Human Centric AI: IGF & (UN) Values Credentials? Human Rights Tooling Progress?

Hi All,

I’ve been participating a bit in IGF, whilst also watching other parts of
it on YouTube.


FWIW: IGF on YouTube: https://www.youtube.com/@igf

List of sessions:
https://docs.google.com/spreadsheets/d/1JEspoS6w3CF79t7pWAz0KkH4UCvbAQZnYto0P5ZRgyw/edit?usp=sharing


The announcement by Vint & Co about “Human Centric Internet” was a
surprise, historically he was more focused on People Centered Internet;
whilst i was more focused on Human Centric content layers; ie: human
centric web, human centric licensing for software tooling & Human Centric
AI.

In anycase, Whilst very much welcomed; I do wonder what will become of the
use of this term…

So, here I am… continuing to do my best, some years after defining the term
via related W3C works, etc…  noting also, the temporal alignment that has
been in play with myData efforts over the years…

There is soon another open-session to highlight the SIG Topics proposed,
including what I’ve sought to bring about for Human Centric AI..  This
topic is extremely dense. There’s alot to it, technically a lot has been
done; but more needs to be done, and alot more needs to be done in the
social domain.

The objective of this SIG, from my POV is to bring about the tooling needed
to instantiate Human Centric AI; which instrumentally requires, the means
for a personal / private software agent to process information - locally -
for many; and for those without, via a trusted infrastructure provider.

I called this ‘information banking’ infrastructure between 2000-12; but
then, ‘knowledge banking’ infrastructure - which has various requirements,
indeed, i a hopeful ISOC chapters will be foundational providers within
many different jurisdictions; as to support the human rights needs of the
most vulnerable, but it seems we are a long way from this potential future
outcome…

And indeed also; that these works via ISOC may compliment & support W3C
related works & that if successful, we may then use ‘human centric ai’
infrastructure, to promote humanitarian works online; where people
collaborate / cooperatively work with one-another to support / produce,
humanitarian tech / knowledge, outcomes…  support peace infrastructure
projects, and our means to deliver SDGs, etc.. As somewhat considered by an
earlier, pre-covid, SIG application effort (that was incomplete).

https://drive.google.com/file/d/1o1FrGelPmWfA6olhKik--UzSBGH1Rz4o/view?usp=sharing


But today,

I struggled once again to get consideration about the importance of
figuring out how best to support ‘values credentials’ and in particular,
those relating to human rights instruments.  It is strange and quite
frustrating to me, when i’m in human rights groups in forums like IGF or
WSIS, and i ask about how to get these works done - and it seems, they
really don’t understand…

https://soundcloud.com/ubiquitous-au/my-question-human-rights-instruments-for-digital-wallets-valuescredentials


It seems as though, the major focus is in-fact a major global thin-client
model, perhaps with overly expensive edge devices; as so few members of our
human family discuss, how it should be governed.

SO, that’s not what i’m looking for; i’m looking to ensure people have the
means to materially support their own human rights, in courts of law, with
the useful benefit of the evidence about their lives, their experiences…
even if they’re made homeless, due to some problem, that needs peaceful
means for resolution.  Even for refugees, or alleged refugees..  That is,
until the evidence is reviewed.

So; we are a people of many different places, languages, cultures, cohorts,
views, beliefs, systems of law, etc…  around our world; but some of the
instruments that we all know of, and could reasonably be held to comply
with in a court of law; if associated with agreements between one-another,
are the values outlined in the body of human rights instruments.  But these
instruments have been made long before broadband, long before the great
lockdown that occurred as early credential use-cases were implemented
world-wide…  so, the way those instruments are worded, they generally seek
to get governments or ‘states’ to agree to those principles; not people,
not one-another…

So, imagine if we all had our own domain names; and those domain names, had
the tools needed so that we managed our own ‘addressbooks’; and that, the
way we form relationships, whether they be ‘introductions’ or
‘invitations’, with one-another needs some form of agreement; as there is
no ‘thing’ like a facebook terms of service - this doesn’t need ‘social
media platforms’, its simply between one-another.  People are then allowed
to have relationships between one-another, without facebook, or similar…

There’s some old papers about this sort of thing;

https://drive.google.com/drive/folders/1lpeoEFowRcq3VTAp5LH6cFN251O9g9iE

https://web.archive.org/web/20130122051820/http://jeffsayre.com/2010/09/15/web-3-0-smartups-the-social-web-and-the-web-of-data/

I’m now working on a browser-extension to extend ‘the web’ to support the
‘social layer’.

https://github.com/WebCivics/SocialWeb-WebExtensionDev-v4/blob/main/valuecredentials/flow.md

These 'agreements' help to define how the 'data' or commons associated to
the relationship, is managed, noting that relationships change overtime -
forging memories, as are important constituencies to our 'personal
ontology' and related informatics of life.

So, whilst the semantic model for defining declaratively the logic
associated with the meaning of terms; so that logical programming can be
used to help identify and then support discussion about any breaches that
may occur to agreements made between people via systems of agreements that
changes overtime; there nonetheless needs to be work to transform these
important instruments from statements that are defined as seeking the
‘state’ protect these rights of us all; to, what i consider to be a
supportive ecosystem, where people are able to make declarations that they
intend to support those values between one-another…  but who, is it, that
will,


   1.

   Define the language
   2.

   Provide the URIs where the ontologies are thereby hosted?


Ie:
https://docs.google.com/spreadsheets/d/1tJkhVcIqCSdzP6mZ3UxAB8za2z3jARofZF0M9DiUdfw/edit#gid=224570733


Its now been a few years of trying to find out, and its still difficult to
find the ‘interested party’, perhaps, if anyone is at the IGF and they
could get someone to follow-up or find out; that would be great, if not,

Then consider this to be one of the ‘to-do’ items; that I think, would be
good for the ISOC Human Centric AI SIG could get stuck into…  at least, to
make some recommendations in one or more languages only, if not more…

Here are some examples; re: Universal Declaration of Human Rights

Article 12

Original:

No one shall be subjected to arbitrary interference with his privacy,
family, home or correspondence, nor to attacks upon his honour and
reputation. Everyone has the right to the protection of the law against
such interference or attacks.

vs.

The Parties Agree to make best efforts, by their own dealings and
decisions, to ensure that no one is subjected to arbitrary interference of
their privacy, family, home or correspondence, nor to attacks upon their
honour and reputation. The Parties Agree that everyone has the right to the
protection of the law against such interference or attacks.

Article 20

Original:

   1.

   Everyone has the right to freedom of peaceful assembly and association.
   2.

   No one may be compelled to belong to an association.


Vs.


   1.

   The Parties Agree to support one another right to freedom of peaceful
   assembly and association.
   2.

   The Parties recognize and agree that No one may be compelled to belong
   to an association.


Article 24

Original:

Everyone has the right to rest and leisure, including reasonable limitation
of working hours and periodic holidays with pay.

VS.

The Parties recognize and support the right of others to rest and leisure,
including reasonable limitation of working hours and periodic holidays with
pay.

An english example is
https://docs.google.com/presentation/d/1q6Cl__FbUddtKtBPxw2KDviO_PEl7Ti-XYZOT1EQM14/edit#


A VERY QUICKLY produced ISOC related example;

https://github.com/WebCivics/isoc-hc-ai-sig-prep/tree/main/ontology

A list of instruments identified some time ago:
https://www.humancentricai.org/Our-Values.html

There’s also some older work: https://github.com/WebCivics/ontologies but
it still needs alot of work.  Fundamentally also, there’s some work on
natural language ontologies; to improve Human Centric AI supports, that is
hoped to be done - thought to be complementary to these works…

In my view, these sorts of works also thereby extend to many other ‘safety
protocol’ related works, including ‘agent labelling’ (ie: being able to
denote the works / involvement of software agents)

I hope this provides a little more information about the sorts of things I
hope to work through inter-nationally, such as to produce the humanitarian
infrastructure needed at this digitally transformative stage of
humanitarian development.  I hope we can get the basics done, as to then
uplift our capacity to make progress with respect to much needed peace
infrastructure for future medicinal earth outcomes; and foundations for
humanitarian sociosphere / biosphere, advancement.


But at the moment; i'm still quite frustrated by the apparent inability to
figure out how to better engage with others, as to improve support for our
means to make use of UN Human Rights values, in our electronic agreements
with one-another; which i know is different to platforms where agreements
are issued, like mandates, but i don't think that should be the only way of
defining the 'digital transformation' of agreements.


so, as noted. frustrated, any means for anyone to make progress in this
area is very much appreciated; noting lastly, the prior work of unicode to
provide the UDHR in XML https://github.com/unicode-org/udhr

Kind Regards,

Timothy Holborn