Decisions and motivations

[lkorczowski]

Hi,

Thanks you very much for this great proposal. However, it would be really interesting to have the explanation of the decisions for the content of the manifesto.

Indeed, for example, the choice of using Hippocrates' name is pertinent for communication purpose but is it really accurate? There are many professional oaths (e.g. Archimedean) and deontology code (e.g. lawyers, notary, etc.) so why used "Hippocratic oath" specifically ? It could be trivial for many (maybe just a commodity decision) but the consequences are important... just as our algorithms.

I believe that the oath motivations and decisions should reflect the oath itself : be open. Maybe it is and I missed the discussions and I'm sorry if it is the case (maybe a link explaining the decision process would be enough).

If it is not, would a decision tree such as https://www.kialo.com/ be interesting ?

Thanks !

[Soliine]

Hello Louis,

Thanks a lot for your feedback and suggestions! It's very cool, nice and encouraging to read your message, thanks for taking the time! They are great inputs. Here's my replies:

" However, it would be really interesting to have the explanation of the decisions for the content of the manifesto." > I agree, it'd be great and we will start doing so. We built the project last year during a 3-month acceleration program organized by French non-profit Data for Good. We could only meet once a week and time was counted so we didn't document our decisions online during these first months but now that we have more time, it's a good idea to start sharing why we wrote what we wrote. To do this, we need to talk about tools and see where we publish and organize this! I replied about Kialo.com below.

"Indeed, for example, the choice of using Hippocrates' name is pertinent for communication purpose but is it really accurate? There are many professional oaths (e.g. Archimedean) and deontology code (e.g. lawyers, notary, etc.) so why used "Hippocratic oath" specifically ? It could be trivial for many (maybe just a commodity decision) but the consequences are important... just as our algorithms." > Good point. We indeed discussed about it. We finally decided to call it "Hippocratic oath for data scientists and for each person working with data" for communication purpose indeed and to have non-data scientists onboard (like product manager, etc). We wanted to be quickly identified by people who don't know anything about data and these professions, who would not understand the project if we name it with the name of a great data scientist, but who know Hippocrates' oath and understand that something similar could be applied to data "science". But nothing is definitive! It was our choice last year, it was also a way to play with this sacred name (our url is hippocrate.tech) and it was quite efficient in terms of public understanding and media coverage but we can still discuss about. We mostly took decisions IRL during focus groups that took place every Wednesday night in Paris (we briefly explained it in French here : https://hippocrate.tech/team/), so the pain is that we don't have any trace of these discussions on an online forum but we can still think of a way to attach discussion to each part of the oath's text. This Github as well as the Framavox was though to let anyone discuss any part of the oath.

"I believe that the oath motivations and decisions should reflect the oath itself : be open. Maybe it is and I missed the discussions and I'm sorry if it is the case (maybe a link explaining the decision process would be enough)." > Agreed. You indeed missed the IRL discussions we had during focus groups during which we collaboratively wrote the text. But we could open a GitHub issue or a Framavox discussion for each sentence or group of sentences of the oath where we would explain them and discuss them. We would publish the link under each sentence on hippocrate.tech.

"If it is not, would a decision tree such as https://www.kialo.com/ be interesting ?" > It's an interesting tool. Do you think the "pro/cons" structure of conversations would be relevant for our project ? Also, I tend to prefer free software solutions like Framavox. At the moment, I would open a thread for each sentence of the oath on Framavox, put the URL of the thread under each sentence and we would explain the decisions behind these sentence on these threads and have people comment and suggest editions on these thread.

[lkorczowski]

Thanks for the thorough answers and I want to thanks all the contributors for the time and effort put in this project.

I think we agree on fact that we need to open the decision process (and making it online is the key here). Great !

IMO, the online process should respect the following specifications:

• The online decision process should strengthen the decisions while keeping an agile framework (some decisions could be re-evaluated based on new inputs) by keeping tracks of the arguments and counter-arguments. Fictive example: "All the pro/con arguments of the term "Hippocratic" should be explained"
• The online decision process should formalize the range and the limitations (boundaries, grey areas, etc.) of the oath by giving theoretical motivations and practical examples. Fictive example: "The term of "Hippocratic oath" should be explained in the application of medicine, what are the philosophy, what can't be done by the oath-taker in medicine and why it matters for data science"
• The online decision process should be user-friendly, should not overwhelm new users, and be participative. Moreover, any argument/counter-argument should be easily added while keeping track of the context (ergonomic). Fictive example: "I am a medical doctor, I don't know data science but I think that the use of the term of "Hippocratic oath" is confusing, I want to add my arguments easily without disturbing the rest of other statements on this issue (that are still valid)"

If we agree on those affirmations (we need a decision tree here ! hahaha), the next question would be : how to execute that goal ?

• I understand the appeal of a free software, I do not have any problem with Framavox. However I doubt that, in the current state, Framavox succeeds to have a efficient decision service. To my knowledge it offers not a lot more that a standard forum with a voting module but I could be wrong. Therefore I am not sure that Framavox would respect the previous affirmations
• Kialo has the advantage to have a hierarchical argumentation that helps to see immediately the weakness/validity of any argument. However it is far from being perfect, but I don't know any better platform.
• "pro/cons" structure of conversation is the basis of dialectic method, I do not know any other formal efficient way to dispute arguments in details (when dozen or hundred people can contributes), do you have an other of method in mind ?

[sammous]

Hello @lkorczowski,
Thank you for your very interesting feedback.
As Soline said, the goal of this project was to empower people to build together an oath, that's why there was IRL sessions every week where everyone was invited to discuss and debate the decisions. We used then Framavox and Github as tools to continue to have open decisions and be location independant, but maybe it is not the best tools for such purpose.
I didn't know kialo and it seems to be a cool tool to debate decisions, and is more user friendly than Github or Framavox, but unfortunately doesn't seem open source.
Maybe we could still contact them to know if they respect GDPR etc... ? Also it would be a way to "internationalize" the oath too ?