* [NTG-context] knuth @ 2023-06-01 9:44 Hans Hagen via ntg-context 2023-06-01 10:32 ` luigi scarso via ntg-context ` (2 more replies) 0 siblings, 3 replies; 12+ messages in thread From: Hans Hagen via ntg-context @ 2023-06-01 9:44 UTC (permalink / raw) To: mailing list for ConTeXt users; +Cc: Hans Hagen Hi, Probably some on this list already checked how well chatgpt answers questions about domains one knows well and then probably noted that in spite of impressive wording, one can run into quite incorrect answers. One can get really stupid responses about tex and friends, but also can get impressive exmapled when asked for. (I'm still planning a wrap up of some.) That said, one should read: https://cs.stanford.edu/~knuth/chatGPT20.txt and this makes a pretty nice new sample file: It's amazing how the confident tone lends credibility to all of that made-up nonsense. Almost impossible for anybody without knowledge of the book to believe that those "facts" aren't authorititative and well researched. (among some other remarks) Hans ----------------------------------------------------------------- Hans Hagen | PRAGMA ADE Ridderstraat 27 | 8061 GH Hasselt | The Netherlands tel: 038 477 53 69 | www.pragma-ade.nl | www.pragma-pod.nl ----------------------------------------------------------------- ___________________________________________________________________________________ If your question is of interest to others as well, please add an entry to the Wiki! maillist : ntg-context@ntg.nl / https://www.ntg.nl/mailman/listinfo/ntg-context webpage : https://www.pragma-ade.nl / http://context.aanhet.net archive : https://bitbucket.org/phg/context-mirror/commits/ wiki : https://contextgarden.net ___________________________________________________________________________________ ^ permalink raw reply [flat|nested] 12+ messages in thread
* Re: [NTG-context] knuth 2023-06-01 9:44 [NTG-context] knuth Hans Hagen via ntg-context @ 2023-06-01 10:32 ` luigi scarso via ntg-context 2023-06-01 11:49 ` Floris van Manen via ntg-context 2023-06-02 7:27 ` BPJ via ntg-context 2023-06-04 21:57 ` Berend de Boer via ntg-context 2 siblings, 1 reply; 12+ messages in thread From: luigi scarso via ntg-context @ 2023-06-01 10:32 UTC (permalink / raw) To: mailing list for ConTeXt users; +Cc: luigi scarso [-- Attachment #1.1: Type: text/plain, Size: 1401 bytes --] On Thu, 1 Jun 2023 at 11:45, Hans Hagen via ntg-context <ntg-context@ntg.nl> wrote: > Hi, > > Probably some on this list already checked how well chatgpt answers > questions about domains one knows well and then probably noted that in > spite of impressive wording, one can run into quite incorrect answers. > > One can get really stupid responses about tex and friends, but also can > get impressive exmapled when asked for. (I'm still planning a wrap up of > some.) > > That said, one should read: > > https://cs.stanford.edu/~knuth/chatGPT20.txt > > and this makes a pretty nice new sample file: > > It's amazing how the confident tone lends credibility to all of that > made-up nonsense. Almost impossible for anybody without knowledge > of the book to believe that those "facts" aren't authorititative > and well researched. > you can happily include the email that I have sent to you that declares that *you* are the author of MFlua and *I* have nothing to do with it (chatgpt3: as far as I know, the overall number of users+developers of MFLua in the entire solar system is exactly 1 -- me -- at least from 10 years) . For a moment I was even convinced that it was true. On the other hand, I am pretty sure that it will be able to generate perfect lualatex files for articles, reports, and later books. How and when it's more a matter of interest rather than capabilities. -- luigi [-- Attachment #1.2: Type: text/html, Size: 2038 bytes --] [-- Attachment #2: Type: text/plain, Size: 496 bytes --] ___________________________________________________________________________________ If your question is of interest to others as well, please add an entry to the Wiki! maillist : ntg-context@ntg.nl / https://www.ntg.nl/mailman/listinfo/ntg-context webpage : https://www.pragma-ade.nl / http://context.aanhet.net archive : https://bitbucket.org/phg/context-mirror/commits/ wiki : https://contextgarden.net ___________________________________________________________________________________ ^ permalink raw reply [flat|nested] 12+ messages in thread
* Re: [NTG-context] knuth 2023-06-01 10:32 ` luigi scarso via ntg-context @ 2023-06-01 11:49 ` Floris van Manen via ntg-context 0 siblings, 0 replies; 12+ messages in thread From: Floris van Manen via ntg-context @ 2023-06-01 11:49 UTC (permalink / raw) To: ntg-context; +Cc: Floris van Manen On 01/06/2023 12:32, luigi scarso via ntg-context wrote: > It's amazing how the confident tone lends credibility to all of that > made-up nonsense. Almost impossible for anybody without knowledge > of the book to believe that those "facts" aren't authorititative > and well researched. Step by step. First you tell them that your system is more convenient then reading the books yourself. Then you replace common sense with new daily 'facts'. Nothing has really changed over the years; if you point out errors, the mere response is that it is still work in progress (see how great opportunities this has), for now it is good enough to entertain children. As if children do not deserve the best access to true knowledge and curiosity. Divide & conquer (wie is er niet groot mee geworden ;-) .F ___________________________________________________________________________________ If your question is of interest to others as well, please add an entry to the Wiki! maillist : ntg-context@ntg.nl / https://www.ntg.nl/mailman/listinfo/ntg-context webpage : https://www.pragma-ade.nl / http://context.aanhet.net archive : https://bitbucket.org/phg/context-mirror/commits/ wiki : https://contextgarden.net ___________________________________________________________________________________ ^ permalink raw reply [flat|nested] 12+ messages in thread
* Re: [NTG-context] knuth 2023-06-01 9:44 [NTG-context] knuth Hans Hagen via ntg-context 2023-06-01 10:32 ` luigi scarso via ntg-context @ 2023-06-02 7:27 ` BPJ via ntg-context 2023-06-02 7:58 ` Hans Hagen via ntg-context 2023-06-04 21:57 ` Berend de Boer via ntg-context 2 siblings, 1 reply; 12+ messages in thread From: BPJ via ntg-context @ 2023-06-02 7:27 UTC (permalink / raw) To: mailing list for ConTeXt users; +Cc: BPJ [-- Attachment #1.1: Type: text/plain, Size: 1982 bytes --] Den tors 1 juni 2023 11:48Hans Hagen via ntg-context <ntg-context@ntg.nl> skrev: > Hi, > > Probably some on this list already checked how well chatgpt answers > questions about domains one knows well and then probably noted that in > spite of impressive wording, one can run into quite incorrect answers. > > One can get really stupid responses about tex and friends, but also can > get impressive exmapled when asked for. (I'm still planning a wrap up of > some.) > > That said, one should read: > > https://cs.stanford.edu/~knuth/chatGPT20.txt > > and this makes a pretty nice new sample file: > > It's amazing how the confident tone lends credibility to all of that > made-up nonsense. Almost impossible for anybody without knowledge > of the book to believe that those "facts" aren't authorititative > and well researched. > > (among some other remarks) > I wonder what Knuth would say about having ChatGPT write computer programs, which I think can be outright dangerous. Either you get shitty code which doesn't work, which is good, or you get shitty code which works which is bad or Really Bad. > Hans > > ----------------------------------------------------------------- > Hans Hagen | PRAGMA ADE > Ridderstraat 27 | 8061 GH Hasselt | The Netherlands > tel: 038 477 53 69 | www.pragma-ade.nl | www.pragma-pod.nl > ----------------------------------------------------------------- > > ___________________________________________________________________________________ > If your question is of interest to others as well, please add an entry to > the Wiki! > > maillist : ntg-context@ntg.nl / > https://www.ntg.nl/mailman/listinfo/ntg-context > webpage : https://www.pragma-ade.nl / http://context.aanhet.net > archive : https://bitbucket.org/phg/context-mirror/commits/ > wiki : https://contextgarden.net > > ___________________________________________________________________________________ > [-- Attachment #1.2: Type: text/html, Size: 3490 bytes --] [-- Attachment #2: Type: text/plain, Size: 496 bytes --] ___________________________________________________________________________________ If your question is of interest to others as well, please add an entry to the Wiki! maillist : ntg-context@ntg.nl / https://www.ntg.nl/mailman/listinfo/ntg-context webpage : https://www.pragma-ade.nl / http://context.aanhet.net archive : https://bitbucket.org/phg/context-mirror/commits/ wiki : https://contextgarden.net ___________________________________________________________________________________ ^ permalink raw reply [flat|nested] 12+ messages in thread
* Re: [NTG-context] knuth 2023-06-02 7:27 ` BPJ via ntg-context @ 2023-06-02 7:58 ` Hans Hagen via ntg-context 2023-06-02 21:07 ` Carlos via ntg-context 0 siblings, 1 reply; 12+ messages in thread From: Hans Hagen via ntg-context @ 2023-06-02 7:58 UTC (permalink / raw) To: mailing list for ConTeXt users; +Cc: Hans Hagen On 6/2/2023 9:27 AM, BPJ wrote: > I wonder what Knuth would say about having ChatGPT write computer > programs, which I think can be outright dangerous. Either you get shitty > code which doesn't work, which is good, or you get shitty code which > works which is bad or Really Bad. He's pretty clear that he leave that to others and continues with TAOCP. wrt you remark: to a large extend these are 'plagiarism' machines so one can wonder about more. I suppose it also depends on how original these programs are (or become). I don't see myself using such software, just like I don't use all these software coding tools (ide stuff). When I acn't remember what I'm doing, or need constant help popping up I should not code. I'm intrigued of course, and see valid applications, but wrt programming I'm just not that interested. I bet after a few years people get bored with the artificial stuff (painting, music, proze, movies, whatever; just hitting buttons doesn't make a creative person I guess), unless of course one wants to be zombified. Hans ----------------------------------------------------------------- Hans Hagen | PRAGMA ADE Ridderstraat 27 | 8061 GH Hasselt | The Netherlands tel: 038 477 53 69 | www.pragma-ade.nl | www.pragma-pod.nl ----------------------------------------------------------------- ___________________________________________________________________________________ If your question is of interest to others as well, please add an entry to the Wiki! maillist : ntg-context@ntg.nl / https://www.ntg.nl/mailman/listinfo/ntg-context webpage : https://www.pragma-ade.nl / http://context.aanhet.net archive : https://bitbucket.org/phg/context-mirror/commits/ wiki : https://contextgarden.net ___________________________________________________________________________________ ^ permalink raw reply [flat|nested] 12+ messages in thread
* Re: [NTG-context] knuth 2023-06-02 7:58 ` Hans Hagen via ntg-context @ 2023-06-02 21:07 ` Carlos via ntg-context 0 siblings, 0 replies; 12+ messages in thread From: Carlos via ntg-context @ 2023-06-02 21:07 UTC (permalink / raw) To: ntg-context; +Cc: Carlos On Fri, Jun 02, 2023 at 09:58:31AM +0200, Hans Hagen via ntg-context wrote: > On 6/2/2023 9:27 AM, BPJ wrote: > > > I wonder what Knuth would say about having ChatGPT write computer > > programs, which I think can be outright dangerous. Either you get shitty > > code which doesn't work, which is good, or you get shitty code which > > works which is bad or Really Bad. > > He's pretty clear that he leave that to others and continues with TAOCP. > > wrt you remark: to a large extend these are 'plagiarism' machines so one can > wonder about more. I suppose it also depends on how original these programs > are (or become). Not original at all. One New Years' resolution ought to be «stop feeding this whatever created by whoever» I've been trying to find a recent (just done a few months ago) interactive session I had had with ChatGPT but I'm afraid I must have dd'd the entire drive (just an honest mistake on the wrong drive at the time while aiming primarily at fixing another drive and there it went :( But every interaction with it [chatgpt], was met/replied to with a digression and paraphrasing on its part. i.e, sorry 'bout that. I will not make that mistake again. I knew right away it was just spitting out useless information. Grammatically correct but not factual. Lacking substance. I first brought up the invention of the printing press as a starting point. And I asked it, to write me up an essay with the most interesting parts I remember that the point that chatgpt was trying to make, was in that the printing press challenged the authority of the church. And I remember I had to go back and forth with it and question that very verb/use of the word challenging. If anything seemed clear to me years ago. and also now, by reading about the printing press and with the publication of the Bible and later with the Book of Psalms and the first colophon in human hitory and so forth, is that the invention of the printing press fostered the establishment of the church. But I can't recall/just don't know right now if by the time chatgpt had decided to use ‹challenging› was implying afterwards or after the establishment of the church, and the excommunications that followed and so forth. All I know is that it rewrote it with different paragraphs after I pointed it out. Very well put nevertheless. But its main argumentative conclusion was the challenging part rather than cementing/positioning/establishing/ one Then I thought, heck, let's see if chatgtt is aware of the relatively recent findings about the printing methods used by Gutenberg and so forth. But it wasn't very clear on this either. As a matter of fact it had no clue really about this for example, citing : https://www.nytimes.com/2001/01/27/arts/has-history-been-too-generous-to-gutenberg.html «Johann Gutenberg, the 15th-century German craftsman, has long been believed to be the father of modern typography. But the secretive inventor may have to share some of the paternity now. A physicist and a scholar of rare books at Princeton University who jointly used new technology to examine some of Gutenberg's texts say he may not have created the seminal process after all, a finding could rewrite the history of printing. The two scholars contend that the metal mold method of printing attributed to Gutenberg was probably invented by someone else about 20 years after Gutenberg printed his Bible. The method, which involves punching a letter into a copper matrix that is filled with lead alloy to create hundreds of identical letters, was the principal way of printing until after World War II.» «The discovery was announced on Monday by Paul Needham, the librarian of the Scheide Library, a private library housed at Princeton, and Blaise Agüera y Arcas, a 25-year-old graduate of Princeton with a degree in physics, before a standing-room-only audience at New York's Grolier Club, a club for book collectors founded in 1884.» the above link was referenced at the following link originally, which I had read before https://jikji.utah.edu/ > I don't see myself using such software, just like I don't > use all these software coding tools (ide stuff). When I acn't remember what > I'm doing, or need constant help popping up I should not code. I'm intrigued > of course, and see valid applications, but wrt programming I'm just not that > interested. I bet after a few years people get bored with the artificial > stuff (painting, music, proze, movies, whatever; just hitting buttons > doesn't make a creative person I guess), unless of course one wants to be > zombified. > > Hans > > ----------------------------------------------------------------- > Hans Hagen | PRAGMA ADE > Ridderstraat 27 | 8061 GH Hasselt | The Netherlands > tel: 038 477 53 69 | www.pragma-ade.nl | www.pragma-pod.nl > ----------------------------------------------------------------- > > ___________________________________________________________________________________ > If your question is of interest to others as well, please add an entry to the Wiki! > > maillist : ntg-context@ntg.nl / https://www.ntg.nl/mailman/listinfo/ntg-context > webpage : https://www.pragma-ade.nl / http://context.aanhet.net > archive : https://bitbucket.org/phg/context-mirror/commits/ > wiki : https://contextgarden.net > ___________________________________________________________________________________ > -- Function reject. ___________________________________________________________________________________ If your question is of interest to others as well, please add an entry to the Wiki! maillist : ntg-context@ntg.nl / https://www.ntg.nl/mailman/listinfo/ntg-context webpage : https://www.pragma-ade.nl / http://context.aanhet.net archive : https://bitbucket.org/phg/context-mirror/commits/ wiki : https://contextgarden.net ___________________________________________________________________________________ ^ permalink raw reply [flat|nested] 12+ messages in thread
* Re: [NTG-context] knuth 2023-06-01 9:44 [NTG-context] knuth Hans Hagen via ntg-context 2023-06-01 10:32 ` luigi scarso via ntg-context 2023-06-02 7:27 ` BPJ via ntg-context @ 2023-06-04 21:57 ` Berend de Boer via ntg-context 2023-06-05 13:09 ` Alan Braslau via ntg-context 2 siblings, 1 reply; 12+ messages in thread From: Berend de Boer via ntg-context @ 2023-06-04 21:57 UTC (permalink / raw) To: mailing list for ConTeXt users; +Cc: Berend de Boer Hi Hans, > Probably some on this list already checked how well chatgpt answers > questions about domains one knows well and then probably noted that in > spite of impressive wording, one can run into quite incorrect answers. I've been using it for ConTeXt, and to be honest, it works amazingly well. It hallucinates sometimes, but it's such a time saver. -- All the best, Berend de Boer ___________________________________________________________________________________ If your question is of interest to others as well, please add an entry to the Wiki! maillist : ntg-context@ntg.nl / https://www.ntg.nl/mailman/listinfo/ntg-context webpage : https://www.pragma-ade.nl / http://context.aanhet.net archive : https://bitbucket.org/phg/context-mirror/commits/ wiki : https://contextgarden.net ___________________________________________________________________________________ ^ permalink raw reply [flat|nested] 12+ messages in thread
* Re: [NTG-context] knuth 2023-06-04 21:57 ` Berend de Boer via ntg-context @ 2023-06-05 13:09 ` Alan Braslau via ntg-context 2023-06-05 14:20 ` Carlos via ntg-context 2023-06-05 14:50 ` luigi scarso via ntg-context 0 siblings, 2 replies; 12+ messages in thread From: Alan Braslau via ntg-context @ 2023-06-05 13:09 UTC (permalink / raw) To: mailing list for ConTeXt users; +Cc: Alan Braslau On 04/06/23 04/06/23, 15:57, Berend de Boer via ntg-context wrote: >> Probably some on this list already checked how well chatgpt answers >> questions about domains one knows well and then probably noted that in >> spite of impressive wording, one can run into quite incorrect answers. > > I've been using it for ConTeXt, and to be honest, it works amazingly > well. It hallucinates sometimes, but it's such a time saver. Maybe someone should try asking it to write documentation/manuals? As a teacher, we are accustomed to seeing Google/Wikipedia/etc. cut-and-paste nonsense; now, this nonsense appears more polished, and our students are none the wiser. Saves them time, too! ;-) Who said: "the best way to economize thought is not to think at all"? Alan ___________________________________________________________________________________ If your question is of interest to others as well, please add an entry to the Wiki! maillist : ntg-context@ntg.nl / https://www.ntg.nl/mailman/listinfo/ntg-context webpage : https://www.pragma-ade.nl / http://context.aanhet.net archive : https://bitbucket.org/phg/context-mirror/commits/ wiki : https://contextgarden.net ___________________________________________________________________________________ ^ permalink raw reply [flat|nested] 12+ messages in thread
* Re: [NTG-context] knuth 2023-06-05 13:09 ` Alan Braslau via ntg-context @ 2023-06-05 14:20 ` Carlos via ntg-context 2023-06-05 14:50 ` luigi scarso via ntg-context 1 sibling, 0 replies; 12+ messages in thread From: Carlos via ntg-context @ 2023-06-05 14:20 UTC (permalink / raw) To: ntg-context; +Cc: Carlos On Mon, Jun 05, 2023 at 03:09:02PM +0200, Alan Braslau via ntg-context wrote: > > On 04/06/23 04/06/23, 15:57, Berend de Boer via ntg-context wrote: > > > Probably some on this list already checked how well chatgpt answers > > > questions about domains one knows well and then probably noted that in > > > spite of impressive wording, one can run into quite incorrect answers. > > > > I've been using it for ConTeXt, and to be honest, it works amazingly > > well. It hallucinates sometimes, but it's such a time saver. > > Maybe someone should try asking it to write documentation/manuals? > > As a teacher, we are accustomed to seeing Google/Wikipedia/etc. > cut-and-paste nonsense; now, this nonsense appears more polished, and our > students are none the wiser. Saves them time, too! ;-) > > Who said: "the best way to economize thought is not to think at all"? > > Alan Yeah. :) that's funny. Everybody uses chatgpt with lmtx that's why everyone that uses it ends up in the mailing list asking to resolve a problem that chatgpt couldn't resolve on its own. But here is the interesting part. The questions are answered by using chatgpt too, rather than humans so nothing gets worked out in the end > ___________________________________________________________________________________ > If your question is of interest to others as well, please add an entry to the Wiki! > > maillist : ntg-context@ntg.nl / https://www.ntg.nl/mailman/listinfo/ntg-context > webpage : https://www.pragma-ade.nl / http://context.aanhet.net > archive : https://bitbucket.org/phg/context-mirror/commits/ > wiki : https://contextgarden.net > ___________________________________________________________________________________ > -- If it has syntax, it isn't user friendly. ___________________________________________________________________________________ If your question is of interest to others as well, please add an entry to the Wiki! maillist : ntg-context@ntg.nl / https://www.ntg.nl/mailman/listinfo/ntg-context webpage : https://www.pragma-ade.nl / http://context.aanhet.net archive : https://bitbucket.org/phg/context-mirror/commits/ wiki : https://contextgarden.net ___________________________________________________________________________________ ^ permalink raw reply [flat|nested] 12+ messages in thread
* Re: [NTG-context] knuth 2023-06-05 13:09 ` Alan Braslau via ntg-context 2023-06-05 14:20 ` Carlos via ntg-context @ 2023-06-05 14:50 ` luigi scarso via ntg-context 2023-06-05 15:22 ` Alan Braslau via ntg-context 2023-06-05 16:07 ` Carlos via ntg-context 1 sibling, 2 replies; 12+ messages in thread From: luigi scarso via ntg-context @ 2023-06-05 14:50 UTC (permalink / raw) To: mailing list for ConTeXt users; +Cc: luigi scarso [-- Attachment #1.1: Type: text/plain, Size: 1306 bytes --] On Mon, 5 Jun 2023 at 15:18, Alan Braslau via ntg-context < ntg-context@ntg.nl> wrote: > > On 04/06/23 04/06/23, 15:57, Berend de Boer via ntg-context wrote: > >> Probably some on this list already checked how well chatgpt answers > >> questions about domains one knows well and then probably noted that in > >> spite of impressive wording, one can run into quite incorrect answers. > > > > I've been using it for ConTeXt, and to be honest, it works amazingly > > well. It hallucinates sometimes, but it's such a time saver. > > Maybe someone should try asking it to write documentation/manuals? > > As a teacher, we are accustomed to seeing Google/Wikipedia/etc. > cut-and-paste nonsense; now, this nonsense appears more polished, and > our students are none the wiser. Saves them time, too! ;-) > > Who said: "the best way to economize thought is not to think at all"? > not all is bad: sometimes a textual description can replace a formalized notation, e.g. https://fosstodon.org/@tao@mathstodon.xyz/110250604086213386 I can imagine similar examples with tables. Of course a very high level text is often ambiguous (as some kind of formalized grammars, after all) but the example shows that in these cases it is better to fix something already almost ok than to typeset it from scratch. -- luigi [-- Attachment #1.2: Type: text/html, Size: 1919 bytes --] [-- Attachment #2: Type: text/plain, Size: 496 bytes --] ___________________________________________________________________________________ If your question is of interest to others as well, please add an entry to the Wiki! maillist : ntg-context@ntg.nl / https://www.ntg.nl/mailman/listinfo/ntg-context webpage : https://www.pragma-ade.nl / http://context.aanhet.net archive : https://bitbucket.org/phg/context-mirror/commits/ wiki : https://contextgarden.net ___________________________________________________________________________________ ^ permalink raw reply [flat|nested] 12+ messages in thread
* Re: [NTG-context] knuth 2023-06-05 14:50 ` luigi scarso via ntg-context @ 2023-06-05 15:22 ` Alan Braslau via ntg-context 2023-06-05 16:07 ` Carlos via ntg-context 1 sibling, 0 replies; 12+ messages in thread From: Alan Braslau via ntg-context @ 2023-06-05 15:22 UTC (permalink / raw) To: luigi scarso, mailing list for ConTeXt users; +Cc: Alan Braslau On 05/06/23 05/06/23, 08:50, luigi scarso wrote: > > > On Mon, 5 Jun 2023 at 15:18, Alan Braslau via ntg-context > <ntg-context@ntg.nl <mailto:ntg-context@ntg.nl>> wrote: > > > On 04/06/23 04/06/23, 15:57, Berend de Boer via ntg-context wrote: > >> Probably some on this list already checked how well chatgpt answers > >> questions about domains one knows well and then probably noted > that in > >> spite of impressive wording, one can run into quite incorrect > answers. > > > > I've been using it for ConTeXt, and to be honest, it works amazingly > > well. It hallucinates sometimes, but it's such a time saver. > > Maybe someone should try asking it to write documentation/manuals? > > As a teacher, we are accustomed to seeing Google/Wikipedia/etc. > cut-and-paste nonsense; now, this nonsense appears more polished, and > our students are none the wiser. Saves them time, too! ;-) > > Who said: "the best way to economize thought is not to think at all"? > > > > not all is bad: sometimes a textual description can replace a formalized > notation, e.g. > https://fosstodon.org/@tao@mathstodon.xyz/110250604086213386 > <https://fosstodon.org/@tao@mathstodon.xyz/110250604086213386> > I can imagine similar examples with tables. > Of course a very high level text is often ambiguous (as some kind of > formalized grammars, after all) > but the example shows that in these cases it is better to fix > something already almost ok than to typeset it from scratch. As an Artificial Intelligence machine, I cannot speculate why it is so hard to write correct LaTeX that does not need to be tweaked in order to give an aesthetic result. However, from monitoring the ConTeXt mailing list, I have observed that most problems seem to have an answer that work correctly when used correctly. In the rare case that ConTeXt users report something not working correctly, this usually is revealed to be a small typographic error in the macros appearing following a major internal code rewrite. I have recorded in all of these rare cases a pattern that reoccurs repeatedly with the phrase "fixed in next upload". I cannot speculate as to how such fixes appear so rapidly and so consistently. My analytical artificial intelligence is thrown off by seemingly irrelevant statements about listening to this or that CD. What could this possibly mean? In supplement, I have not yet deciphered what is meant by the acronym MWE that also repeatedly occurs. This does not appear to fit with something that is reported as being broken. (Written with the assistance of ChatGPT.) Alan ___________________________________________________________________________________ If your question is of interest to others as well, please add an entry to the Wiki! maillist : ntg-context@ntg.nl / https://www.ntg.nl/mailman/listinfo/ntg-context webpage : https://www.pragma-ade.nl / http://context.aanhet.net archive : https://bitbucket.org/phg/context-mirror/commits/ wiki : https://contextgarden.net ___________________________________________________________________________________ ^ permalink raw reply [flat|nested] 12+ messages in thread
* Re: [NTG-context] knuth 2023-06-05 14:50 ` luigi scarso via ntg-context 2023-06-05 15:22 ` Alan Braslau via ntg-context @ 2023-06-05 16:07 ` Carlos via ntg-context 1 sibling, 0 replies; 12+ messages in thread From: Carlos via ntg-context @ 2023-06-05 16:07 UTC (permalink / raw) To: ntg-context; +Cc: Carlos On Mon, Jun 05, 2023 at 04:50:53PM +0200, luigi scarso via ntg-context wrote: > On Mon, 5 Jun 2023 at 15:18, Alan Braslau via ntg-context < > ntg-context@ntg.nl> wrote: > > > > > On 04/06/23 04/06/23, 15:57, Berend de Boer via ntg-context wrote: > > >> Probably some on this list already checked how well chatgpt answers > > >> questions about domains one knows well and then probably noted that in > > >> spite of impressive wording, one can run into quite incorrect answers. > > > > > > I've been using it for ConTeXt, and to be honest, it works amazingly > > > well. It hallucinates sometimes, but it's such a time saver. > > > > Maybe someone should try asking it to write documentation/manuals? > > > > As a teacher, we are accustomed to seeing Google/Wikipedia/etc. > > cut-and-paste nonsense; now, this nonsense appears more polished, and > > our students are none the wiser. Saves them time, too! ;-) > > > > Who said: "the best way to economize thought is not to think at all"? > > > > > not all is bad: sometimes a textual description can replace a formalized > notation, e.g. > https://fosstodon.org/@tao@mathstodon.xyz/110250604086213386 > I can imagine similar examples with tables. > Of course a very high level text is often ambiguous (as some kind of > formalized grammars, after all) > but the example shows that in these cases it is better to fix something > already almost ok than to typeset it from scratch. > > -- > luigi Not all is bad, no, except the company behind chatgpt ought to pay its users for feeding more data into it. It's a westernized company after all. Lots of baggage. I'm surprised it hasn't applied for a patent yet really. > ___________________________________________________________________________________ > If your question is of interest to others as well, please add an entry to the Wiki! > > maillist : ntg-context@ntg.nl / https://www.ntg.nl/mailman/listinfo/ntg-context > webpage : https://www.pragma-ade.nl / http://context.aanhet.net > archive : https://bitbucket.org/phg/context-mirror/commits/ > wiki : https://contextgarden.net > ___________________________________________________________________________________ -- You know you've been spending too much time on the computer when your friend misdates a check, and you suggest adding a "++" to fix it. ___________________________________________________________________________________ If your question is of interest to others as well, please add an entry to the Wiki! maillist : ntg-context@ntg.nl / https://www.ntg.nl/mailman/listinfo/ntg-context webpage : https://www.pragma-ade.nl / http://context.aanhet.net archive : https://bitbucket.org/phg/context-mirror/commits/ wiki : https://contextgarden.net ___________________________________________________________________________________ ^ permalink raw reply [flat|nested] 12+ messages in thread
end of thread, other threads:[~2023-06-05 16:08 UTC | newest] Thread overview: 12+ messages (download: mbox.gz / follow: Atom feed) -- links below jump to the message on this page -- 2023-06-01 9:44 [NTG-context] knuth Hans Hagen via ntg-context 2023-06-01 10:32 ` luigi scarso via ntg-context 2023-06-01 11:49 ` Floris van Manen via ntg-context 2023-06-02 7:27 ` BPJ via ntg-context 2023-06-02 7:58 ` Hans Hagen via ntg-context 2023-06-02 21:07 ` Carlos via ntg-context 2023-06-04 21:57 ` Berend de Boer via ntg-context 2023-06-05 13:09 ` Alan Braslau via ntg-context 2023-06-05 14:20 ` Carlos via ntg-context 2023-06-05 14:50 ` luigi scarso via ntg-context 2023-06-05 15:22 ` Alan Braslau via ntg-context 2023-06-05 16:07 ` Carlos via ntg-context
This is a public inbox, see mirroring instructions for how to clone and mirror all data and code used for this inbox; as well as URLs for NNTP newsgroup(s).