A quick conversation with ChatGPT
|
01-20-2023, 02:50 PM
Post: #1
|
|||
|
|||
A quick conversation with ChatGPT
My short conversation with ChatGPT
Me: What are the advantages and disadvantages of pocket calculators using reverse polish notation entry model? ChatGPT: Advantages of pocket calculators using reverse polish notation (RPN) entry model include: Faster and more efficient data entry: In RPN, the operator is entered after the operands, which eliminates the need for parentheses and reduces the number of keystrokes needed to perform a calculation. Easier to use for complex mathematical operations: RPN is well-suited for use with stack-based operations, which are commonly used in advanced mathematical and scientific calculations. More accurate results: Because RPN eliminates the need for parentheses, it reduces the possibility of operator precedence errors. Disadvantages of pocket calculators using RPN entry model include: Steep learning curve: RPN can be difficult for users who are not familiar with the notation, making it less accessible to casual users. Limited compatibility with other calculators: RPN is not as widely used as traditional infix notation, so it may not be compatible with other calculators or software. Less intuitive: RPN can be less intuitive for users who are used to the traditional infix notation. Me: What can you tell my about the HP-41 calculator that was popular in the Eighties? ChatGPT: The HP-41C was a popular programmable calculator produced by Hewlett-Packard in the late 1970s and early 1980s. It was known for its advanced features and versatility, as well as its long battery life. Some of the key features of the HP-41C include: Programmability: The HP-41C was fully programmable, allowing users to create and save their own custom programs. This made it useful for a wide range of applications, including engineering, scientific, and financial calculations. Expandability: The HP-41C had a modular design, which allowed users to add additional hardware and software modules to expand its functionality. These modules included things like a printer, a magnetic card reader/writer, and a barcode reader. Advanced mathematical functions: The HP-41C had a wide range of advanced mathematical functions built-in, including trigonometric, logarithmic, and statistical functions. Large memory: The HP-41C had a large amount of memory, which allowed users to store many programs and data sets. RPN: The HP-41C used Reverse Polish Notation (RPN) entry, which made it faster and more efficient to use for complex mathematical operations. Durability: The HP-41C was known for its durability and ruggedness, which made it popular among engineers and scientists who needed a reliable calculator for field work. The HP-41C was widely adopted in the engineering, science and financial fields, it was also popular among hobbyists and enthusiasts, many of them developed programs and shared them with others, which expanded the calculators capabilities even more. Me: Are there any recent developments for the HP-41 calculators? ChatGPT: The HP-41C was discontinued in the early 1990s and there have been no recent developments for the calculator itself. However, there is still a dedicated community of users, developers and enthusiasts that continue to use and support the HP-41C. They share software, information, hardware and ideas on how to improve the capabilities of the calculator. Some of the things that have been developed for the HP-41C include: Emulators: Software emulators that allow the HP-41C to be used on modern computers and mobile devices. These emulators are often more accurate and powerful than the original calculator, and they can run programs and perform calculations that the original HP-41C could not. Hardware upgrades: Some enthusiasts have developed hardware upgrades for the HP-41C, such as memory expansions and interface modules that allow the calculator to be connected to modern computers and devices. Third-party software: There are many third-party software programs available for the HP-41C, which can be used to expand its functionality and capabilities. This includes programs for statistics, finance, engineering and other fields Support communities: There are online communities dedicated to the HP-41C, where users can share tips, ask questions, and find information on how to use and maintain the calculator. The HP-41C's design and architecture has had a lasting impact, it's design influenced later HP calculators and other calculator manufacturers, its RPN notation influenced the design of HP's financial and scientific calculators in the following decades. |
|||
01-20-2023, 05:51 PM
Post: #2
|
|||
|
|||
RE: A quick conversation with ChatGPT
Juergen,
you killed me.. If you need to speak about HP41, you are for sure welcome in the HPCC regular monthly meetings. Speaking with a Bot? haha (I appreciate the post. I was just having a big laugh). HP71B 4TH/ASM/Multimod, HP41CV/X/Y & Nov64d, PILBOX, HP-IL 821.62A & 64A & 66A, Deb11 64b-PC & PI2 3 4 w/ ILPER, VIDEO80, V41 & EMU71, DM41X |
|||
01-26-2023, 06:52 PM
Post: #3
|
|||
|
|||
RE: A quick conversation with ChatGPT | |||
01-26-2023, 07:27 PM
Post: #4
|
|||
|
|||
RE: A quick conversation with ChatGPT | |||
01-26-2023, 09:10 PM
Post: #5
|
|||
|
|||
RE: A quick conversation with ChatGPT
(01-26-2023 07:27 PM)Albert Chan Wrote:(01-26-2023 06:52 PM)John Galt Wrote: ChatGPT's command of the English language is less than impressive. If you can't tell the difference by chatting with it, then it passed the Turing test with flying colors! Tom L Cui bono? |
|||
01-27-2023, 03:24 PM
Post: #6
|
|||
|
|||
RE: A quick conversation with ChatGPT
(01-26-2023 06:52 PM)John Galt Wrote:(01-20-2023 02:50 PM)Jurgen Keller Wrote: ... it's design influenced later HP calculators and other calculator manufacturers, its RPN notation influenced the design... [/i]ChatGPT's command of the English language is less than impressive. One has to keep in mind that GPT-3.5 (the model behind chat GPT) is trained on a lot of literature but also a snapshot of the internet (see "Common Crawl"). Thus there is likely a lot of low quality input in the training database (mine included). This explains also because sometimes there are more errors than one would expect. Still the Turing test is something that is most likely passed. Wikis are great, Contribute :) |
|||
01-29-2023, 01:39 PM
Post: #7
|
|||
|
|||
RE: A quick conversation with ChatGPT
(01-20-2023 05:51 PM)floppy Wrote: Juergen, I prefer to speak to real persons, too. I was just curious about all this hype about ChatGPT. I think I've missed to renew my HPCC membership a few years ago. Do you usually have particular agendas for the meetings, or is it more kind of a free chat about (HP) calculators? |
|||
01-29-2023, 01:45 PM
Post: #8
|
|||
|
|||
RE: A quick conversation with ChatGPT
(01-27-2023 03:24 PM)pier4r Wrote:(01-26-2023 06:52 PM)John Galt Wrote: ChatGPT's command of the English language is less than impressive... Yep, I've seen other conversations that contained really bad mistakes. The problem is that people tend to believe fake news if you repeat it thousands of times. We've seen many examples in the last few years, in particular in politics. |
|||
01-29-2023, 03:06 PM
Post: #9
|
|||
|
|||
RE: A quick conversation with ChatGPT
(01-29-2023 01:39 PM)Jurgen Keller Wrote:My recommendation:(01-20-2023 05:51 PM)floppy Wrote: Juergen, - contact HPCC via their internet page - you should receive a webex invit - to the moderator: announce topics you want to present during the meeting; or you are interested in - if none, listen at the meeting when it happens (others comes with topics) I would say: an interactive discussion of a bunch of people with a wide knowledge and democratic behaviour, with the mindset to bring new things making turning the galaxy. HP71B 4TH/ASM/Multimod, HP41CV/X/Y & Nov64d, PILBOX, HP-IL 821.62A & 64A & 66A, Deb11 64b-PC & PI2 3 4 w/ ILPER, VIDEO80, V41 & EMU71, DM41X |
|||
01-31-2023, 09:04 PM
Post: #10
|
|||
|
|||
RE: A quick conversation with ChatGPT
On the subject of RPN/Forth, here is an interesting conversation (not mine):
ChatGPT and Forth Question write a prime test program. The program should take a number from the stack and return 1 on the stack if it is prime and 0 if it is not. The program should be written in Forth. ChatGPT Here is a simple prime test program in Forth: sql: prime? ( n -- f ) 2 dup >r begin dup * dup >r dup 2 < over 0= until r> drop r> swap 2 < if drop 0 exit then 1 ; Explanation:
The full conversation is on reddit: https://www.reddit.com/r/Forth/comments/...and_forth/ Not much of a threat to Forth programmers - Rob "I count on old friends to remain rational" |
|||
02-01-2023, 01:00 PM
(This post was last modified: 02-01-2023 01:01 PM by Maximilian Hohmann.)
Post: #11
|
|||
|
|||
RE: A quick conversation with ChatGPT
Hello!
(01-31-2023 09:04 PM)robve Wrote: Not much of a threat to Forth programmers These are just teething problems, it's learning fast! But honestly, I am a bit bewildered about the apparent hostility and negativity against artificial intelligence shown in this forum. When the pocket calculator was introduced in the 1970ies it was a great technological advance, and all of us who witnessed that have been fascinated ever since. Hence this website and forum... Now, after several decades of research, AI finally starts to come out from the labs to become part of our lives. This is another major technological step forward, in my estimation many oders of magnitudes greater than the pocket calculator. Why not greet it as happily as we greeted the first calculators back then? Regards Max |
|||
02-01-2023, 06:17 PM
Post: #12
|
|||
|
|||
RE: A quick conversation with ChatGPT
ChatGPT tells me to use a negative interest rate to find FV.
Why? Quote:The HP-12C calculator uses Reverse Polish Notation (RPN) to perform its calculations. RPN is a mathematical notation in which the operands (numbers) are entered before the operation. Here's an example of how to perform a simple interest calculation using the HP-12C: HP-12C Gold / HP-12C Platinum |
|||
02-01-2023, 11:22 PM
Post: #13
|
|||
|
|||
RE: A quick conversation with ChatGPT
(02-01-2023 01:00 PM)Maximilian Hohmann Wrote: But honestly, I am a bit bewildered about the apparent hostility and negativity against artificial intelligence shown in this forum. In 1982 the economist Bernard Avishai wrote "The danger from computers is not that they will eventually get as smart as men, but that we will meanwhile agree to meet them half-way." ChatGPT is a great piece of research but it's very much only half-way. The hostility, well my own at least, is not against ChatGPT itself but the way it is seemingly being adopted uncritically. |
|||
02-01-2023, 11:42 PM
(This post was last modified: 02-01-2023 11:48 PM by pier4r.)
Post: #14
|
|||
|
|||
RE: A quick conversation with ChatGPT
(02-01-2023 01:00 PM)Maximilian Hohmann Wrote: Hello! Could you see this picture? https://pbs.twimg.com/media/Fn1ALZgaAAIj...name=large (I am not sure, since it could be behind registration). In general I think the skepticism is not against the tool, but how it gets advertised. It is a bit like the self driving marketing of tesla, that eventually will be done, but not as fast and as secure as the brand was marketing. edit: further combining those AI tools (a bit like the "programs" of the geth in mass effect) one can get much better results. For example chatGPT + wolfram Alpha, wolfram can save the answer more often than not when chatGPT simply is on the wrong way. An example of a collection of tools: https://www.futurepedia.io/ . Wikis are great, Contribute :) |
|||
02-02-2023, 01:02 AM
Post: #15
|
|||
|
|||
RE: A quick conversation with ChatGPT
(02-01-2023 11:22 PM)BruceH Wrote:(02-01-2023 01:00 PM)Maximilian Hohmann Wrote: But honestly, I am a bit bewildered about the apparent hostility and negativity against artificial intelligence shown in this forum. Well said!! I agree fully with Max, I'm a bit surprised by the jollies folks seem to get denigrating imperfect results (perhaps not knowing their attempts to use are actually improving it...). But I am bewildered at folks wanting to turn it loose on real world issues, and apparently expecting it to be up to the task. In Max's metaphor, a bit like asking the HP-45 to solve demanding real-time 3D simulations and poo-pooing it for not being able to do it yet. While I guess the key word in there is yet, it's the expectation that's at fault, not the technology. --Bob Prosperi |
|||
02-02-2023, 01:09 AM
Post: #16
|
|||
|
|||
RE: A quick conversation with ChatGPT
(01-29-2023 01:39 PM)Jurgen Keller Wrote: I prefer to speak to real persons, too. I was just curious about all this hype about ChatGPT. Hi Jurgen - Send an email or PM to Mark Power (username here is, unsurprisingly, 'Mark Power') and request an invitation to the meetings; we typically meet on the 1st and 3rd Saturday's of each month. Meetings are usually unstructured, but only because we would rather meet even if there is nothing structured to present; there is always something calculator geeks find to discuss, including show-and-tell about the latest toys (both calculators and not), troubleshooting discussions about ongoing repair projects and other related topics. Drop-in when it's convenient. --Bob Prosperi |
|||
02-02-2023, 04:06 AM
(This post was last modified: 02-02-2023 03:25 PM by robve.)
Post: #17
|
|||
|
|||
RE: A quick conversation with ChatGPT
(02-01-2023 01:00 PM)Maximilian Hohmann Wrote: Hello! The big difference with the 50s up to the new millennium is that almost all new technologies created during those decades were developed by scientists and engineers to empower humans with advances that filled a need rather than a want. The technological creations made humans smarter, healthier, more creative more productive, and more scientifically or technically informed by exploring new discoveries. Marketing took a much smaller role back then, than it does today. Today's technologies appear to be for a large part created to fill a marketing opportunity, not to advance humanity, but to entertain or accelerate productivity even without humans. Most of today's AI gets incrementally better, never reaching a level that is promised. AI can't effectively reason with counterfactuals for example, which is critical to compete with humans. To understand the importance of reasoning, read The Book of Why by ACM Turing award winner J. Pearl. Perhaps today's AI makes us more productive, for example to assist with answers like chatGPT does, which of course was always a major aim of corporations to have us produce (and consume) more stuff. Poking fun of AI is not a form of negativity IMO. Back in the 80s we laughed about Elisa (an AI therapist). Accepting a technology without criticism, because we're looking for something new to be excited about, feels utterly empty and unrewarding, especially when over time it becomes clear that the "new thing" under delivers. The investors got their ROI and will just close up shop. Who cares? It will all be forgotten in a year or two. We have seen this over and over again. Most startups talk about an exit strategy in their business plan. In other words, there is no long term viability plan. It scares them. Since most of us care more about our children and their future than about ourselves, ask yourself this question: would you fully trust an AI (e.g. chatGPT) to offer medical, career, and financial advice to your children? Or would you like your children to learn by gaining experience and confidence through interactions with humans in our society? What if AI has it 95% right, but the remaining 5% will kill someone or make him/her bankrupt or very unhappy with the choices made? Who is responsible for this? Sure, humans are also flawed, but there are a lot of us with difference experiences and experiences that are closer to resonate with others who seek advice. If someone gives bad advice, we move on to interact with others and forget about the ordeal. That will be hard to do when there is only one (or maybe two?) AI that offers occasional bad advice. - Rob "I count on old friends to remain rational" |
|||
« Next Oldest | Next Newest »
|
User(s) browsing this thread: