- From: Timothy Holborn <timothy.holborn@gmail.com>
- Date: Tue, 26 Sep 2023 07:50:25 +1000
- To: public-humancentricai@w3.org
- Message-ID: <CAM1Sok3RMRzBiLgYs+waYgfp+V74TSMOS3nMKgSVJ0UGu48SBQ@mail.gmail.com>
Please find the content from the email I have forwarded to the ISOC Open
Forum (copy/paste), below...
any questions, feedback, concerns or otherwise; let me know.
NB:
https://lists.w3.org/Archives/Public/public-humancentricai/2023Sep/0003.html
Hi All,
I'm writing to the list to introduce my works and ideas about the future of
Human Centric AI.
Whilst it is my hope that we will successfully bring about this SIG and
through that engage in the discussions and works that provide our sense of
what the requirements for these new types of systems and agents will be, i
thought i would share with you some of my considerations, thoughts and
(draft) works that have been done overtime.
I note that the submission has attracted support from some wonderfully
skilled, talented and highly capable individuals who are very well placed
to help to advance the beginnings of this topic, as the efforts become
coordinated should this opportunity succeed in attracting sufficient
support, as required to bring it about.
I will follow-up with more details about the social aspects of the history
at a later stage, whilst encouraging others to reach-out and become added
to the list!
<https://www.humancentricai.org/>
There is also a historical website: https://www.humancentricai.org/ and
humancentricai.xyz has also been obtained, to support both the W3C &
hopefully also ISOC works. Without support, these works are unlikely to
develop to the level required to bring the outcomes about in a way that
could otherwise be achieved, for humanity.
These works are not new, indeed, i have made a significant effort to bring
about the tooling to support the creation of these agents, over a very long
period of time..
Quick History:
Knowledge Banking Doc 2012 (A Summary from back then)
<https://drive.google.com/file/d/13jjiVON6uodPX3fwoiZbvQ43IjzJnOLc/view?usp=sharing>
(NB:
More expansive document early 2013
<https://drive.google.com/file/d/1EOTzwJmgJhuFl7uvKhlUwmUO6FcQXd0_/view?usp=sharing>
)
Why Fixing "Fake News" is a difficult problem to solve 2017
<https://drive.google.com/file/d/1fvrYlTYUCYQMjW4pTel6QOIvJ03GtQT7/view?usp=sharing>
The Human Centric infosphere 2018
<https://drive.google.com/file/d/1bDwFjfgvS_snBhnH_TUZgFTrfpW3BzIv/view?usp=sharing>
Protecting Citizens & Critical Identifier Attacks 2018
<https://drive.google.com/file/d/1fqaK2eplB-UkQK84_tgjGMX1LvEjTN57/view?usp=sharing>
I spoke at the auIGF in 2014. Then, as I was publically defining the
'Human Centric' concept via the w3c works,
https://lists.w3.org/Archives/Public/public-rww/2016Feb/0015.html
https://lists.w3.org/Archives/Public/public-credentials/2016Feb/0031.html
In 2017 i produced: https://2017.trustfactory.org/ as was part of that
work, i first published: www.youtube.com/watch?v=e9vROTibKiE on christmas
day 2016...
The Credentials
<https://www.w3.org/community/credentials/2014/08/06/call-for-participation-in-credentials-community-group/>(from
the Web Payments <https://www.youtube.com/watch?v=m8cIYzy5MIA> works) led
to the 'digital identity' rollout via other forms of market participants,
globally, during covid... FWIW: I disagreed, noting that there are
different ideologies that exist and it is my opinion that it is quite
important to ensure that there are different options allowed; so long as,
they provide support for human rights & rule of law as well as if not
better than, the alternatives. As far as i can tell, their solutions most
often define people via a wallet with a shared private key; whereas, it is
my strong view that these types of solutions cannot support Human Centric
AI, at least, not in the way i intended to create the tools to support.
I had already been working on this concept of trying to figure out how to
deliver the first 'knowledge banks
<https://drive.google.com/file/d/1MfD4QIGyXioS_liFY66HRwkcrjBAhdtc/view?usp=sharing>'
as i called them, having made some progress locally ('internet australia'
or ISOC-AU) whilst also having prepared documentation for a 'Web Civics'
SIG in 2019
<https://drive.google.com/file/d/1o1FrGelPmWfA6olhKik--UzSBGH1Rz4o/view?usp=sharing>
...
Whilst the notion of needing to support the AI Agents, which is the
interface to the 'data storage' environment (ie: web-payments, credentials
/ verifiable claims, rww - as is now advanced as solid); was known long
ago, i didn't previously lead with this term, as is now the case; but
therein, one of the major threats to successfully achieving the goals of
these works, was that whomever implements the first network for storing
peoples 'data', could be the last. That if a commercial approach to support
human rights was undertaken; then whoever implements the first 'network' if
done commercially, would likely seek to ensure that they're the only
provider available. This was sought to be addressed as a major risk /
threat; which, has been the main underlying reason for seeking to achieve
the implementation of this infrastructure via ISOC...
now, many years later; Human Centric AI SIG running for Election!
The ISOC Human Centric AI SIG (special interest group
<https://www.internetsociety.org/sigs/>) proposal for election is a
significant milestone; whereby you, the members, are now provided the
option to vote on whether or not ISOC should advance these very important
works.
I hope this outline will provide enough information about the many aspects
and aspirational goals that i hope to achieve by progressing works to bring
about Human Centric AI Agents, at large.
The basis upon which Human Centric AI agents can be brought about, requires
the means to ensure that people have a meaningful useful relationship with
the records about their lives. The means to achieve these goals requires
the information to be stored in particular ways; as well as the means to
process many 'social' factors, in addition to the technical aspects.
Whilst there is a very wide-spectrum of 'things to do';
1. The internet today is without 'digital embassies', places where the most
vulnerable are provided the means to retain access to the digital records
of their lives...
Whether the threat of losing all the information about their lives is due
to hardware failure, cyber attacks, poverty, alleged crime or serving time
in prison; or circumstances pertaining to 'alleged' refugees; who need to
prove the merits of their circumstance, before being afforded refugee
status; the means to ensure HUMAN RIGHTS are afforded to all, requires some
sort of 'infrastructure provider' to help to ensure these capacities are
part of the 'digital transformation' outcomes...
In-order for a 'human centric ai' agent (software agent / processor); to
function, it requires the comprehensive use of the information about the
human being who 'owns' it.. the data-subject..
This capability is not merely a 'wallet', rather, it involves a series of
complex requirements in-order to support, at the core of it, 'selfhood' of
the natural person... it's about support for 'personal ontology', which is
itself a very complex topic...
*"personal ontology is concerned with the kinds of things that might
exemplify those properties, such as immaterial souls, human animals,
embodied minds, bundles of perceptions, etc." source: rutgers.edu
<https://rucore.libraries.rutgers.edu/rutgers-lib/48409/>*
The implications are profound.
This capability / that's similar yet different to a 'hosting service'; then
needs to be defined in a manner that is portable, between providers; and
there are many policy areas that need to be explored in-order to deliver
the means to best support this infrastructure.
2. 'Web Civics', 'Permissive Commons', 'Human Rights', Social Web and
Ending Digital Slavery.
These 'Human Centric AI' agents; require 'commons' information structured
in a particular way, as to then compute informatics processes both; in
relation to the aspects pertaining to 'selfhood' (the intimate, private,
'vault' of self); and in-turn also, the social or broader world as does
in-turn relate to 'personhood' as is in philosophy and law, a term that is
closely tied with legal and political concepts of citizenship, equality,
and liberty.
More specifically, whilst 'personhood' infrastructure often now exists for
legal personalities (companies, organisations, incorporated entities and in
some regions, software); the means to support natural persons in their own
private capacity and/or with associative supports for persons for whom
guardianship relates (ie: children, elderly, severe disability; or
temporarily in circumstances such as hospitalisation); does not really
exist...
The internet also lacks the infrastructure needed to support humanitarian
works online, at scale, on fair-terms; which in-turn acts to extinguish our
capacities to produce the infrastructure needed to remedy these incredibly
significant problems.
At an 'informatics' level; the notion of 'permissive commons', refers to an
'uplift' of 'open-data', which has historically been provided in very basic
forms of informatics resources, lacks many features and is largely not
decentralised...
*"The Legal interpretation of the term 'commons' is defined as Something
that is common is owned equally by two or more people, such as a piece of
land. A Tenancy in Common is an interest in land wherein at least two
people share ownership." Source: legal dictionary
<https://legal-dictionary.thefreedictionary.com/common>*
Therein; within this 'social' domain; there are different levels of
'centricity', including 'human centric', 'agent centric' and 'entity
centric' layers that relate to the stewardship of different kinds of
informatics resources. These resources then need to be made available in a
way that supports tamper-evidence, provenance, version control and the
means to make use of those resources locally; by Human Centric AI agents,
that are then involved in producing new records (temporally); that makes
use of those resources via links...
In-effect; the requirements for a 'human centric ai' agent, to function
effectively, locally, on consumer electronics (ie: laptops, desktops, local
home-network devices, etc); requires both the decentralisation of the Giant
Global Graph <https://en.wikipedia.org/wiki/Giant_Global_Graph> as well as
extensive language works.
Therein; whilst it is hoped that this may be performed in a manner that is
supported by 'entity centric' considerations; such as, that the use of a
version of UN human rights instruments that can support the utility of
those terms for agreements between one-another, being supported by the UN
(which has been a historically difficult objective to achieve); somehow,
these informatics resources need to be produced and held in trust or
furnished some form of morally honourable stewardship; and the volume of
works needed, are extensive.
Whilst the 'smart data storage' systems are able to be federated (ie: a
person has an online 'backup' on a 'knowledge bank' // host; as well as a
home server (ie: mac mini); and elements on their devices (ie: laptops /
desktops / mobile phones); akin to a 'personal cloud',
The 'ontologies' that most pervasively exist today are built upon the old
'mainframe' based requirements, in which the vast majority of the tools
define 'consumers' as 'things', not people. the ontologies do not easily
support our humanity, our selfhood, our 'status' as 'observers'; nor our
natural languages...
Whilst there are many misnomers out there about AI and how different forms
of AI science functions; in-order to fabricate 'human centric ai', we need
to create spatio-temporal (space / time) natural language ontologies; that
are formatted with both graph & vector informatics, in a way that can be
employed on consumer electronics with a relatively small memory /
computational footprint.
With the study and computational implementation of natural languages; comes
also, the process through which our history, culture, the history of words
and events become embodied within the structured data-resources; that are
then called upon, contextually, in relation to our activities both 'of
ourselves', and indeed also socially; as to form 'sense-making' apparatus...
upon which, all else employs to produce the informatics requirements
needed; to support the way these 'human centric ai' agents are made able to
function...
including the means through which 'values credentials' or instruments that
are able to be used when forming relationships electronically with
one-another online, such as to include but not be limited to human rights
based instruments; codes of conduct, laws, etc… that are then able to be
appended to reflect the changes in law and/or relationships that occur
throughout life, whilst still maintaining the ability to support the nature
of historical informatics.
The consequential implication of these works is that, we need to figure out
a way through which people are able to be paid fairly for their useful
works & contributions; in a manner that does not lead to rendering
obligations in perpetuity, or some form of 'tax' on our 'thoughtware'...
This in-turn requires the construction of solutions that address these
challenges both technically (ie: appropriate economic mechanisms),
organisationally (stewards, governance) and legally (ie: 'human rights'
software licensing, contracts, etc.).
What this is thought likely to end-up doing; is to lead to a circumstance
where we will be testing a framework to support human rights (ie: fair work
terms, means to address (digital) slavery, exploitation, corruption, etc.);
as we seek to construct the tooling needed, to produce 'human centric ai'
infrastructure; that is then intended, to be equipped to support the
development of 'peace infrastructure projects' world-wide; as required, to
achieve SDGs (sustainable development goals).
3. (Cyber) Safety (Security) implications
So, it is important to note that the objective of these works should be to
provide an OPTION rather than a 'mandate'. There are various alternatives
that have been invested into by various people at levels of great
significance. interoperability & portability is considered to be an
important part of the requirements... However, these systems are different.
The tooling that has been advanced over the decades, considers the use of
Signed RDF resources, decentralised informatics, the use of Domain Names,
DNS records, IPv6 subnets (Security extensions), TLS certs, DLTs, TLS,
OIDC, WebAuthn, etc... The technical methods are intended to support the
needs of legitimate law-enforcement & cyber security requirements (rule of
law / legal requirements, evidentiary systems, etc.); as well as providing
comprehensive protections against 'insider threat' vectors, etc.
Part of the 'human centric' concept; is that, it is not 'governments' who
breach peoples human rights, it is people, human beings - who have a job in
a public institution; who is either carrying out their duties in a lawfully
honourable manner, or is not; and per 'rule of law' the place to decide
disputes where required, is in a court of law, should peace be sought;
rather than corruption or protection under some statute that pertains to
acts of war.
Historically, there has been an unfortunately low level of support for rule
of law; so much so, that it is common that many people are engaged in
unlawful activities, when considered in association to commonplace
activities, that are not better addressed, by law. but if there is no
support for rule of law, then the importance of addressing these issues is
mute.
SO, the informatics environment creates what i call 'social encryption'; in
which, through the moral inclusion of natural persons as a first-class
'internet citizen' (via their 'human centric ai' agent, a 'webizen'); the
cryptographically associative temporal records of what people do, are
spread across the tenants involved in whatever happened; and this in-turn,
becomes an incredibly secure 'social' 'fabric', where the ability to
entirely mute the facts of what happened to anyone involved, becomes a
problem about hacking the internet… or the decision of a judge in a court
of law…
not just one 'node'... or database... Or privateer…. Law.
This is in-turn supported via the 'permissive commons' aspects; whereby,
there are means to temper threats, via law... but that also, the means to
engage in acts of violence and/or manipulate circumstances improperly (ie:
corruption activities) is significantly disaffected.
this is thought likely to challenge many commonly accepted practices, in
many different areas... as the informatics systems, promote an elevation
of our capacities as a human family; from an 'information age', to an
ontologically astute series of fabrics for the promotion of a 'knowledge
age'.. this is in-turn considered essential, should we figure out how to
empower every human being on the planet with the means to find something
useful to do, that may positively benefit others; and that, they can just
go do it...
To achieve SDGs, it is my view, that we want to ensure there is support for
'medicinal earth' outcomes; what this means, is that we need to construct
the knowledge fabrics needed in-order to support the means for people to
rehabilitate the soils, environments and build medicinal villages and
societies; where the knowledge of how to make use of flora, fauna and
funga; when blended with the use of a personal human centric ai agent,
provides the means for people to know how to support their wellbeing via
what they consume; and, how they are then also able to work towards the
construction of outcomes that improve our biosphere, our sociosphere and
the socioeconomic standing and opportunities available to themselves, their
children and more broadly their communities.. our communities...
Whilst these goals invariably require economic recognitions; methods,
systems and processes; the implication also becomes, one of legal
responsibility in relation to the gainful work activities that people do
either online, or in a manner empowered by online systems.
4. Harmoniously 'competitive' environments for the promotion of
humanitarian 'digital transformation' works...
It has been my observation that the vast majority of persons involved in
forming political decisions about the future of artificial intelligence, in
particular, whilst they have highlighted the importance of 'human centric
ai'; seemingly, the lens in which they consider what is best; more
specifically focuses upon the advancement of global 'platforms' and in-turn
also, the desires of governments; which often also, acts to negate the
needs of citizens whose troubles relate to mistreatment by government
employees and contractors.
It appears that technically; there is simply a lack of awareness of the
possibilities, which are often considered therefore 'theoretical' until
such a time, as they have a fully produced embodiment made available to
them as a consumer; to then be empowered, to think about.
so, there's a need to seek out the 'old skool geeks' in our world; to help
produce those RD&D outcomes, whilst is it also CRITICAL that we seek out
supports from professionals in many, many different fields; to work through
the various implications via its many lenses. Therein; a variety of
different approaches have been rapidly developed and whilst i consider
these approaches to be different, the means to ensure that there is
interoperability, portability and ecosystem supports to ensure other
ecosystems can work with human centric ai systems and vice versa; without
impugning the properties of Human Centric AI systems and/or one or any
classes of these forms of systems; in some form of undeclared manner, that
would act detrimental to the underlying moral worth & cyber-security
properties, of these systems.
5. ISOC
It is my strong belief; that the fastest and best available option to
provide this infrastructure is via ISOC chapters.
This is then in-turn able to support the promotion of other 'digital
transformation' works, in a time-frame believed to be far more quickly
brought about, than what may otherwise come about via the UN; particularly
given the apparent 'platform centric' focus of its efforts based upon how
it has been informed of the options. There have been several occasions
where professionals from the political field have told me that they do not
understand the technology and are dependent upon what they are told by
'industry experts', but in this situation; those who are most equipped to
provide the level of professional engagement sought by the public sector,
is in-fact the global platforms who have every interest in seeking to
ensure that they are THE critical REQUIREMENT for the future of
THOUGHTWARE. As all legal personalities (legal personhood) have the right
not to self-incriminate (a rule of law principle); there has been an
absence of representation for the needs of human beings, and this situation
must be turned around somehow...
The productivity losses alone, warrant the investment needed to deliver the
outcomes we need to produce a 'trust factory' to deliver 'human centric ai'
that can then be trusted as a form of technological infrastructure to
support our thoughtware... our human agency, life.
This is not intended to mean - that there should be no other options;
indeed, if that were the case then the meaning of 'human centric ai' would
likely become THE TOPIC that people seek to fight over, to define, as to
maximise profits; and we, will likely be consumed.
There are many questions that remain outstanding, that are not simply about
the 'technical standards' although works in those areas is also critical;
rather, there are various notions of 'platform providers' but in these
frameworks, people are in-effect their own 'platform provider' in many
ways; and whilst there are many benefits that are envisaged possible (ie:
paying people for useful work via micropayments and/or otherwise (therefore
also taxation), local jurisdictional supports (therefore improved access to
justice); as these systems have not existed previously, the ramifications
of how they are best able to be supported to promote human rights (inc.
rule of law); is yet to be better understood, as is necessary, prior to
encoding rules into 'ontologies'. Given the differences, there have been
suggestions that this is in-turn some form of 'new internet', or similar...
some means that seeks to distinguish how these systems operate as to
support law; as to ensure that any measures taken to address the problems
with 'platforms' do not act to lock-humanity-out, of being furnished the
tooling needed for human rights, human dignity, peace via these forms of
Human Centric AI Ecosystems. In some ways, it is not unlike the advent of
Linux, as is now instrumental.
Most people who have had an instrumental historical role with the creation
and development of the internet, have had an involvement at some stage with
Internet Society. Whilst some may not still be involved, it is hoped that
the right type of purposeful objective may well lead to many becoming more
involved, once again...
It is my view; that whilst the challenges are significant, the best known
opportunity to address these challenges is via uplifting the infrastructure
that is; Internet Society... In-turn, there are many 'social issues' that
need to be addressed, which is thought to be part of the 'wheel-house' of
ISOC in relation to internet governance infrastructure; and relationships
had or held historically, internationally.
ROLE WITH W3C
W3C is the historical venue for establishing 'web standards' which is
loosely, both the 'world wide web' as well as the 'web of data', or the
underlying tooling that has been defined over decades, to power software
agents (ai).
W3C does not historically get involved in works pertaining to the 'social
factors'. rather, W3C is the place to support the development of 'web
standards' that are provided support via the patent pools of its members.
Whereas Internet Society and its chapters do Actively engage in 'internet
policy' related matters. This has historically been associated with
regional IGF events throughout the world, which then leads to the
international IGF that has been a critical means through which the UN / ITU
has in-turn been informed by the global internet community at large.
Whilst these frameworks are seemingly a topic of discussion via the UN
Digital Compact (digital transformation agenda, etc.); i am not aware of
any other existing capabilities, that may be extended to support
humanitarian 'knowledge banks' for those most in need; and, as a means to
progress 'web civics' works, on peace infrastructure projects whilst
looking to provide support for human rights via this approach that is
squarely focused upon the practice of doing so via 'Human Centric AI'
systems.
One of the finer points herein; is that one of the long-term challenges has
been, to seek to ensure that whatever the first implementation is, does not
become the last implementation; but rather, a means to support the stimulus
required to support what some have called a 'human centric internet', but
is moreover about the means to make use of technology as tools; to promote,
supports for human rights; particularly, for those who most require the
useful employment of those rights, in a manner that best supports peace.
Yet moreover, whilst the 'social factors' are more about the works that
'should' be done via ISOC (not w3c); the outcomes are then likely to inform
what works could or should be progressed, via the W3C Human Centric AI
<https://www.w3.org/community/humancentricai/> group, set-up earlier this
year... Importantly also, the means to figure out how to fund the work of
people doing humanitarian work, in a manner whereby humanity is the
intended beneficiary; is significantly uplifted should these ISOC works
become successful; and in-turn, equipped by supporters.
6. Values Credentials
I have started, today, writing some AI compatible versions of the ISOC Code
of Conduct and Mission Statement. I am yet to seek to address the privacy
policy…
https://github.com/WebCivics/isoc-hc-ai-sig-prep/tree/main/ontology
This is the beginnings of an example of how we are able to create online
resources that provide declarative semantics, that then provide the means
to support the use of Human Centric AI agents. These processes require a
great deal of work, to ensure supports of Human Rights; and address the
many difficult use-cases that relate to the procedural requirements
associated with the dynamics of multi-agent systems; and how to seek to
ensure optimisation in a way that can better ensure systems promote
supports that ensure decisions are correct in the eyes of history not
simply otherwise, as assumed at the time a claim was made. These
considerations have many implications, that in-turn need to be worked
through as to support human rights based outcomes.
See Also Draft Example of UDHR Values Credentials
<https://docs.google.com/presentation/d/1q6Cl__FbUddtKtBPxw2KDviO_PEl7Ti-XYZOT1EQM14/edit#slide=id.p>.
(more notes below
7. Technical & Related Notes
Whilst i have been working on a particular form of ecosystem, to support
private and personal AI agents that operate in a particular way; socially,
this is not something that i seek to define personally; rather, the moral
and honourable practice method, is seeking to engage with others and figure
out what we, as a community whom are involved, agree upon; after,
discovery, discussion, processing the options; Nonetheless,
It is my belief that the best available option for producing a 'backwards
compatibility spec'; which is basically about defining a set of
requirements, that means systems are required to be able to export to those
defined specifications; as to ensure users ("owners") can be migrated
between systems whilst maintaining the functional use of those systems even
if they've been migrated elsewhere, or that new rules have been applied
(ie: due to court orders); is most likely to be premised upon W3C
Standards works, notwithstanding some known issues with respect to the
documentation of these works; Therein, it is important to consider that not
all application of technical standards are expressly defined in the
documentation of those standards...
Some of the older resources;
- https://web.archive.org/web/20181116103202/https://crosscloud.org/
- https://drive.google.com/drive/folders/1lpeoEFowRcq3VTAp5LH6cFN251O9g9iE
- Example of decentralised 'micro-blogging' app that used RWW servers:
https://www.youtube.com/watch?v=IhwAiTOFPrc
- TimBL demonstrating 'solid' on his laptop (thought likely to be developed
over decades?) https://www.youtube.com/watch?v=gZB6d-4klmU
The libraries now support verifiable claims / credentials, etc.
There are several implementations, ranging from 'personal' solutions
through to enterprise (ie: BBC) and government (IE: flanders, belgium)
-
https://pdsinterop.org/
-
https://solidproject.org/users/get-a-pod
-
If you would like to test one out; see,
-
https://github.com/pdsinterop
-
https://github.com/nodeSolidServer/node-solid-server
-
Related works: https://github.com/solid/
There are a number of differences in relation to the approach; that is
thought important for support of 'human centric ai' in the form, that i
have been working towards... In-part also, this relates to the use of
domain names and consequently therefore also, IPv6.
Herein; whilst Solid may not end-up being the solution for Human Centric
AI, it is considered to be an option for commencing the process of
constructing the decentralised informatics systems, that are then able to
be uplifted overtime to better support Human Centric AI Agents & related
systems. The way these systems work is in a decentralised manner.
Diagram: interface Example (NB also: from ~2015 http://dev.webcivics.org/ )
The image below illustrates some dev work that was done earlier this year,
where a web-extension provides the capacity for people to 'log in' to their
Human Centric AI (IdP via a Solid Server); and then launch 'apps'
(html/css/js) locally in the browser; the implication being, that the
'social web' is baked into the browser (rather, underlying data-service,
per image below); this then means that people can interact with any URI
(web-page); and that their 'social' environment is defined by
themselves... not platforms... Solid can support implementations now, that
support the ability to start to get a comprehension of how these systems
work; and therefore also, what are the many implications that may need to
be addressed via ISOC works.
Work is presently being done to investigate how this may in-turn be
integrated with a blockchain solution for micropayments / tokens /
financial & non-financial currencies (which requires a 'vault' as is
different to a 'wallet' as 'identity' is defined via the 'inforg
<https://medium.com/webcivics/the-semantic-inforg-the-human-centric-web-reality-check-tech-50e2fa124ed4>
')
It is important to note; that the underlying 'data services' that provide
support for functions such as 'social-relationships' with other 'entities'
are not managed by the 'app', but rather the underlying data-service… This
means that the context in which people manage the information in
association to their relationships with others is in-turn - human centric;
the 'social graph' or complex 'address book' / permissive
commons-structures work in 'any app'...
Illustrative Diagram: Systems
The below diagram was also produced earlier this year; as i was working
through the process of figuring out how to define the 'natural language
ontologies' and the systems needed to support it; that then provide the
means for people to use natural language, to define how they seek to
categorise relationships, for 'connections' with one-another and various
other functions...
In this diagram; the term 'web 3.0' refers to the sort of definitions made
earlier: web.archive.org/web/20130128161834/http://jeffsayre.com/...
<https://web.archive.org/web/20130128161834/http://jeffsayre.com/2010/09/13/web-3-0-powering-startups-to-become-smartups/>
but that, not all functions can be supported either in the 'web browser' or
via an extension; so there needs to be an underlying 'agent' (the human
centric ai - agent); that supports the functionality required, to process
information locally, act as an API interface for other forms of AI packages
(ie: transformer models, etc); and also, as a local 'server' that is then
able to support the decentralised functionalities, etc...
But herein are various considerations relating to the 'human centric'
nature of the informatics design, whereby the means to support
relationships without platform intermediaries is supported; but this
in-turn leads to the need to consider various implications & the need for
'safety protocols'...
These works are also being advanced to consider how addresses may be
applied to 'things' (as distinct to people); as the means to figure out how
to support humanitarian (tech) work online in a manner that promotes human
rights, is considered to be a top priority; without seeking to make 'poor
choices'... Therein, whilst there is personal cryptography involved; it is
more of a vault, than a wallet.
NB: some more unfinished earlier work, illustrating in-part, some of the
future objectives... webcivics.github.io/CooperativeProjects/projects.html
Earlier this year, i also set-up; https://www.humancentricai.org/ and
www.w3.org/community/humancentricai and acquired the 'humancentricai.xyz'
domain for use with standards communities; whereby the future hope, is that
sufficient progress is made as to bring about the opportunity to migrate
stewardship of 'web civics' and related efforts to any such future
safe-harbour as to improve the means through which these long-term goals,
can be progressed and hopefully therefore also; realised…
The implementation I have been working on employs domain names, DNS records
& IPv6 (security extensions, etc.); and, it has a 'host' framework to
support pseudo-anonymised 'informatics'; but that largely, the means in
which the AI agent operates, is local for people who have a laptop or
similar... This has then sought to be developed further with 'group'
systems (ie: places of work, schools, etc.) that leverage off the existence
of underlying 'human centric ai' supports provided to individuals.
Then through the use of domain names as the primary means to form linkable
identifiers (inc. use of DNS records); relationships are defined between
people (/agents) in a manner that employs 'value credentials' that denote
the shared 'values' that are expected to be supported by one-another in
relation to that relationship; whereby, the terms and nature of
relationships change overtime, and the artefacts end-up being supported via
the 'permissive commons' frameworks; with legal support.
FINAL NOTES
Operationally, that is, in relation to the 'modal' structure of how i've
been doing these works; i've observed many significant differences between
the process in which i have sought to prosecute the development of
technology to support human rights in this particular sort of way; as an
individual, that has seemingly not been possible for the vast majority of
others, whose works are facilitated through the lens of their role as an
agent for an employer; sought, to represent the interests of their employer
first & foremost.. It is my opinion that this dutiful methodology has had
many very negative repercussions.
Somehow, people need to be able to declare the requirements that are most
important to them; because, they are human... and without that humanity,
being furnished the capacities needed to be supported as to grow
peacefully, towards one's fullest potential:
en.wikipedia.org/wiki/Maslow%27s_hierarchy_of_needs
It seems to me that there's not much good purpose to go create 'human
centric ai' supporting systems; if, the objectives that are sought by those
who empower the capacity for people to exist, seek to make unconscionable
demands. There are a plurality of very difficult human rights use-cases
that have not been addressed very well in the public forums where web
standards, and similar, are created... there's a lot to do... but, the
essence of it, becomes substantively about forging the apparatus through
which we may better understand ourselves; and through us, our world.. both
the space & inter-space… the nature, of who we are as persons & what it is,
we do…
I have provided an Appendix of resources below, that provides some more
information. However, I would like to note that the protocol for making
decisions is hoped to be via group decisions and that the links illustrate
various factors that relates to the ISOC work sought to be done; in
consideration of the complexities, but that these works are not necessarily
ISOC works other than as necessary to support enablement. Group efforts are
sought to be undertaken that may therefore be achieved through discussion
and efforts to seek to ensure complementary good faith engagement on works
seeking to form cooperative outcomes; which may lead to differences that
could, for instance, be catered for by classifications or other means of
categorisation.
Overtime, it is hoped that the innovation tools to support work online will
develop and that through the use of these tools, our means to make better
efforts more productively will be improved.
If you have any questions, concerns, issues, ideas or suggestions; or if
you're simply excited and want to reach out, please do not hesitate to do
so…
https://www.youtube.com/watch?v=lHB_G_zWTbc
Sincerely and with Kind Regards,
Timothy Charles Holborn.
https://www.linkedin.com/in/ubiquitous/
https://www.humancentricai.org/
http://webdev.intent.eco
APPENDIX
The Peace Infrastructure Projects Library
<https://drive.google.com/drive/folders/1tYFIggw8MIY5fD2u-nbwFRM6wqrhdmQZ>
This is currently presented as a google drive folder, whilst it is hoped to
create a decentralised 'semantic library' in future. The Library is
constantly updated, contributions are welcomed and in particular it is
encouraged that others form a similar folder; and then provide a link to
that folder which can in-turn be added to one-anothers libraries… The
source of this content is primarily via linkedin posts; however, when i
perform research online or are forwarded documents, i also add them to this
library. The categorisation is difficult, as many documents relate to
multiple topics; and the means to more easily make use of; and make
(social) annotations is not well supported presently. Note also, that i am
influenced by my language skills (english) and place (Australia). It is
understood that there are different resources available to people with
different skills to my own, and it is desirable that the means to uplift
our ability to share knowledge is brought about via Human Centric AI
Systems.
Within this folder, the Tech Ethics Folder
<https://drive.google.com/drive/folders/1uwGax8GvZA2jzJ_UFIoYppijZX4vDsoL>
contains
various publications about ethics and human rights. The DevOps
<https://drive.google.com/drive/folders/1C-_jdQwtWxyjftrvI0yUSEKfhzeaKTV0>
folder
contains various technical resources, and in particular therein, the
DataScience
<https://drive.google.com/drive/folders/18kpYSPz4yQMMn0IOfaPT7JyYD1wXEGMn>
folder
contains a large amount of information about various kinds of AI Sciences,
whilst other miscellaneous AI publications are in the AI MISC
<https://drive.google.com/drive/folders/1tFYQBE3i0hyrsBCE_33TiLHtGA7XLBtC>
folder..
The Blockchain & Web3
<https://drive.google.com/drive/folders/1L-0fhDBLZJW-7OUxrdkUPmX3F9zQDOvx>
folder
contains files about that, alongside SSI (self sovereign identity); which
is considered to be a different method for defining human identity, than is
achieved via these Human Centric AI related approaches.
The Gov & Defence related docs
<https://drive.google.com/drive/folders/1RzE_QQbBRyE0druVpvffQoT5KDms_5mT>
folder
contains files and folders for regions, multi-governmental groups and Cyber
<https://drive.google.com/drive/folders/1BVrg_6ss0v0RlnahjaM26eG40fvRfEZr> from
a policy related intelligence point of view, as is supplemented by PsyOps
<https://drive.google.com/drive/folders/1Gsi4pjnIcjSdhyco3RI6j7py2qm4g5t5>.
The ID - IoT - GIS - ARBR (metaverse)
<https://drive.google.com/drive/folders/11E7MvuGkM54RD5CeJvsG81ECS5H87UKS>
folder
contains files about 'things' and larger scale considerations such as
industry 4.0, smart cities and mobility as a service, 'metaverse', etc..
The Selfhood
<https://drive.google.com/drive/folders/1qtXrWitII3X_5fnbU0pZRikCrLMs21I->
folder
seeks to collate information that pertains to human lives; education,
health, psychology, consciousness studies and knowledge representation; and
other social factors; and the QM
<https://drive.google.com/drive/folders/1GQ-h6Ym69xMJxZSSMBjARwKxnKKqNqAu>
folder
(quantum) collates some of the information about quantum physics & related
applications.
It is thought probable that quantum physics has something to do with the
function of 'mind' or 'consciousness'; however, the mechanism is unknown.
There is years of research into the field of 'consciousness studies'... I
have a YouTube Playlist
<https://www.youtube.com/playlist?list=PLCbmz0VSZ_voTpRK9-o5RksERak4kOL40> as
well as a Timeline
<https://cdn.knightlab.com/libs/timeline3/latest/embed/index.html?source=1r-bo83ImIEjSCmOFFMcT7F79OnCHDOGdkC_g9bOVFZg&font=Default&lang=en&hash_bookmark=true&initial_zoom=4&height=750#event-consciousness-qm-ai-studies-video-edition>
Presentation
that helps to provide further insights into the progression of works and
theories in this complex field.
NOTE: About 'language models', until recently a large language model was
something like Babelnet cyc FrameNet Framester Lemon linguist list SUMO
WordNet, etc. However more recently, due to the emergence of transformer
models the term 'large language model' is now far less clear.
Consequently, the resources about language models is now in multiple places
in the library. The work to create a solution for spatio-temporal
n-dimensional storage and use of resources such as language models, is
described in the Spatio-Temporal: Human Centric AI Package Format
<https://docs.google.com/document/d/1Fwx3-YYyKgeigaScoMVoTFc3V2p-0jVwOg0IvMr8TZs/edit>
document.
METHODS STUDY: The method seeks to provide practice examples of how
decentralised knowledge sharing could or should work. This then supports
the means to better understand future requirements for systems to support a
'hyper media library' and related works.
Some Examples include;
-
Ted Nelson - Xanadu Space: https://www.youtube.com/watch?v=En_2T7KH6RA
-
British Museum - ResearchSpace -
https://www.youtube.com/watch?v=MaAv0SE7wis
Youtube Libraries (video libraries)
*Medicinal Earth*
The Medicinal Earth concept is an upper-level 'objective' concept that I
have started to use to inform the approach that I am seeking to bring
about, particularly given the declared objective of WHO to address 'digital
identity' as a matter of great importance for human health (and mental
wealth)..
The concept is about the idea that we use ICT / AI in a way that promotes
our ability to understand how to create medicinal environments and in-turn
through the use of knowledge, learn how to use those medicinals in our
daily lives to support our health and wellbeing; both individually, and
in-turn also moreover socially.
https://www.youtube.com/playlist?list=PLCbmz0VSZ_vpr1Hok5YfMyAFlKcz0eUxx
I have also started on a list for fungi or funga, which was previously not
well known by me.
https://www.youtube.com/playlist?list=PLCbmz0VSZ_vo4aRHKPha7jGOeB_22hCWL
This in-turn feeds into the broader body of knowledge sought to be
developed around,
Biosphere Management;
https://www.youtube.com/playlist?list=PLCbmz0VSZ_vovZZfPYe136pRjE6QotzQX
Solving The Plastics Problem
https://www.youtube.com/playlist?list=PLCbmz0VSZ_vrugX0s7t3rdGqyGtkYKDWz
Improving Water Quality
https://www.youtube.com/playlist?list=PLCbmz0VSZ_vpwHpGRTOKI47pPiWrdDQla
Housing
https://www.youtube.com/playlist?list=PLCbmz0VSZ_vob_3UTGlns9smlCYEWbEP5
https://www.youtube.com/playlist?list=PLCbmz0VSZ_voRLoOc1QIrb8pAXDjG4S61
Energy
https://www.youtube.com/playlist?list=PLCbmz0VSZ_voj61kp4n0hyTfj1folDGBz
In-effect, the useful outcome sought to be brought about by Human Centric
AI systems and related supports; is to figure out how to address the
real-world issues in our biosphere / sociosphere, that leads to negative
consequences that act to inhibit healthy people and healthy environments
that are needed to provide the supports required to deliver those health &
mental wealth fundamentals. This means that we need complex knowledge
informatics systems; that are then able to be employed personally, with the
use of personal information to provide context in relation to aspects of
'selfhood'. It is commonly the case that there are low-cost means to
address many problems, but the barrier is whether or not the knowledge is
available to identify and implement solutions to any such problems.
*Consciousness*
The studies into 'consciousness' (as afore mentioned) are applied in terms
of seeking to attenuate the production of technology (AI / knowledge
informatics sciences) that does not act to unduly interfere with natural
agents. This is increasingly now a field of study, by leading AI experts.
https://www.youtube.com/playlist?list=PLCbmz0VSZ_voTpRK9-o5RksERak4kOL40
Human Agency & Digital Identity
https://www.youtube.com/playlist?list=PLCbmz0VSZ_vp7r7VvuGuaSipjxp9bI1_D
Science & Spirituality
https://www.youtube.com/playlist?list=PLRtuDWE10sVcfZmHqqH2qXPY0I1eHch7j
Quantum
https://www.youtube.com/playlist?list=PLCbmz0VSZ_vod7oKdLBUCEZkH0wCaFh8_
*(note; fairly sure quantum systems, prefer vectors...)*
*Economic*
See Draft Doc: Knowledge Economics: Project
<https://docs.google.com/document/d/1OBktQmBSt-Crq4r1A3PLJWEjMvQ0uBRo780ySo1BOrg/edit#heading=h.z6ne0og04bp5>
Knowledge Economics: Project
<https://docs.google.com/document/d/1OBktQmBSt-Crq4r1A3PLJWEjMvQ0uBRo780ySo1BOrg/edit?usp=sharing>
I have added the 'human rights' aspects into this economic section, as
human rights are either economically supported or conversely the abuse of
human rights for profit / gainful purpose, is materially supported…
Thereafter, are a range of considerations highlighted via various lists.
Human Rights
https://www.youtube.com/playlist?list=PLCbmz0VSZ_vrnPSQ41DnH0FpY1f9fgAR2
Knowledge Capital
https://www.youtube.com/playlist?list=PLCbmz0VSZ_vqrjn5j3Ldedw-NWUFKNhIW
Economy
https://www.youtube.com/playlist?list=PLCbmz0VSZ_vpYsIEl_gb9BUwFnp7EHh08
BOT models
https://www.youtube.com/playlist?list=PLCbmz0VSZ_vpjjpCi4DfO49p_3-QqwLuX
Mission Model Canvas
https://www.youtube.com/playlist?list=PLjDPolp9BAOHy3WBfBIULusFNWBs6amyn
Innovation Markets
https://www.youtube.com/playlist?list=PLCbmz0VSZ_vplID3vlHvdvN6k3ozI94im
Coops
https://www.youtube.com/playlist?list=PLCbmz0VSZ_vrkH-HlXct_Z3HKeZF6bB00
Digital Receipts (note; note 'ontologically' defined in more useful ways,
yet)
https://www.youtube.com/playlist?list=PLCbmz0VSZ_vrg0f9Id-s2qoDJOs56YdIB
eCash (a candidate blockchain method)
https://www.youtube.com/playlist?list=PLCbmz0VSZ_vp0GJ1WgR2b9fvI89n0m9cn
Pervasive Surveillance
https://www.youtube.com/playlist?list=PLCbmz0VSZ_vqLxdLjyFae7ozEiAIKJw0Q
Attention Economy
https://www.youtube.com/playlist?list=PLCbmz0VSZ_vp9SeEKiGnvM7NdxoZvaorN
TECH:
These tech resources are mostly fairly basic… but they may help get a
grasp of some of the more basic concepts. Note that the previously
mentioned lists about Consciousness also have tech in it.
AI
https://www.youtube.com/playlist?list=PLCbmz0VSZ_vr5buCx-9K6DFmcs_0BtxRC
Ethics & governance of AI (MIT)
https://www.youtube.com/playlist?list=PLj62-wQeg_DjuoWS4A_VYzLWtEVUZj1IO
IoT
https://www.youtube.com/playlist?list=PLCbmz0VSZ_vpewcqcz-XNu4NV3tCAl6Py
New Internet
https://www.youtube.com/playlist?list=PLCbmz0VSZ_voyWh284YNHT44344S4-DU_
What is Linked Data?
https://www.youtube.com/playlist?list=PLCbmz0VSZ_vqZemkHpzb7-GWmzcUdGgfP
Semantic Web
https://www.youtube.com/playlist?list=PLCbmz0VSZ_vqWAb5NHkaEvsT9RA8--QkY
Semantic Reasoning
https://www.youtube.com/playlist?list=PLCbmz0VSZ_vqWAb5NHkaEvsT9RA8--QkY
Virtuoso RDF Views over RDBMS Data Sources
https://www.youtube.com/playlist?list=PLCbmz0VSZ_vqsYV4B_oXulV3ncJbFj0Hb
Ontology Tools (note; mostly via OWL presently)
https://www.youtube.com/playlist?list=PLCbmz0VSZ_vr4BRBxfGhAPXboA9iAdZlQ
RWW (solid), Credentials & Payments
https://www.youtube.com/playlist?list=PLCbmz0VSZ_vqI-SfZ3VRo8bs6jJF-Dmrs
NOTE: I have started to work on a 'RWW / Solid' based video library; but
have not finished it. https://github.com/mediaprophet/gitpages-video-project
(Basically, all it needs is the JSON file…)
After WSIS2023; Some work was done on seeking to figure out how to support
the means to define a solution for SDG Learning & Knowledge Hubs
<https://docs.google.com/document/d/1D63FlICIOXcnLx_PYs0ByLYvc1B-BXqhhXGdaW_rIMI/edit>
SDG Learning & Knowledge Hubs: ProjectConcept
<https://docs.google.com/document/d/1D63FlICIOXcnLx_PYs0ByLYvc1B-BXqhhXGdaW_rIMI/edit>
"
The energy calculations need more work; as has, the work on figuring out
how to define both human centric ai (for digital identity), and the
'medicinal earth' concept; as that relates to what might be best taught to
people for useful purposes.
Therein, there is both a 'recognition of prior learning' and ability to
provide micro-credential 'wrappers' around educational content to test &
quality what people have learned…
This is in-turn something that is expected to enhance productivity; but
that also, feeds into the Human Centric AI - underlying series of
requirements; to better understand what people know, have studied and have
gained skills, knowledge and capacities to support various kinds of work
activities.
seeAlso notes about Annotations - Knowledge Clouds
<https://docs.google.com/document/d/16R9CqN18O3v358jC5GpqJ59Vlw7-7HjxGpY_BE2eA-k/edit>
Annotations - Knowledge Couds WIP
<https://docs.google.com/document/d/16R9CqN18O3v358jC5GpqJ59Vlw7-7HjxGpY_BE2eA-k/edit>
…
https://github.com/orgs/WebCivics/repositories
Webizen AI Repos
https://github.com/orgs/WebizenAI/repositories
The Webizen AI repositories have some early work defining or working to
integrate some host systems, as well as forming the basis to the client.
In this method, that was done prior to the events that led to renewed
'standards' / internet governance related works; the host seeks to use a
containerised approach for managing end-user 'hosts' which are then
facilitated via tailscale (wireguard) open-source libraries; that are then
supported by a variant of WebID-TLS & WebID-DNS (that afaik, doesn't
actually exist otherwise); the main differences are the ontological methods
sought to be used, that required testing prior to any work to seek to
standardise in some way.
This then provided a RWW (early solid) based client on the host-machine;
which could then be employed via a web-browser extension, to support
decentralised 'social' apps; whilst, the host environment provided support
for payments, back-ups, continual presence & 'permissive commons' supports;
alongside, IPv6 subnets and routing in connection to a users own domain
name (TLD); which then led to modifications of the DNS supports provided
via the wireguard based last-mile networking solution…
The issue that became important to address; was how to support natural
language ontologies, which are thereby required in-order to support
implementation. Whilst I had plans to create a commercial implementation;
the problem became that I did not want to seek VC investment to do works
that should be 'commons', such as is the case with human languages…
Additionally; some initial work was done on seeking to advance a DID for
GIT, incorporating support for SSH… This has not been resourced either,
noting that there is an express difference between GIT the subversion
control protocol that is part of linux and available on almost any system;
vs. github, which is a centralised product / platform…
The other issue was in seeking to establish 'values credentials'; built
upon the use of natural language supported - human rights instruments and
similar… alongside ontologies for 'selfhood'...
https://github.com/WebCivics/ontologies
UDHR Values Credentials
<https://docs.google.com/presentation/d/1q6Cl__FbUddtKtBPxw2KDviO_PEl7Ti-XYZOT1EQM14/edit#slide=id.p>
UDHR - Inter-Personal 'Value Credentials' Agreements
<https://docs.google.com/presentation/d/1q6Cl__FbUddtKtBPxw2KDviO_PEl7Ti-XYZOT1EQM14/edit#slide=id.p>
UHR Blockchain & Economics
<https://docs.google.com/document/d/1zirrCE-nP9mdFicdWSvEVTQjjG5Asb8q5RBd2LlJkOw/edit#>
UHR Blockchain
<https://docs.google.com/document/d/1zirrCE-nP9mdFicdWSvEVTQjjG5Asb8q5RBd2LlJkOw/edit>
After some years trying to figure out how to get this done (note: UDHR in
XML by Unicode <https://github.com/unicode-org/udhr>)
I then attended WSIS2023 and this led to UN GDC involvement. This is a
draft document (incomplete) that relates to my considerations
<https://docs.google.com/document/d/1xetF-5dhyU-iuLV-Trn2P-m_BHbeK-N-6e0pOcbsx40/edit#>
; 20230430_UNGDC InspiredDoc - DRAFT
<https://docs.google.com/document/d/1xetF-5dhyU-iuLV-Trn2P-m_BHbeK-N-6e0pOcbsx40/edit?usp=sharing>
ISOC 'values credential' (code of conduct) draft in RDF:
https://github.com/WebCivics/isoc-hc-ai-sig-prep
Note also, exhaustive amount of work required to illustrate or demonstrate
considerations about the distinct values, implications, requirements, SWOT,
etc.. alongside regionally contextual issues.
There are many different kinds of 'artificial minds' or AI agents, indeed here
is a list of some of them
<https://docs.google.com/spreadsheets/d/1VixKXjZL31bZRXQS9J1FmvPyDzdkgE8B2-3fzPmRYNc/edit#gid=1227058217>
.
But the question becomes, which sorts of ones, what sorts of
characteristics are most important to you. What sorts of characteristics
are most important for your children, your family, your community. This is
an opportunity to work through all those sorts of considerations; and then,
define the socio-technical tooling required to bring about a different kind
of option, Human Centric AI…
If you are interested in this topic, if you have any questions; or if
you're interested in getting involved in the work being done RIGHT NOW -
Please, feel welcome to reach out... There's alot that still needs to be
done. I am quite sure i missed at least a few things in this document, but
will be working to put up some web-resources; whilst not seeking to be the
person who defines it all.. I hope, we are able to forge a sustainable
means to engage in humanitarian works for the betterment of humanity,
globally.
Tim.
------------------------------
timothy.holborn@gmail.com
www.linkedin.com/in/ubiquitous
Received on Monday, 25 September 2023 21:51:14 UTC