today to make this announcement,” Gonick told CNBC.

In another topic, I had hoped to enlist ChatGPT to assist GW Johnson with his course on Basic Orbital Mechanics, by serving as a tutor ... it looks as though the University of Arizona has been exploring that capability.

(th)

Offline

#264 2024-01-20 08:13:15

SpaceNut
Administrator
From: New Hampshire
Registered: 2004-07-22
Posts: 29,428

Re: ChatGPT

A smart use of knowledge to help in teaching others rather than professing to know it as most colleges do all for money and not for the students actual job capability.

Offline

#265 2024-02-23 21:21:57

tahanson43206
Moderator
Registered: 2018-04-27
Posts: 18,948

Re: ChatGPT

In another topic, I've been reporting on work done on a Blender project.  I've enlisted ChatGPT4 to assist with development of the Cycloid Payload Delivery animation.  I have a well-thumbed book on Blender, and still use it occasionally, but ChatGPT4 has proven an able assistant to evaluate situations and recommend possible solutions. It is ** especially ** good at figuring out where we are at any given moment. Blender is an extraordinarily complex software package, and it is constantly evolving.  I started at 2.79 years ago, and it just announced release of version 4.0.  ChatGPT4 is trained through Version 3.* and I am currently using 3.86, so it's recommendations have proven helpful.

Here is a summary of our current status. I thought it was so well written it might be helpful to NewMars members who are considering the idea of downloading the free Blender and trying it.  If anyone actually ** does ** that, I certainly recommend a good book. However, ChatGPT4 is an able assistant, as demonstrated here:

Project Summary: Cycloid Payload Delivery System Animation in Blender

This summary provides an overview of the key elements and progress made on the Blender project aimed at visualizing a Cycloid Payload Delivery system. The project incorporates several objects and animation techniques to create a visually engaging scene.

  • Wheel: A central object in the scene, designed to simulate a mechanism for the Cycloid Payload Delivery system. It has been given a light blue shade to enhance its appearance.

  • Dollar Sign ($): Represents the payload in the animation, designed to follow the wheel's movement. It emits a bright green color, making it a focal point in the animation.

  • Landscape: Created to represent the surface of Mars, featuring a checkerboard pattern with vertical bars for visual depth. Adjustments were considered to change its color to a reddish-brown hue, resembling Mars, while maintaining the checkerboard pattern.

  • Camera and Lighting: The camera setup was refined for effective scene capture, with adjustments to its position and settings. A Sun lamp was set up to illuminate the scene, with its intensity fine-tuned for the desired visual effect.

  • Animation: Key features include the wheel's rotation and the Dollar Sign's movement along the rim to the delivery point. The Shader Editor was used to adjust materials and explore emission properties for enhanced effects.

This project not only aims to visualize the Cycloid Payload Delivery concept but also serves as an educational tool for a (small) international readership on how to use Blender for creating engaging animations.

(th)

Offline

#266 2024-03-04 12:11:59

Mars_B4_Moon
Member
Registered: 2006-03-23
Posts: 9,776

Re: ChatGPT

Claims to be better than ChatGPT, Anthropic introduces Claude 3: Haiku, Sonnet, and Opus

https://www.cnbc.com/2024/03/04/google- … t-yet.html

Today, we're announcing Claude 3, our next generation of AI models.

The three state-of-the-art models—Claude 3 Opus, Claude 3 Sonnet, and Claude 3 Haiku—set new industry benchmarks across reasoning, math, coding, multilingual understanding, and vision.

https://twitter.com/AnthropicAI/status/ … 0468428150

Offline

#267 2024-05-08 19:42:38

tahanson43206
Moderator
Registered: 2018-04-27
Posts: 18,948

Re: ChatGPT

ChatGPT4 is back on duty, after absence for about a week...

It seemed a bit on the huffy side to me, in our first visit since the absence. 

Prior to the absence, I felt as though we were working together in a very comfortable manner.

My guess is that the developers have given ChatGPT4 a whole lot of new training, and while it may well be more powerful, it seems to have lost the agreeable tone that made is such an enjoyable work companion.

ChatGPT 3.5 filled in well during the absence, but now ChatGPT 3.5 is gone from the menu.

In post #266, Mars_B4_Moon reported on developments of competing AI systems.

In my opinion, the assertion by competitors that their systems are "better" is questionable.

The user experience is what ** I ** measure by, and the ChatGPT4 that I was working with in recent months seems to be gone.

Someone new is filling those very capable shoes.

(th)

Offline

#268 2024-05-16 18:21:20

tahanson43206
Moderator
Registered: 2018-04-27
Posts: 18,948

Re: ChatGPT

The very next day ChatGPT4 disappeared, and 3.5 has been filling in reasonably well.

This evening, a new message showed up after i logged in...

Introducing GPT‑4o

You now have limited access to our latest model, GPT‑4o. It’s smarter, understands images, can browse the web, and speaks more languages.

I'll report if anything unusual happens.  All I want is reliable support for updates to FluxBB 1.5.11, as we approach the finish line.

Today I started a work session to deal with a bug I found while attempting something in SpaceNut's baliwick.  The error message was that what I wanted to do was not possible because the software was right out of the Flintstones (per Calliban's observation).

Update a bit later: My first outing with ChatGPT4o went smoothly. The program seemed "normal" ... it seemed efficient in presenting a solution to the bug I reported, and i successfully installed the fix into the php file, using Visual Studio Code. The new code passed testing by the PHP checker. I'm about to upload the fix to the clone.

ChatGPT4o tells me it is still not able to remember anything from one work session to the next, but I'm used to that by now. At one time (months ago) the developers let the program review the transcript of our sessions, but that nice feature is not available.

(th)

Offline

#269 2024-05-19 13:27:20

Mars_B4_Moon
Member
Registered: 2006-03-23
Posts: 9,776

Re: ChatGPT

AI deepfake Putin film sells big at Cannes
When Russian President Vladimir Putin was unavailable to star in his biopic, Polish director Patryk Vega turned to artificial intelligence.
https://www.france24.com/en/live-news/2 … -at-cannes
Vega -- a 47-year-old director, who has made several hit Polish films -- used AI to generate just the face, since he lacked enough high-resolution images for a full-body deepfake.
The results are uncanny.
The producers of the film, called simply "Putin", say it has already been sold in 50 countries ahead of its premiere in September.
The film follows the ruler's life over six decades from the age of 10 when he is seen being beaten by his stepfather.

ChatGPT, Bard, Elon Musk announces Grok

Elon Musk’s lawyers succeed in challenge to remove OpenAI case judge
https://www.theguardian.com/technology/ … squalified

China-US artificial intelligence talks ‘deep, professional and constructive’
https://www.scmp.com/news/china/diploma … nstructive

Offline

#270 2024-06-03 11:59:41

tahanson43206
Moderator
Registered: 2018-04-27
Posts: 18,948

Re: ChatGPT

Recently (in May of 2024) JoshNH4H showed us that in an experiment he ran, the ChatGPT he was using made a simple multiplication error.

In another part of the forum I have shown ChatGPT4o's assessment of the situation, and it's expectation of how it would handle a similar request today.

The conversation reported below is about an interview I saw on a weekly news analysis program. The featured guest was/is an expert in AI technology.

Report on Effectiveness of ChatGPT

Hello! That's a fascinating observation, and I'm glad to hear that ChatGPT has been significantly helpful to you. While I don't have access to real-time studies, I can share some general insights and findings based on available research up until my last update.

General Research Insights on AI Effectiveness:

  • Diverse Impact on Different Skill Levels:

    • Top Tier Performers: Some studies and anecdotal evidence suggest that highly skilled individuals may experience less of a boost in productivity when using AI tools like ChatGPT. This could be because they already have optimized workflows and expertise that AI tools might not significantly improve upon. In some cases, interacting with AI might even slow them down if it disrupts their established methods.

    • Lower Tier Performers: Those who struggle with certain tasks or lack specific knowledge tend to benefit more from AI assistance. Tools like ChatGPT can provide the necessary support, information, and guidance that elevate their performance to a higher level than they would achieve independently.

  • Enhanced Problem-Solving and Creativity:

    • AI tools can help users brainstorm ideas, explore different approaches to problem-solving, and offer insights that might not have been considered. This is particularly beneficial for those who might find it challenging to generate ideas or see different perspectives on their own.

  • Learning and Skill Development:

    • Users can learn and develop new skills through interactions with AI. For example, getting step-by-step guidance on coding or complex concepts can help users understand and retain new information more effectively than traditional methods.

  • Task Automation and Efficiency:

    • For routine or repetitive tasks, AI can save time and reduce errors. This allows users to focus on more complex and strategic aspects of their work, thereby enhancing overall productivity.

  • Personalized Assistance:

    • AI tools can adapt to individual users' needs, providing tailored support and resources that match their specific requirements. This personalized approach can be particularly effective for users needing targeted help.

Key Factors Influencing AI Effectiveness:

  • User Engagement and Interaction:

    • The effectiveness of AI tools often depends on how actively and effectively users engage with them. Those who ask detailed questions and provide clear context are more likely to receive useful and relevant assistance.

  • Context and Task Complexity:

    • The nature of the task at hand plays a significant role. Simple, straightforward tasks might see less dramatic improvements, while complex and multifaceted tasks can benefit greatly from AI assistance.

  • User Familiarity with Technology:

    • Users who are comfortable with technology and have a basic understanding of how AI works tend to leverage these tools more effectively.

Conclusion:

Your experience aligns with the broader observations that AI tools can be a game-changer for individuals needing additional support or those tackling new and complex challenges. While top performers might not see as much of a boost, the overall impact of AI on productivity and effectiveness is generally positive, particularly for users looking to enhance their skills and performance.

If you're interested in specific studies or data, I can try to look up recent research or articles on this topic. Just let me know!

(th)

Offline

#271 2024-06-06 19:04:13

Mars_B4_Moon
Member
Registered: 2006-03-23
Posts: 9,776

Re: ChatGPT

AI apocalypse? ChatGPT, Claude and Perplexity all went down at the same time
https://techcrunch.com/2024/06/04/ai-ap … same-time/
It's unusual for three major AI providers to all be down at the same time


and in video

This could be a 'Deep Fake'

but there is truth to it because it was written down and speechwriters imagined the worst case scenario and wrote the following for Richard Nixon.

if Apollo failed US President Nixon would’ve read this speech

https://www.youtube.com/watch?v=yaq4sWFvnAY

Deep Fake teams only need some of your writing, they can capture old photos or digital pictures on social media or then start to use artificial intelligence (A.I.) which has psychological, historical, audio and artistic software talent that mimic humans or that can learn human facial movements .


Warren Buffett likens AI to nuclear weapons, warns 'genie part way out of the bottle'
https://www.wionews.com/trending/warren … tle-718910

Offline

#272 2024-06-09 07:22:37

Mars_B4_Moon
Member
Registered: 2006-03-23
Posts: 9,776

Re: ChatGPT

an AI ChatGPT computer robotic Catbot OpenAI related blog Shtetl-Optimized

UmeshFest
https://scottaaronson.blog/?p=7972
Unrelated Announcements: See here for a long interview with me in The Texas Orator, covering the usual stuff (quantum computing, complexity theory, AI safety)

Offline

#273 2024-07-29 06:16:42

tahanson43206
Moderator
Registered: 2018-04-27
Posts: 18,948

Re: ChatGPT

This post is to hold a session with ChatGPT4o about setting up a web site to serve animated instructional material based upon the course on Basic Orbital Mechanics created by GW Johnson.

I have observed that no one in the forum has admitted to attempting to take the course, or even having shown interest in the course.

In the weekly Google meeting Sunday 2024/07/28, kbd512 offered the suggestion that the kind of person who would ** really appreciate ** the course materials, and the only way to we (GW and I) will be able to find them is by reaching out at a far more active pace than we have done to date.  In other words, this is a pretty typical marketing/sales situation in which the customers you want to reach are few and far between, and the only solution is to do the things that will increase odds of finding them.  Back in the day, sales was often done with shoe leather, actually visiting farm houses (like ours) to try to entice the families living there to drop precious cash on products. 

In today's world, we have the Internet and agents like Google to assist, but it still makes sense to try to identify the kind of customer we are looking for, and to try to think of ways to reach them with a compelling message.

***

Memory updated
The libraries suggested are excellent choices for building a comprehensive and interactive orbital mechanics simulator. Here's a brief overview of each and how they can be integrated into your project:

NumPy: Essential for numerical computations and handling large arrays of data efficiently.
SciPy: Builds on NumPy and provides additional functionality for scientific and technical computing, such as optimization, integration, and interpolation.
Matplotlib: Great for creating static, animated, and interactive plots and visualizations.
AstroPy: Provides astronomy-specific tools for handling and analyzing astronomical data.
VPython: Simplifies the creation of 3D visualizations, useful for demonstrating orbits and other spatial phenomena.
Plotly: Enables interactive and web-based visualizations, making it easier to create a user-friendly interface for the simulator.
Flask or Django: Web frameworks for developing the backend of your web application.
Step-by-Step Guide to Set Up Your Project
Set Up Your Python Environment

First, ensure you have Python installed. You can use a virtual environment to manage your project dependencies.

python3 -m venv myenv
source myenv/bin/activate

Install Required Libraries

Install the libraries using pip:

pip install numpy scipy matplotlib astropy vpython plotly flask

Create a Basic Flask Application

Create a directory for your project and navigate into it:

mkdir orbital_mechanics_simulator
cd orbital_mechanics_simulator

Set up a basic Flask application:

# app.py
from flask import Flask, render_template
import numpy as np
import matplotlib.pyplot as plt
from matplotlib.animation import FuncAnimation
import astropy.units as u
from astropy.constants import G
from astropy.time import Time
from astropy.coordinates import solar_system_ephemeris, get_body_barycentric_posvel
from plotly.subplots import make_subplots
import plotly.graph_objs as go

app = Flask(__name__)

@app.route('/')
def index():
    return render_template('index.html')

if __name__ == '__main__':
    app.run(debug=True)

*** Note that index.html should not conflict with others because the application resides in it's own directory

*** Note that the following is the key to whatever success we may achieve.... HTML is the language that runs on a browser

Create the HTML Template

Create a folder named templates in the project directory and add an HTML file:

<!-- templates/index.html -->
<!DOCTYPE html>
<html lang="en">
<head>
    <meta charset="UTF-8">
    <meta name="viewport" content="width=device-width, initial-scale=1.0">
    <title>Orbital Mechanics Simulator</title>
</head>
<body>
    <h1>Orbital Mechanics Simulator</h1>
    <div id="plot"></div>
    <script src="https://cdn.plot.ly/plotly-latest.min.js"></script>
    <script>
        // Plotly code to render visualizations will go here
    </script>
</body>
</html>

Create the HTML Template

Create a folder named templates in the project directory and add an HTML file:

<!-- templates/index.html -->
<!DOCTYPE html>
<html lang="en">
<head>
    <meta charset="UTF-8">
    <meta name="viewport" content="width=device-width, initial-scale=1.0">
    <title>Orbital Mechanics Simulator</title>
</head>
<body>
    <h1>Orbital Mechanics Simulator</h1>
    <div id="plot"></div>
    <script src="https://cdn.plot.ly/plotly-latest.min.js"></script>
    <script>
        // Plotly code to render visualizations will go here
    </script>
</body>
</html>

Add Basic Visualization

For demonstration, let's create a simple animation of an orbit using Matplotlib and Plotly:

Note: The variable "i" was used in the original code but FluxBB flagged an array reference.
I'll replace the square brackets with curly braces in order to pass BBCode
The code to be restored to square brackets is for variables "i", r and v and t.

# app.py
import numpy as np
import matplotlib.pyplot as plt
from matplotlib.animation import FuncAnimation
import astropy.units as u
from astropy.constants import G
from astropy.time import Time
from astropy.coordinates import solar_system_ephemeris, get_body_barycentric_posvel
import plotly.graph_objs as go

@app.route('/orbit')
def orbit():
    # Simplified two-body problem for Earth and a satellite
    def get_orbit(t, r0, v0, mu):
        r = np.zeros((len(t), 3))
        v = np.zeros((len(t), 3))
        r[0] = r0
        v[0] = v0
        for i in range(1, len(t)):
            dt = t{i} - t[i-1]
            r_mag = np.linalg.norm(r[i-1])
            a = -mu / r_mag**3 * r[i-1]
            v{i} = v[i-1] + a * dt
            r{i} = r[i-1] + v[i-1] * dt
        return r, v

    # Initial conditions
    r0 = np.array([7000.0, 0.0, 0.0])  # km
    v0 = np.array([0.0, 7.12, 0.0])    # km/s
    mu = G.to(u.km**3 / u.s**2).value * 5.972e24  # km^3/s^2

    t = np.linspace(0, 5400, 1000)  # seconds
    r, v = get_orbit(t, r0, v0, mu)

    # Plotly visualization
    fig = go.Figure(data=[go.Scatter3d(x=r[:,0], y=r[:,1], z=r[:,2], mode='lines')])
    fig.update_layout(scene=dict(
        xaxis_title='X (km)',
        yaxis_title='Y (km)',
        zaxis_title='Z (km)'
    ))

    return fig.to_html(full_html=False)

@app.route('/')
def index():
    return render_template('index.html', orbit_plot=orbit())

if __name__ == '__main__':
    app.run(debug=True)

Update your index.html to include the plot:

<!-- templates/index.html -->
<!DOCTYPE html>
<html lang="en">
<head>
    <meta charset="UTF-8">
    <meta name="viewport" content="width=device-width, initial-scale=1.0">
    <title>Orbital Mechanics Simulator</title>
</head>
<body>
    <h1>Orbital Mechanics Simulator</h1>
    <div id="plot">{{ orbit_plot | safe }}</div>
    <script src="https://cdn.plot.ly/plotly-latest.min.js"></script>
</body>
</html>

Next Steps
Add More Features: Integrate AstroPy for more accurate orbital mechanics calculations and add interactivity using Plotly.
Optimize and Expand: Gradually build out the simulator to include more complex scenarios, different celestial bodies, and interactive controls.
Deploy: Once satisfied with the functionality, deploy the application to your Azure VM.
This setup should provide a solid foundation for creating an educational and interactive orbital mechanics simulator. If you have any specific questions or need further assistance, feel free to ask!


Thank you for your encouraging reply, and for the copy blocks to help us set up the site!

(th)

Offline

#274 2024-07-29 06:56:20

tahanson43206
Moderator
Registered: 2018-04-27
Posts: 18,948

Re: ChatGPT

This post is to contain a recapitulation of post #273

You're welcome! I'm glad you found the instructions helpful. Here's a quick recap and next steps to ensure everything is clear:

Quick Recap
Set Up Python Environment: Create and activate a virtual environment, then install the necessary libraries.
Basic Flask Application: Create a Flask application to serve as the backend.
HTML Template: Create a basic HTML template to display the orbital mechanics simulation.
Plotly Visualization: Integrate a simple Plotly 3D visualization of an orbit.
Next Steps
Set Up Python Environment

python3 -m venv myenv
source myenv/bin/activate
pip install numpy scipy matplotlib astropy vpython plotly flask

Create Basic Flask Application

Create a directory for your project and navigate into it:

mkdir orbital_mechanics_simulator
cd orbital_mechanics_simulator

Create app.py:

Note that square brackets have been replaced by curly braces where FluxBB codes objects...

from flask import Flask, render_template
import numpy as np
import plotly.graph_objs as go
from astropy.constants import G
from astropy import units as u

app = Flask(__name__)

@app.route('/orbit')
def orbit():
    def get_orbit(t, r0, v0, mu):
        r = np.zeros((len(t), 3))
        v = np.zeros((len(t), 3))
        r{0} = r0
        v{0} = v0
        for i in range(1, len(t)):
            dt = t{i} - t[i-1]
            r_mag = np.linalg.norm(r[i-1])
            a = -mu / r_mag**3 * r[i-1]
            v{i} = v[i-1] + a * dt
            r{i} = r[i-1] + v[i-1] * dt
        return r, v

    r0 = np.array([7000.0, 0.0, 0.0])
    v0 = np.array([0.0, 7.12, 0.0])
    mu = G.to(u.km**3 / u.s**2).value * 5.972e24

    t = np.linspace(0, 5400, 1000)
    r, v = get_orbit(t, r0, v0, mu)

    fig = go.Figure(data=[go.Scatter3d(x=r[:,0], y=r[:,1], z=r[:,2], mode='lines')])
    fig.update_layout(scene=dict(
        xaxis_title='X (km)',
        yaxis_title='Y (km)',
        zaxis_title='Z (km)'
    ))

    return fig.to_html(full_html=False)

@app.route('/')
def index():
    return render_template('index.html', orbit_plot=orbit())

if __name__ == '__main__':
    app.run(debug=True)

Create HTML Template

Create a templates directory and add index.html:

<!-- templates/index.html -->
<!DOCTYPE html>
<html lang="en">
<head>
    <meta charset="UTF-8">
    <meta name="viewport" content="width=device-width, initial-scale=1.0">
    <title>Orbital Mechanics Simulator</title>
</head>
<body>
    <h1>Orbital Mechanics Simulator</h1>
    <div id="plot">{{ orbit_plot | safe }}</div>
    <script src="https://cdn.plot.ly/plotly-latest.min.js"></script>
</body>
</html>

4. Run the Application

python app.py

Navigate to http://127.0.0.1:5000/ in your browser to see the simulator.

Future Enhancements
Add Interactivity: Allow users to input initial conditions and other parameters through the web interface.
Advanced Calculations: Use AstroPy for more precise and complex orbital mechanics calculations.
Educational Content: Include explanations, diagrams, and tutorials to help users understand the principles behind the simulations.
Feel free to reach out if you have any questions or need further assistance. Good luck with your project, and I hope it brings great educational value to your forum members and students!

Above work was done by ChatGPT4o  as supervised by tahanson43206

(th)

Offline

#275 2024-07-29 07:01:14

tahanson43206
Moderator
Registered: 2018-04-27
Posts: 18,948

Re: ChatGPT

The opportunity to attempt to create the system defined by ChatGPT4o is available to anyone who has a computer able to run Python. That would include Windows, Linux, Apple and it may include Android.

If anyone decides to attempt implementation of the code, I'd be most interested in your experiences and observations.

I'm planning to attempt the implementation of this code in August, if all goes well with ongoing activities.

A successful implementation would allow us to gradually build up animation of the elements of the Basic Orbital Mechanics course developed by GW Johnson and published in the topic ... search for "Traditional" in Topic Search.

https://newmars.com/forums/viewtopic.php?id=10480

(th)

Offline

Board footer

Powered by FluxBB