OpenChain and ChatGPT – New Case Studies
The OpenChain Project is releasing the first draft case studies created by ChatGPT on our GitHub. These are not intended to replace our community contributions, but to make it fast for people to add ideas and adjustments. This will specifically address one of the greatest challenges in creating new material: the initial time spent for drafting.
Why?
Our community feedback shows that people usually enjoy commenting and polishing more than drafting. Check them out and let us know what you think!
It took ChatGPT less than ten minutes to create eight case studies:
https://github.com/OpenChain-Project/Reference-Material/tree/master/Adoption-Case-Studies/Official/en/ChatGPT
Hi,
It seems that the below link shows 404-page not found. ☹
BR,
Haksung
보낸
사람:
main@... <main@...>이(가)
다음
사람
대신
보냄: Shane Coughlan <scoughlan@...>
날짜:
화요일, 2023년
5월 16일
오전 6:21
받는
사람:
OpenChain Main <main@...>
참조:
OpenChain Education <education@...>
주제:
[openchain] OpenChain and ChatGPT – New Case Studies
This got a soft launch at the OpenChain Mini-Summit adjacent to Open Source Summit North America.
The OpenChain Project is releasing the first draft case studies created by ChatGPT on our GitHub. These are not intended to replace our community contributions, but to make it fast for people to add ideas and adjustments. This will specifically address one
of the greatest challenges in creating new material: the initial time spent for drafting.
Why?
Our community feedback shows that people usually enjoy commenting and polishing more than drafting. Check them out and let us know what you think!
It took ChatGPT less than ten minutes to create eight case studies:
https://github.com/OpenChain-Project/Reference-Material/tree/master/Adoption-Case-Studies/Official/en/ChatGPT
https://github.com/OpenChain-Project/Reference-Material/tree/master/Adoption-Preparation/Case-Studies/ChatGPT might be a correct URL.
KOIZUMI Satoru @ OLYMPUS
Sent: Tuesday, May 16, 2023 9:48 AM
To: main@...
Cc: OpenChain Education <education@...>
Subject: [EXTERNAL] Re: [openchain] OpenChain and ChatGPT – New Case Studies
|
haksung=sk.com@... からのメールを受け取る頻度は高くありません。このことが重要である理由 |
|
PLEASE NOTE: Please use caution when opening links and attachments in this message. |
Hi,
It seems that the below link shows 404-page not found. ☹
BR,
Haksung
보낸
사람:
main@... <main@...>이(가)
다음
사람
대신
보냄: Shane Coughlan <scoughlan@...>
날짜:
화요일, 2023년
5월 16일
오전 6:21
받는
사람:
OpenChain Main <main@...>
참조:
OpenChain Education <education@...>
주제:
[openchain] OpenChain and ChatGPT – New Case Studies
This got a soft launch at the OpenChain Mini-Summit adjacent to Open Source Summit North America.
The OpenChain Project is releasing the first draft case studies created by ChatGPT on our GitHub. These are not intended to replace our community contributions, but to make it fast for people to add ideas and adjustments. This will specifically address one
of the greatest challenges in creating new material: the initial time spent for drafting.
Why?
Our community feedback shows that people usually enjoy commenting and polishing more than drafting. Check them out and let us know what you think!
It took ChatGPT less than ten minutes to create eight case studies:
https://github.com/OpenChain-Project/Reference-Material/tree/master/Adoption-Case-Studies/Official/en/ChatGPT
This got a soft launch at the OpenChain Mini-Summit adjacent to Open Source Summit North America.
The OpenChain Project is releasing the first draft case studies created by ChatGPT on our GitHub. These are not intended to replace our community contributions, but to make it fast for people to add ideas and adjustments. This will specifically address one of the greatest challenges in creating new material: the initial time spent for drafting.
Why?
Our community feedback shows that people usually enjoy commenting and polishing more than drafting. Check them out and let us know what you think!
It took ChatGPT less than ten minutes to create eight case studies:
https://github.com/OpenChain-Project/Reference-Material/tree/master/Adoption-Case-Studies/Official/en/ChatGPT
Koizumi San sent the correct link:
https://github.com/OpenChain-Project/Reference-Material/tree/master/Adoption-Preparation/Case-Studies/ChatGPT
My apologies for accidentally copy/pasting a link from an older revision of GitHub.
On May 16, 2023, at 9:59, Satoru Koizumi <satoru.koizumi@...> wrote:
https://github.com/OpenChain-Project/Reference-Material/tree/master/Adoption-Preparation/Case-Studies/ChatGPT might be a correct URL.
KOIZUMI Satoru @ OLYMPUS
From: main@... <main@...> On Behalf Of Haksung via lists.openchainproject.org
Sent: Tuesday, May 16, 2023 9:48 AM
To: main@...
Cc: OpenChain Education <education@...>
Subject: [EXTERNAL] Re: [openchain] OpenChain and ChatGPT – New Case Studies
haksung=sk.com@... からのメールを受け取る頻度は高くありません。このことが重要である理由PLEASE NOTE: Please use caution when opening links and attachments in this message.
Hi,
It seems that the below link shows 404-page not found. ☹
https://github.com/OpenChain-Project/Reference-Material/tree/master/Adoption-Case-Studies/Official/en/ChatGPT
BR,
Haksung
보낸 사람: main@... <main@...>이(가) 다음 사람 대신 보냄: Shane Coughlan <scoughlan@...>
날짜: 화요일, 2023년 5월 16일 오전 6:21
받는 사람: OpenChain Main <main@...>
참조: OpenChain Education <education@...>
주제: [openchain] OpenChain and ChatGPT – New Case Studies
This got a soft launch at the OpenChain Mini-Summit adjacent to Open Source Summit North America.
The OpenChain Project is releasing the first draft case studies created by ChatGPT on our GitHub. These are not intended to replace our community contributions, but to make it fast for people to add ideas and adjustments. This will specifically address one of the greatest challenges in creating new material: the initial time spent for drafting.
Why?
Our community feedback shows that people usually enjoy commenting and polishing more than drafting. Check them out and let us know what you think!
It took ChatGPT less than ten minutes to create eight case studies:
https://github.com/OpenChain-Project/Reference-Material/tree/master/Adoption-Case-Studies/Official/en/ChatGPT
Just to clarify:
(1) OpenChain always has a practice of community-reviewing material to ensure it looks correct.
(2) LLMs do not work by copying existing text. They work by the analysis of existing text to create prediction systems for guessing the next word in a sequence. On occasion, they may produce text mirroring existing text, but it should not be assumed that LLMs are creating Derivative Works under copyright law. Roughly the same applies to a human who learnt from existing material, and may or may accidentally replicate something too close to it when producing things on the topic. Both need editorial oversight to catch and correct issues.
Regards
Shane
On May 16, 2023, at 10:44, Divya Mohan <divya.mohan0209@...> wrote:
Hey Shane,
This seems like a great idea, modulo the two caveats below:
- We must ensure that it is always manually verified.
- We must ensure that the content ChatGPT is providing, since it is essentially "learning" these from other sources on the, is not trademarked under a different license.That might prove to be a conflict of interest.
However, I get a 404 on clicking the above link and on digging a little, it seems like the Adoption-Case-Studies folder is missing in the repo.
Is this an expected output due to the folder permissions?
Cheers,
Divya
On Tue, May 16, 2023 at 3:06 AM Shane Coughlan <scoughlan@...> wrote:
This got a soft launch at the OpenChain Mini-Summit adjacent to Open Source Summit North America.
The OpenChain Project is releasing the first draft case studies created by ChatGPT on our GitHub. These are not intended to replace our community contributions, but to make it fast for people to add ideas and adjustments. This will specifically address one of the greatest challenges in creating new material: the initial time spent for drafting.
Why?
Our community feedback shows that people usually enjoy commenting and polishing more than drafting. Check them out and let us know what you think!
It took ChatGPT less than ten minutes to create eight case studies:
https://github.com/OpenChain-Project/Reference-Material/tree/master/Adoption-Case-Studies/Official/en/ChatGPT
--
Best regards,
Divya Mohan
(https://linktr.ee/divyamohan)
Hi Divya
Just to clarify:
(1) OpenChain always has a practice of community-reviewing material to ensure it looks correct.
(2) LLMs do not work by copying existing text. They work by the analysis of existing text to create prediction systems for guessing the next word in a sequence. On occasion, they may produce text mirroring existing text, but it should not be assumed that LLMs are creating Derivative Works under copyright law. Roughly the same applies to a human who learnt from existing material, and may or may accidentally replicate something too close to it when producing things on the topic. Both need editorial oversight to catch and correct issues.
Regards
Shane
> On May 16, 2023, at 10:44, Divya Mohan <divya.mohan0209@...> wrote:
>
> Hey Shane,
>
> This seems like a great idea, modulo the two caveats below:
>
> - We must ensure that it is always manually verified.
> - We must ensure that the content ChatGPT is providing, since it is essentially "learning" these from other sources on the, is not trademarked under a different license.That might prove to be a conflict of interest.
>
> However, I get a 404 on clicking the above link and on digging a little, it seems like the Adoption-Case-Studies folder is missing in the repo.
>
> Is this an expected output due to the folder permissions?
>
> Cheers,
> Divya
>
> On Tue, May 16, 2023 at 3:06 AM Shane Coughlan <scoughlan@...> wrote:
> This got a soft launch at the OpenChain Mini-Summit adjacent to Open Source Summit North America.
>
> The OpenChain Project is releasing the first draft case studies created by ChatGPT on our GitHub. These are not intended to replace our community contributions, but to make it fast for people to add ideas and adjustments. This will specifically address one of the greatest challenges in creating new material: the initial time spent for drafting.
>
> Why?
>
> Our community feedback shows that people usually enjoy commenting and polishing more than drafting. Check them out and let us know what you think!
>
> It took ChatGPT less than ten minutes to create eight case studies:
> https://github.com/OpenChain-Project/Reference-Material/tree/master/Adoption-Case-Studies/Official/en/ChatGPT
>
>
>
>
>
>
> --
> Best regards,
> Divya Mohan
> (https://linktr.ee/divyamohan)
>
In both cases, an editorial review would be required to catch it was the actual suggestion but I think I did not sum it up correctly.But, yes, overall, I agree with the approach suggested.https://github.com/OpenChain-Project/Reference-Material/tree/master/Adoption-Case-Studies/Official/en/ChatGPT => This link still points me to a 404, however.Cheers,DivyaOn Thu, 18 May, 2023, 11:54 Shane Coughlan, <scoughlan@...> wrote:Hi Divya
Just to clarify:
(1) OpenChain always has a practice of community-reviewing material to ensure it looks correct.
(2) LLMs do not work by copying existing text. They work by the analysis of existing text to create prediction systems for guessing the next word in a sequence. On occasion, they may produce text mirroring existing text, but it should not be assumed that LLMs are creating Derivative Works under copyright law. Roughly the same applies to a human who learnt from existing material, and may or may accidentally replicate something too close to it when producing things on the topic. Both need editorial oversight to catch and correct issues.
Regards
Shane
> On May 16, 2023, at 10:44, Divya Mohan <divya.mohan0209@...> wrote:
>
> Hey Shane,
>
> This seems like a great idea, modulo the two caveats below:
>
> - We must ensure that it is always manually verified.
> - We must ensure that the content ChatGPT is providing, since it is essentially "learning" these from other sources on the, is not trademarked under a different license.That might prove to be a conflict of interest.
>
> However, I get a 404 on clicking the above link and on digging a little, it seems like the Adoption-Case-Studies folder is missing in the repo.
>
> Is this an expected output due to the folder permissions?
>
> Cheers,
> Divya
>
> On Tue, May 16, 2023 at 3:06 AM Shane Coughlan <scoughlan@...> wrote:
> This got a soft launch at the OpenChain Mini-Summit adjacent to Open Source Summit North America.
>
> The OpenChain Project is releasing the first draft case studies created by ChatGPT on our GitHub. These are not intended to replace our community contributions, but to make it fast for people to add ideas and adjustments. This will specifically address one of the greatest challenges in creating new material: the initial time spent for drafting.
>
> Why?
>
> Our community feedback shows that people usually enjoy commenting and polishing more than drafting. Check them out and let us know what you think!
>
> It took ChatGPT less than ten minutes to create eight case studies:
> https://github.com/OpenChain-Project/Reference-Material/tree/master/Adoption-Case-Studies/Official/en/ChatGPT
>
>
>
>
>
>
> --
> Best regards,
> Divya Mohan
> (https://linktr.ee/divyamohan)
>
Correct link here:
https://github.com/OpenChain-Project/Reference-Material/tree/master/Adoption-Preparation/Case-Studies/ChatGPT
Your review of the case studies would be appreciated :)
Actually, one of the surprising things about the ChatGPT work was that it created reasonable, if short, case studies with a very simple but logical structure. It clearly went into “basic format of case studies” mode, and executed with (mostly) accurate reference to OpenChain and our standards.
tl;dr: this looks like a good avenue to go from “blank page” to “some material to get started” around simple tasks like basic case studies etc.
Regards
Shane
On May 18, 2023, at 15:31, Divya Mohan <divya.mohan0209@...> wrote:
In both cases, an editorial review would be required to catch it was the actual suggestion but I think I did not sum it up correctly.
But, yes, overall, I agree with the approach suggested.
https://github.com/OpenChain-Project/Reference-Material/tree/master/Adoption-Case-Studies/Official/en/ChatGPT => This link still points me to a 404, however.
Cheers,
Divya
On Thu, 18 May, 2023, 11:54 Shane Coughlan, <scoughlan@...> wrote:
Hi Divya
Just to clarify:
(1) OpenChain always has a practice of community-reviewing material to ensure it looks correct.
(2) LLMs do not work by copying existing text. They work by the analysis of existing text to create prediction systems for guessing the next word in a sequence. On occasion, they may produce text mirroring existing text, but it should not be assumed that LLMs are creating Derivative Works under copyright law. Roughly the same applies to a human who learnt from existing material, and may or may accidentally replicate something too close to it when producing things on the topic. Both need editorial oversight to catch and correct issues.
Regards
ShaneOn May 16, 2023, at 10:44, Divya Mohan <divya.mohan0209@...> wrote:
Hey Shane,
This seems like a great idea, modulo the two caveats below:
- We must ensure that it is always manually verified.
- We must ensure that the content ChatGPT is providing, since it is essentially "learning" these from other sources on the, is not trademarked under a different license.That might prove to be a conflict of interest.
However, I get a 404 on clicking the above link and on digging a little, it seems like the Adoption-Case-Studies folder is missing in the repo.
Is this an expected output due to the folder permissions?
Cheers,
Divya
On Tue, May 16, 2023 at 3:06 AM Shane Coughlan <scoughlan@...> wrote:
This got a soft launch at the OpenChain Mini-Summit adjacent to Open Source Summit North America.
The OpenChain Project is releasing the first draft case studies created by ChatGPT on our GitHub. These are not intended to replace our community contributions, but to make it fast for people to add ideas and adjustments. This will specifically address one of the greatest challenges in creating new material: the initial time spent for drafting.
Why?
Our community feedback shows that people usually enjoy commenting and polishing more than drafting. Check them out and let us know what you think!
It took ChatGPT less than ten minutes to create eight case studies:
https://github.com/OpenChain-Project/Reference-Material/tree/master/Adoption-Case-Studies/Official/en/ChatGPT
--
Best regards,
Divya Mohan
(https://linktr.ee/divyamohan)
Hi all,
I did skim through the new case studies yesterday, and agree they are surprisingly (depressingly) good. I did wonder whether it would be worth recording the prompt used to create each one – but I also confess that I'm not sure what purpose that would serve, apart from slightly assuaging my general fear of using such generated material.
I think the fact that each case study is referring to a clearly fictional/hypothetical situation is good – IMO, it's important to be clear that "this is a path you could take", rather than "this is something that Company X actually did successfully, and they are available to answer any questions you may have." 😊 It also means that I find it less annoying that the documents lack the gnarly details ("Yes, but how?").
At the risk of increasing the number of such studies, it did occur to me that one likely scenario not touched on but worthwhile considering is one we face a lot at ADI: Company X acquires Company Y, and the processes of the two formerly-separate orgs need to be aligned.
(Did I just volunteer myself to write a playbook? Globbits. Mental note: drink more coffee before replying to Shane.)
steve
From:
main@... <main@...> on behalf of Shane Coughlan <scoughlan@...>
Date: Thursday, 18 May 2023 at 07:48
To: OpenChain Main <main@...>
Cc: OpenChain Education <education@...>
Subject: Re: [openchain] OpenChain and ChatGPT – New Case Studies
[External]
Hi Diva
Correct link here:
https://urldefense.com/v3/__https://github.com/OpenChain-Project/Reference-Material/tree/master/Adoption-Preparation/Case-Studies/ChatGPT__;!!A3Ni8CS0y2Y!-JI5r1F5IeFDsHbq63lwAh7eVBl2M25_S61DnkmLjWqDLroUe3DCJhzoj0-aqGotps3i4WCzxwJ9udksi0XpYcNgLiwrSQo$
Your review of the case studies would be appreciated :)
Actually, one of the surprising things about the ChatGPT work was that it created reasonable, if short, case studies with a very simple but logical structure. It clearly went into “basic format of case studies” mode, and executed with (mostly) accurate reference
to OpenChain and our standards.
tl;dr: this looks like a good avenue to go from “blank page” to “some material to get started” around simple tasks like basic case studies etc.
Regards
Shane
> On May 18, 2023, at 15:31, Divya Mohan <divya.mohan0209@...> wrote:
>
> In both cases, an editorial review would be required to catch it was the actual suggestion but I think I did not sum it up correctly.
>
> But, yes, overall, I agree with the approach suggested.
>
>
https://urldefense.com/v3/__https://github.com/OpenChain-Project/Reference-Material/tree/master/Adoption-Case-Studies/Official/en/ChatGPT__;!!A3Ni8CS0y2Y!-JI5r1F5IeFDsHbq63lwAh7eVBl2M25_S61DnkmLjWqDLroUe3DCJhzoj0-aqGotps3i4WCzxwJ9udksi0XpYcNgTDjBLXc$ =>
This link still points me to a 404, however.
>
> Cheers,
> Divya
>
> On Thu, 18 May, 2023, 11:54 Shane Coughlan, <scoughlan@...> wrote:
> Hi Divya
>
> Just to clarify:
>
> (1) OpenChain always has a practice of community-reviewing material to ensure it looks correct.
>
> (2) LLMs do not work by copying existing text. They work by the analysis of existing text to create prediction systems for guessing the next word in a sequence. On occasion, they may produce text mirroring existing text, but it should not be assumed that
LLMs are creating Derivative Works under copyright law. Roughly the same applies to a human who learnt from existing material, and may or may accidentally replicate something too close to it when producing things on the topic. Both need editorial oversight
to catch and correct issues.
>
> Regards
>
> Shane
>
> > On May 16, 2023, at 10:44, Divya Mohan <divya.mohan0209@...> wrote:
> >
> > Hey Shane,
> >
> > This seems like a great idea, modulo the two caveats below:
> >
> > - We must ensure that it is always manually verified.
> > - We must ensure that the content ChatGPT is providing, since it is essentially "learning" these from other sources on the, is not trademarked under a different license.That might prove to be a conflict of interest.
> >
> > However, I get a 404 on clicking the above link and on digging a little, it seems like the Adoption-Case-Studies folder is missing in the repo.
> >
> > Is this an expected output due to the folder permissions?
> >
> > Cheers,
> > Divya
> >
> > On Tue, May 16, 2023 at 3:06 AM Shane Coughlan <scoughlan@...> wrote:
> > This got a soft launch at the OpenChain Mini-Summit adjacent to Open Source Summit North America.
> >
> > The OpenChain Project is releasing the first draft case studies created by ChatGPT on our GitHub. These are not intended to replace our community contributions, but to make it fast for people to add ideas and adjustments. This will specifically address
one of the greatest challenges in creating new material: the initial time spent for drafting.
> >
> > Why?
> >
> > Our community feedback shows that people usually enjoy commenting and polishing more than drafting. Check them out and let us know what you think!
> >
> > It took ChatGPT less than ten minutes to create eight case studies:
> >
https://urldefense.com/v3/__https://github.com/OpenChain-Project/Reference-Material/tree/master/Adoption-Case-Studies/Official/en/ChatGPT__;!!A3Ni8CS0y2Y!-JI5r1F5IeFDsHbq63lwAh7eVBl2M25_S61DnkmLjWqDLroUe3DCJhzoj0-aqGotps3i4WCzxwJ9udksi0XpYcNgTDjBLXc$
> >
> >
> >
> >
> >
> >
> > --
> > Best regards,
> > Divya Mohan
> > (https://urldefense.com/v3/__https://linktr.ee/divyamohan__;!!A3Ni8CS0y2Y!-JI5r1F5IeFDsHbq63lwAh7eVBl2M25_S61DnkmLjWqDLroUe3DCJhzoj0-aqGotps3i4WCzxwJ9udksi0XpYcNgiZ8EJHc$ )
> >
>
>
>
>
>
>
>