WIP: HumanCentric AI - Charter Draft

Hi All,

Welcome to the people who have joined recently.  May I first extend a
welcoming motion, for people to be encouraged to introduce themselves and
what it is that they'd like to see become part of this group's works; and
in-turn also, the future of Human Centric AI systems and personal
experiences...

Note also; i've been alerted to TPAC:
https://www.w3.org/2023/09/TPAC/Overview.html whereby: W3c needs to know if
the Community Group(s) would like to meet during TPAC 2023; and that we are
asked to discuss & let them know *by 8 May at the latest.*

Also, I am interested to learn which days / times best suit participants
for zoom meetings, and what communications platforms members would prefer
for 'unofficial' group work, as may be specific to w3c group members who've
signed-up, so that all involved are on the same footing...

I've historically used various platforms; at the moment, I prefer discord.
although i have earlier used zulip, gitter, irc (w3c has an IRC channel for
the group, in-effect); and various others solutions.  in theory, we could
also look to build an app that uses something like solid, however if that
was something that i was tasked with, without additional resources, then
i'd consider whatever i make experimental - at this stage; and that,
moreover, that i am of the opinion that efforts should be made to encourage
a decentralised approach to human centric ai related works, whereby
official w3c works end-up being done via official w3c processes, channels
and platforms... but again, open to suggestions.

Attached is a link to the Human Centric AI CG Charter - Draft document that
I've been working on myself..

Link is:
https://docs.google.com/document/d/1ragkWVvqO5xqqYS4XPh2m4dqm8fWjH8vjKtZDzSB_kc/edit#


The document is in British/Australian English (rather than US English); and
I've not done a full compatibility check with pre-requisites, if there's
any significant issues - please let me know.

EXPLANATORY NOTES:

1. It is still considered an early draft. Input is sought.

2. It seeks to provide a basic summary definition of what Human Centric AI
is and/or does; and thereby seeks to provide some context for the group.

3. It seeks to provide a basis upon which an affirmation of honourable
intents is able to be conferred upon group members.  A teaching I took on
board was that an affirmation assumes that a person is honourable; whereas
being sworn to act in a particular way, considers that the requirement to
act honourably is unusual.

4. It refers to the UN related "Digital Transformation Agenda', whilst
loosely, and thereby, some variations based upon
https://www.un.org/en/about-us/un-charter/ are used to define 'community of
practice' related considerations.

5. It considers transparency & accountability, probity, duties,
responsibilities and the reasonable limitations that now currently apply.

In-effect, the intended useful purpose of the method; seeks to bestow
personal responsibility at the center of the structures, and to thereby
seek to support governance upon the basis that where there are issues, they
should for the most-part (save exception); relate to how a person /
persons, may have breached their own values; and thereby seek to form means
to be supportive.

6. It outlines the role of committees and related structures.
The consideration being, that if this does in-fact become empowered with
the support i think it deserves, that there is a lot that could be done;
and that each constituency, may be better equipped via a focused group
effort, whereby the 'brains trusts' for the nuances and specific
considerations associated to that particular constituency, may be best
presented by its team structures, et.al.

7. Authorised Representatives & Delegation
It seeks to provide for 'authorised representatives' to assist by taking on
particular commissions and/or missions, on behalf of the group.

It also considers that - not all people have good relationships with all
other people; and that, there may be issues that do not need to be
laundered in public domain; and/or that people may become busy and/or
unavailable due to various reasons, alongside other circumstances whereby
the means to delegate agency becomes a useful mechanism in-order to serve
the common good.

8. Registries
It injects a concept of employing part of the administrative email system,
as a baseline means to (privately); log the amount of effort that has been
usefully provided towards the activities and/or enabling required works
associated to the groups purposes / efforts.

There is no financial mechanisms associated to this, however, i am
effectively putting forward a suggestion that we seek to provide some level
of accounting, on a best-efforts basis, to better honour the time and
effort people put into these works; and to thereby,

8.1 uses the practice method as a sort of tutorial / experimental
methodology, to better understand the relationship between ESG, SDGs, Human
Rights and the challenges we have in seeking to support the efficacy of
those terms / concepts.

8.2 that we seek to improve hygiene about provenance, whereby the efforts
of people become part of an accounting system, even if it is just
providing capacities for acknowledgement; that may in-turn also, act to
better address issues where works are levied with the encumbrances of false
attribution; or similar.

The Proposed method is to use the internal email system that currently
provides chair-persons and w3c notifications when a person has joined, etc.

i am very open to having a broader discussion about this, and if anyone is
aware of similar sort of practices - other than git / github (ie: commits);
or indeed also, google docs (which has an API to pull contributor info,
afaik); let me know.  afaik, it's a bit 'innovative'....  thoughtfully
so... imo.

9. The role of the Chair-person(s)

It provides information about the role of chair-persons, both as to be
mindful that some may not understand corporate governance generally; and
also, in terms as applied to the charter for the group.

Therein note also; that i consider the role that i am currently fulfilling
as chair-person to be an interim measure as a form of 'care-taker'
provision; whereby, once the group is put together sufficiently, and that
there is a level of 'common sense', shared values, clarity on objectives
and that group members have had the opportunity to get to know one-another
a bit at least, etc.  Then there should be a formal election for the
role(s) of chair-person(s).

I have sought  to employ some of the 'city of london' (londinium) related
principles; although not alot and/or, only loosely, but therein the idea of
a 1 year term, rather than any longer.  Similarly also, the concepts
relating to committees (ie: livery companies / aldermen like) this in-turn
seeks to form an innate structure where the role of persons is to serve the
interests of the group.

10. Elections
Provides general outline about elections, nominations, mechanisms through
which the group may discharge its capacities to stand an officer down /
relieve them of their role/duties, etc.

It also makes statements as to seek people to be honest about their level
of knowledge of internet governance, w3c and related bodies of knowledge;
thought usefully important to the roles; without seeking to necessarily
preclude any such persons without that knowledge, from seeking to be
elected to a role of importance (ie: chairperson, or committee or
authorised representative, etc.). It also outlines the nomination periods,
etc.

11. Decision process
this is based upon the terms drafted from the Credentials CG; and seeks to
outline how resolutions are able to be defined.

12. The Registry
The registry concept supplements the considerations made by (8) above, with
some procedural methodologies / suggestions.

13. How changes to the charter are to be performed, once it has been
formally approved by resolution.

14. WIP Activity Schedule / Works.

I have started on this, however - have not spent so much time on it; yet,
notwithstanding existing works otherwise, that are sought to be presented
to the group for consideration.

Fundamentally, the principle idea would be that the group identify
different project opportunities, and that they then become listed as part
of the artifacts to the charter; once they've been approved as 'official cg
projects', notwithstanding the means - as should be encouraged, for people
to identify, explore, incubate and develop ideas, until such a point as
they may then be presented for group resolution; and thereby also, the
means to establish a committee (sub-committee in-effect) for that groups
activities specifically.

To some degree, the considered method to address this, relates to my former
experience working at an ISP where the SFOA (standard form of
agreement)[2]; basically had, the main part of the document, providing
general terms, which was then supplemented by constituencies to the
agreement that were produced as part of the product development cycle, for
new products/service.

in this way; the ability to define terms for new products, did not overtly
incur and/or interfere with the terms generally, for the products offered
by the ISP. attached is a link to an example of this document, as is
otherwise noted[2].

Personally, i think one of the things that could be started on immediately,
is to do work on,

a. formatting UN Human Rights Instruments based RDF 'ontologies', so that
we're then able to make use of them & empower others to do so also.

b. forming some resources that seek to provide a framework for
discerning the qualities of different types of AI agents.

c. improving the underlying works on defining both; online attack vectors
(noting some previous work was done on 'online harms'), as well as seeking
to evaluate 'digital justice' related requirements, and in-turn also
'safety protocols', which imo; starts with AI employable human rights based
instruments, or what i call 'values credentials'..

d. I've been working on natural language 'systems', that somewhat augment
the way 'ontologies' function. This work in-turn also led to considerations
about how to ensure support for all languages of prayer, which is something
that is not able to be achieved otherwise today. Irrespective of whether or
not these systems do become part of WIP, the underlying requirement of
greater importance is ensuring that systems are equipped with some sort of
useful portability / export/import function, as to ensure people do not
become trapped in a particular product and/or service.

otherwise noting; that i have a list, and i'm actively seeking input from
others about what their views and thoughts are.

AND THAT; may of my considerations and assumptions are linked with fairly
in-depth comprehension of semantic web related ecosystems; and in-turn,
early works (and intended applications of) rww (solid), web-payments,
credentials/verifiable claims, etc.  if people are not aware of how
technologies like solid (and/or some of the 'my data' related solutions)
function; then, i'd suspect that the basis upon which sense-making is able
to be formed, is different.

RELATED ISSUES AND RELATED CONSIDERATIONS

a. It is presently 12 point font, lots of spacing and may be considered a
long document..
i. PROs; the means to produce apparatus, relating to 'human centric'
information/knowledge management approaches, is (sadly) an expansive
endeavor of a non-trivial nature; that requires illustration of structures
that are unlike other systems. These structures are considered to be
required in-order to get to the point where Human Centric AI is deliverable.

There are also various 'human rights' related considerations that are
declaratively noted in the charter, which both require more texts; but may
therefore also, provide improved means for resilience should disputes arise
in future that need to be considered in the context of human rights related
texts and associative deliberations.

ii. CONs; its as longer document than would otherwise be sought to be
defined.  Improvements to the draft should improve the articulation in more
succinct means, and/or reduction could be improved by referring to other
documents as constituencies and/or meanings; but, this wouldn't actually
act to reduce the total amount of terms for the charters entire
embodiment...

I am actively seeking input...

b. The headings and 'flow' of the document could be improved.  There may
also be opportunities to integrate statements, ie: the registry components
are presently in two different areas of the document.

c.  the creation of this document; considers also, that W3C and in
particular therein, W3C CGs have a particular role in the broader Internet
Governance ecosystem[3].  Historically (~2016-9) whilst working on related
works, i had a view that part of the solution to support various
requirements may well be via forming a global Topic Chapter / SIG in
ISOC[4]; as to thereby support discussions of a broader - social - nature,
and in-turn act to support and be supported by regional ISOC chapters[5].

Whilst conversations are underway, discussing the future of internet
governance via the UN GDC[6] works, amongst other places; part of the
consideration / deliberation, is that the works undertaken here, could or
should, perhaps optimistically, be structured in such a way that is of
empowering usefulness, for future scalability and growth.

As such, part of the consideration innately sought to be 'built into' this
charter, is the way it may in-time be taken-up and employed; and/or, how an
broader 'community of practice' may in-turn employ compatible charters,
without unnecessary future delays (to the best of our abilities / best
efforts, etc.); as may in-turn, be better equipped as a consequence of the
works undertaken within this CG...

note also; that whilst there is no capacity to support ensuring people are
able to be paid fairly for useful work, that may have fairly remarkable
consequences upon humanity, noting my previous experience relating to
credentials[7]; particularly where people are engaged for the express
purpose of seeking to act as an agent for the good of humanity,
independently; in a manner that is different, to others who are part of
quorum, as representatives of various w3c members[8];  ISOC 'topic
chapters' or SIGs are different, as they do require an incorporated entity
and may therefore become part of the solution in how we better support the
human rights of people engaged in work for the good of humanity.  However,
this is presently an 'unknown'; nonetheless, seeking to provide some
capacity for future positive change in this area, which may in-turn provide
support as to inform approaches that may be employed more broadly, is
considered - useful...

d. Current and future resources.
I have obtained a domain name (humancentricai.xyz) as well as establishing
a github organisation, to host code. There are also w3c resources[9],
including zoom.  I have not yet added reference to these methods / systems,
noting that I'm still working through it...  I've investigated github apps
as a potential means to support polls / electronic voting systems;

I am also unclear whether and/or how secret ballots may be supported and/or
in what circumstances, and need to do more investigations.

e. skillsets and accessibility / diversity

i. I would also like to get a sense of what methods of engagement people
are familiar with; imo, there's great benefits to be gainfully supported;
by supporting diversity, and that means that some participants may not be
web developers specifically, but rather - luminaries in various fields,
many of which don't require a person to know how to use git, or IRC, etc.

Additionally, if the group elected to make - perhaps experimentally - some
solid based apps / functional stuff; then what might the implications be..
 a recent TimBL Video re: solid is,
https://www.youtube.com/watch?v=Ay7GSLX9tHM

Noting also, that the implementation that i've been working on; isn't
presently 'solid based', but one of the safety protocols is about ensuring
that people can be migrated to some functional alternative (in the event of
a problem, etc.), which i think is most-likely solid.  AND THAT, i'm not
interested in 'owning language' or other basic things people need for
freedom of thought / human rights purposes; and that, broadly otherwise,
safety protocols are needed ASAP, imo.

IMO: it would be good for the communities efforts to get and seek-out
wide-ranging inputs, as to help us inform our approach, and in-turn better
define what may be considered 'fit for purpose', et.al.

ii. IMO, historical works are very focused on producing code /
specifications outputs; based upon the requirements for incrementally
developing new constituencies for 'the web', etc.  There are some big
shifts happening, and whilst alot of work has been done - many don't know
much about it generally, whether it be about the w3c[10], internet
governance[3], the role of patent pools[11] and the related implications
associated to W3C[12] as a venue for related works, etc.

And those that do understand those sorts of considerations very well, may
not be best empowered without input from others; that may be brought about
by presentations to groups via online learning sessions or other means, as
to support the process of philosophical engineering[13], as does apply to
Human Centric AI related requirements; now, sometime on from the time TimBL
presented his thoughts on manipulating reality and semantics back in may
1994[14].

Overall, nowtherefore, figuring out how to better support diversity; and
in-turn, collaboratively figure out the critical path requirements; is
thought to be amongst the sorts of things / objective purposes & Tasks that
we could get onto ASAP.

Finally also, should the work on the human rights stuff be undertaken, then
we may well update the charter once that's done to make use of the work, in
a manner that's not unlike the way creative commons[15] now functions.

Faithfully,

Timothy Holborn


LINKS:
[1] DRAFT CHARTER DOCUMENT LINK:
https://docs.google.com/document/d/1ragkWVvqO5xqqYS4XPh2m4dqm8fWjH8vjKtZDzSB_kc/edit?usp=sharing

[2] NETSPACE SFOA EXAMPLE:
https://docs.google.com/document/d/1Ia5y3IDemMZ-9y-LlJeSIM85PRabp5zUpMVaJZy3ojc/edit

[3] https://en.wikipedia.org/wiki/Internet_governance
[4] https://www.internetsociety.org/sigs/
[5] https://www.internetsociety.org/chapters/
[6] https://www.un.org/techenvoy/global-digital-compact
[7]
https://www.w3.org/community/credentials/2014/08/06/call-for-participation-in-credentials-community-group/

[8] https://www.w3.org/Consortium/Member/List
[9] https://www.w3.org/community/about/tool/
[10] https://www.w3.org/Consortium/
[11] https://en.wikipedia.org/wiki/Patent_pool
[12] https://www.w3.org/Consortium/Patent/
[13] http://www.w3.org/2007/09/map/main.jpg
[14] note, snippet - full clip noted in comments
https://www.youtube.com/watch?v=UkjyCPuTKPw
[15] https://creativecommons.org/ns