W3C home > Mailing lists > Public > public-solid@w3.org > March 2019

Re: The Default

From: Melvin Carvalho <melvincarvalho@gmail.com>
Date: Thu, 21 Mar 2019 11:27:22 +0100
Message-ID: <CAKaEYh+pmewf=byUNDDQy7LpxfrAb3ehmnX0We4jrj4yHbYEBA@mail.gmail.com>
To: Mitzi László <mitzil@inrupt.com>
Cc: Timothy Holborn <timothy.holborn@gmail.com>, public-solid <public-solid@w3.org>
On Thu, 21 Mar 2019 at 11:22, Mitzi László <mitzil@inrupt.com> wrote:

> Hi Timo,
> Thank you doing a re-take and sharing ideas to build on.

Could you outline the purpose of this doc.

Are these personal thoughts, or designed to be part of some "official"
discourse, or our group.

> Mitzi
> On 16 Mar 2019, at 23:35, Timothy Holborn <timothy.holborn@gmail.com>
> wrote:
> Heya,
> I realised i did not respond in a manner that was supportive of your
> initial attempts.  I've thereafter put your doc into a google Doc[1] and
> have started to make comments (edits, in comment mode) - noting, i was
> unable to transfer ownership which was my intention (google told me, it
> needed to be transferred from the same domain - so, i couldn't do it).
> The first question i have - is it your intention to produce this document
> as a w3c Solid CG document? noting the particulars with respect to the IPR
> management implicit with W3C CG (and later IG/WG) works.  If not, then i'm
> happy to start a new document - whilst assuming otherwise that this must
> have been your intent
> Second point is that specs usually go into a standard type of template,
> here's an example[2].  I note therein - there's a bunch of related specs
> (ie: LDP, SPARQL, RDF, HTML/HTML5, HTTP, Verifiable Claims (?), WebID,
> WebID-TLS, WebID-OIDC, LDN, WebSockets, Web of Things, WebRTC, various
> ontologies - ie: FOAF & the list goes on...) that could be culminated into
> a document that helps to provide pointers to the ecosystem components.
> Therein, the other method used for documentation is Wiki pages.  I was
> alerted to a wiki form that's solid compatible[3][4] whilst noting, the
> 'official W3 solid CG' wiki is here[5].
> Thereafter - here's a bunch of Thoughts / notes... IMHO.
> 1. W3C SPECs vs. Ideological frameworks
> This is always problematic.  There are various motivators for various
> persons, who are involved in 'standards work' via various means - some,
> investment backed (with not enough revenue for sustainability) where
> investors have different ideas / needs; others, 'starving artists', others
> - corporates with incorporated agenda, etc.
> W3C Specs from the CG need members buy-in to work.  some examples includes
> history with payments works[6][7] which are now instrumental constituent
> objects to 'solid', afaik.
> Specs do need objective parts; yet, imho - enabling a variety of
> implementation methods is more considerate of the needs of 'standards'
> rather than the means to use a standards process to cement a 'rent seeking
> position', that isn't necessarily going to work out for the document
> authors / editors in any case.
> Which brings me to point (2)
> 2. 'human being centric' or 'human centric' vs. 'person centred' or 'agent
> centric'
> The solid specs, afaik, essentially provide flexibility about
> implementation strategies.  These implementation strategies will be
> employed via a specified ideological approach by spec implementers - but
> this is not necessarily the same across the ecological environment brought
> about by the spec's existence - which does in-turn, seek to support
> interoperability rather than ideological application of a certain view upon
> all.
> Therein - there is the thing i call 'human centric'[8] requires a bunch of
> apparatus that is external to W3C, and now seemingly underway.
> Alternatives may include 'person centric'  which then brings about
> implications about companies being considered by text of law persons[9]
> (which then, brought about a means to make a distinction about natural
> persons by works relating to the concept of natural persons being
> 'consumers'[10]).
> Agent centric extends beyond the 'person centric' stuff - and starts to
> add software, such as AI agents - whether they be the 'things' 'driving'
> 'self driving vehicles', or otherwise.  These sorts of considerations are
> technically required, no-matter how the accountability frameworks are
> designed to bring accountability in defence of victims.
> Some may prefer a world where if the AI agent does the wrong thing, it's
> considered in a manner similar to 'act of god' and nothing is done;
> others, are very frightened about what the implications of this traditional
> type of 'corporate view' may have upon humanity and the natural world.
> yet - the problem is, when defining w3c specifications for 'solid' - the
> solid platform may well be a good option for being employed as the
> underlying 'web operating system' or 'network based operating system'
> foundation for a self-driving vehicle (whether it be car or wheelchair)
> that may in-turn communicate with other agents to ensure the occupant isn't
> harmed (particularly good for wheelchairs, seeking to ensure someone
> doesn't come-out of their driveway and run-over the wheelchair, for
> instance); therein - this is an 'agent centric' approach (using the vocab
> definition used by FOAF).
> 3.  Distinction between a WebID and a WebID-[AUTH-SEQUENCE] - enormously
> important stuff.  There is also 'patterning' happening with the creds work,
> that needs to be addressed by way of implementations, which IMHO in-turn
> requires definition of the 'intelligent agent' thing - that i think is
> intended to exist within the 'solid specs'?
> 4. DYO (Define Your Own) Robot
> The robot needs ontologies.  If the robot is controlled by a company (or
> old-world-robot) then this needs to be declared as to make a master/slave
> arrangement between the responsible actor, and its subordinates, which
> in-turn need to be addressable for other agents.
> 5. SemWeb Addressable URIs (Inc. DIDs?)
> This is rather kinda important, imho.  A definition needs to exist.  I'm
> not sure if its about HTTP, or SPARQL Addressability, etc.  I've looked
> into various related links[11][12][13]; whilst not knowing / understanding
> what the 'standards related' interop spec - should look like, at this stage.
> 6. pseudo-anon WebID's / URIs
> Providers could, in theory, provide a URI string that supplies a WebID &
> related data/support stuff, in a manner that keeps the real identity of the
> URI owner confidential to the provider (which is therefore able to be
> subjected to lawful request, facilitate KYC/AML, etc.);  I do not see these
> semantics declared in almost any of the solid related works (perhaps its an
> 'at this stage' type of problem?).
> 7. Final thoughts on document
> IMHO - this document needs to be broken down into constituent objects.
> there are many issues with it ATM, so far as i consider, and some of the
> resolutions do indeed need new work to be done.  embracing the leadership
> role of inrupt is essential to doing so, noting that whilst it is amongst
> potential lead implementers (thinking also of openlinksw in particular, but
> not exclusively) - there are some questions in there that are less about
> specs, and more about business models.
> Therein - this is difficult work to do.  There is nothing about payments
> in the document, and the fact is that people only exist due to
> socio-ECONOMIC capacities made available for them to do so, there is no
> world where human beings (homo sapiens = wise man) live with dignity &
> wisdom whilst being devoid in their lives of economic attributions, or that
> they be considered a cost/burden upon society, as a consequence of seeking
> to do good (as apposed to dealing drugs or the many other socially bad
> things to do, that harm people, but have direct economic attributions that
> assist those who make such choices do indeed get paid for doing so.
> The dignity equastion, or kindness equations relating to 'knowledge
> workers' needs to be addressed; by implication therein, the concern nations
> may have is that the purpose of government relates to distribution of funds
> (which means people must be economically attributable, other than as a
> consumer) and law;  so, in-order to make it clear that solid is NOT a plan
> to strip nations of their revenue & systems of evidence that enable their
> 'rule of law' to function (as is indirectly the implication) as to be
> managed worldwide - - > solid needs to demonstrate how it is that its
> systems design 'philosophical engineering' strategies, improve attributable
> personhood whilst also attending to emerging issues in IoT, AR, AI, etc.
> I think it is the case that solid is the best placed project able to
> demonstrate how it is designed from the ground-up in a manner that is able
> to meet these various challenges, but therein are the pragmatics of
> invest-ability which in-turn relates to ontological design[14].
> This in-turn creates an underpinning object relating to the manner through
> which solid seeks to render 'treatment' over what be deemed 'the commons',
> and how informatics artifacts relating to the commons is made discoverable
> and under what terms should 'commons' be made manageable, and able to be
> used privately by persons - in a manner that could be akin to the ability
> to use language objects in your head, when looking at a tree.
> How and what does Solid do, in consideration of what may in-future be
> considered 'thought crime'...    How does solid - define ontologically, the
> distinction between what an agent thought about - and what an agent did,
> and how it may be discovered in relation to what be considered 'low
> distortion ratios' for 'reality' as experienced by 'data subjects' as a
> part of their life.  Therein a diagram (created sometime ago)
> https://www.webizen.net.au/wp-content/uploads/2018/09/Credential-enabled-Identity-5.svg
> - whereby that view formed by a judge - that's kinda what it gets down to
> at the end of the day...
> Living in a world where there is data about wrong-doing that causes
> death/injury, but due to the access control frameworks asserts to the data
> about any such situation forming a revocation category from its use to save
> a life, in a court of law - that's a fairly undignified implementation
> strategy, particularly where the rationale of 'privacy' is used to form any
> such allowance for beneficiaries.
> Thereafter - the document (PG:13) asks "Who should be responsible for
> governing the WebID registry? An institution? How to build trust? How to
> finance?"
> There's a few problems to this concept, imho. for instance, a 'WebID' may
> be psuedo-anonymous.  Yet the answer, in-turn brings about the need to
> undertake a bunch of other works, that are not about w3c standards.  It
> could be a w3 business group - but i'm not sure what the economics are to
> establishing any such group.
> IMHO - there's a seperate 'thing' that's about an 'ethics bound
> implementation alliance' - which i think may be different to timBLs work
> on a contract for the the web (as a whole), whilst noting - perhaps it
> could grow into becoming an 'optional protocol', much like many other
> 'human rights', agreements of international standing...
> hope that helps.
> I suggest, the economic & ontology work are perhaps amongst the highest of
> priorities; but this is hard to define in a manner i'm satisfactorily
> convinced has merit - given there are so many moving parts. Theory is, if
> we have ontology work happening, and a means to do
> micro-payments attributions relating to the works of persons (even if, the
> POC is managed centrally via inrupt as to manage KYC/AML related stuff)
> then, we'd be in a better place to parcel up work, and get it all done.
> If there's a list of commercial sponsors that are known today - awesome -
> very interested to see the list of who they are...  ASAP.
> IMHO - we're messing with the infrastructure that's forming a diffused
> distortion array impactful upon those living with consciousness[15][16].
> So, unlike other territories of commodification of natural resources, it's
> kinda important this stuff is considered in a manner that keeps the dignity
> of others - at the heart of ones purpose; or moreover - these are my
> thoughts,
> I have faith in those of others, but i'm not entirely sure how an
> 'on-boarding' approach should or could work...   Some of the thinking,
> imho, is fairly foreign to all too many agents...
> but therein - perhaps that has more to do with my reality, than those of
> others.  Noam speaks of 'moral grammar'[17] - but if we make a world where
> 'fake news' is the information source - then, it's not really people that
> will be responsible anymore? If someone is trained to hold false-belief's
> purposefully, via income generating commodification mechanisms - how it is
> their fault, if they say or do things that are factually horrific, whilst
> potentially reasonable from their status as an 'observer' of whatever
> information it is they've been fed (and what it is they've never been able
> to make known).
> Timo.
> [1]
> https://docs.google.com/document/d/15emiG_B0XKhJgv7dq9T0mNd9eGXcJTWPGsEuVHOaDGc/edit?usp=sharing
> [2] https://w3c-ccg.github.io/did-spec/
> [3] https://github.com/bourgeoa/tiddlywiki-node-solid-server
> [4] https://bourgeoa.solid.community/public/tiddlywiki/
> [5] https://www.w3.org/community/solid/wiki/Main_Page
> [6] http://manu.sporny.org/2016/browser-api-incubation-antipattern/
> [7] https://www.w3.org/blog/news/archives/5862
> [8] https://lists.w3.org/Archives/Public/public-rww/2016Feb/0015.html
> [9] https://en.wikipedia.org/wiki/Corporate_personhood
> [10]
> https://en.wikipedia.org/wiki/United_Nations_Guidelines_for_Consumer_Protection
> [11] https://en.wikipedia.org/wiki/Information-centric_networking
> [12] https://en.wikipedia.org/wiki/Content_centric_networking
> [13] https://irtf.org/icnrg
> [14] https://www.youtube.com/watch?v=aigR2UU4R20
> [15]
> https://www.youtube.com/playlist?list=PLCbmz0VSZ_voTpRK9-o5RksERak4kOL40
> [16]
> https://www.technologyreview.com/s/613092/a-quantum-experiment-suggests-theres-no-such-thing-as-objective-reality/
> [17] https://www.youtube.com/watch?v=Kr2K8mo-A5g&t=4898s
> On Sat, 16 Mar 2019 at 20:33, Mitzi László <mitzil@inrupt.com> wrote:
>> Hi W3C Solid Community Group,
>> In preparation for our call next Thursday I wanted to share some thoughts
>> on the following agenda item:
>>    - Discuss possibility of Solid Design Requirements Specification in
>>    particular the potential for defining the default data sharing settings in
>>    such a way that the user is protected while able to engage at a minimum
>>    level.
>> I have begun to write the Solid spec chronologically i.e. detailing the
>> technical requirements when they are relevant to the user journey. It is a
>> very rough draft. The purpose of this thought experiment is not to restrict
>> the path, rather to identify where the default design is critical and if
>> there are any technical requirements that if done by a single party would
>> result in a conflict of interest to the core values of Solid. I would like
>> to talk about the minimum.
>> As homo sapiens, the default tends to be our choice, we are lazy. Rather
>> than fight our natural wiring (which anyone who went on a diet can tell you
>> is tough) I think we should reflect on the default to make sure it
>> represents our more considered choices and defined values.
>> Pat’s work on G consent could be a very useful reference to build on
>> http://openscience.adaptcentre.ie/ontologies/GConsent/docs/ontology
>> Depending on our conversations next week perhaps this could be a new
>> repository on the Solid GitHub.
>> Please excuse me for using Microsoft Word, however, it illustrates the
>> point I am trying to make rather neatly.
>> Mitzi
Received on Thursday, 21 March 2019 10:27:57 UTC

This archive was generated by hypermail 2.4.0 : Friday, 17 January 2020 17:26:39 UTC