Re: CredWeb Outros (was Re: long time no chat, CredWeb friends)

Sandro wrote:
"I'm particular curious how people now feel about the Credible Web Mission."

I believe that the mission statement is seriously weakened by its almost
exclusive focus on the consumption of content or claims about content. The
stated mission is to enable consumers to "tell when content is
reliable...,' etc. and to help consumers "find the content they want..."
This is all good, but what's missing is an explicit charge to facilitate
the expression of the "credibility indicators" upon which other consumers
might rely. Certainly, it is implied that some entity(ies) should be able
to make claims, but, I believe it would be useful to stress that *all*
consumers should be capable of providing "credibility indicators." Rather
than entrusting the task of making statements about credibility to some
restricted class of organizations who use specialized tools, we should be
enabling every consumer to be a "pro-sumer;" one who both consumes content
as well as produces claims about content.

Rather than building new systems which are dedicated to the providing
credibility data, I believe that it would make a great deal more sense to
focus instead on leveraging those systems that already exist for making
statements about web resources. By this I mean the SocialWeb systems such
as Twitter, ActivityPub, Mastodon, etc. Given that a very high percentage
of SocialWeb posts links to or comments about other web resources, these
systems already provide much of what is needed to provide an effective
foundation for a credibility-indicator publishing and dissemination system.

For instance, earlier today, I noticed that the National Academies of
Science recently published a report on *Accelerating Decarbonization in the
USA
<https://nap.nationalacademies.org/catalog/25931/accelerating-decarbonization-in-the-united-states-technology-policy-and-societal>*.
Since I wanted to see what others had to say about it, I went to Twitter
and searched for the report's URL in the Twitter search bar. The result was
a list of tweets either just simply linking to the report or providing
comments on it. Because I recognized some of the Tweeters who endorsed the
report, particularly Dave Roberts, I now have the sense that this report
might be worth reading. I regularly do the same thing to uncover commentary
on web resources that interest me.

But, while Twitter provides a crude annotation system that I can use to
discover informal "credibility indicators," it is clear that the experience
could be greatly improved. For instance, if instead of simply providing for
"likes" or prose comments, it was possible to publish structured data (
ClaimReview <https://schema.org/ClaimReview>?) in a SocialWeb post, my
client, or some application, might be able to summarize the claims in some
manner useful to me. While getting Twitter to adopt such things might be
problematic, it might be more fruitful to work with the W3C SocialWeb
community group <https://www.w3.org/community/socialcg/> to incorporate W3C
Annotations <https://www.w3.org/TR/annotation-model/> into
ActivityStreams/ActivityPub
<https://www.w3.org/TR/activitypub/>and to provide a standard method for
encoding structured credibility-indicators as social annotations. Ideally,
these mechanisms would eventually be incorporated into tools like
Mastodon... Once it is possible to publish credibility data into the
already established SocialWeb data channels, it would then be useful to
move on to addressing other issues. For instance, the development of
systems that filter or usefully summarize credibility-indicators coming
from a large group composed of individuals having widely varying
credibility. (The "Credibility Web" stuff that Sandro has sometimes spoken
of in the past would be useful in filtering...) (Note: For another view of
the intersection between W3C Annotations and Credibility, see Jon Udell's
2018 post on the subject
<https://misinfocon.com/how-web-annotation-can-help-readers-spot-fact-checked-claims-ccbf9246dd68>
.)

My strong belief is that CredWeb could make a very useful contribution by
addressing the problem of incorporating credibility-indicators as
annotations within the SocialWeb. If I were to edit the mission statement
today, I would do so with the aim of focusing efforts on leveraging the
existing SocialWeb systems which already provide crude methods for the
expression and distribution of credibility-indicators by all web consumers
(or, at least all those with SocialWeb accounts).

bob wyman

On Mon, Oct 23, 2023 at 3:00 PM Sandro Hawke <sandro@hawke.org> wrote:

> Thanks Scott and Sara-Jayne.   Just to clarify, after talking with Scott,
> the meeting is *Wednesday, 1 Nov 2023* (9 days from now).  Zoom link will
> be posted to credweb.org.
>
> Also, credit to An Xiao Mina, who co-founded the group. She went on to
> other things, but she was a key part of the early stages of the group.
>
> Continuing what Sara-Jayne started, I suggest everyone reading this post
> an "outro".  (We started the group with folks doing "intros".)  I'm
> particular curious how people now feel about the Credible Web Mission
> <https://www.w3.org/community/credibility/>, and what kind of work they
> might still be doing in the field.
>
> For myself, I'm still 100% on board with the first paragraph of the
> mission, but not the use of RDF in the second part.
>
> I'm working full time to make this a reality. I've pivoted a few times in
> my approach, including moving away from W3C, and these days I see the best
> path to the Credible Web being an integrated data/app platform.
>
> That is: make credibility data-sharing a part of general data sharing
> (including personal data). That way, the credibility system can use the
> rest of the data sharing system to manage credibility assessments, and the
> rest of the data sharing system can use the credibility system to maintain
> trust and safety.
>
> I'm calling this platform with integrated credibility the Human Data Forum
> <https://humandataforum.org>. If you want to chat about it, or just want
> a beta-test invitation when they're available, let me know. That's slated
> for December.
>
> I hope you're all doing well, and will see some of you on the call Nov 1.
>
>      -- Sandro
>
>
>
> On 10/23/23 13:57, Sara-Jayne Terp wrote:
>
> I can't make the meeting because I'm on stage at OODAcon then (talking
> about trust landscapes with Heather McMahon), so here's a written update.
> I went back to live in the UK, and am now one of the tech leads on ARCD - a
> project using machine learning to respond to cyber attacks at machine speed
> / in places where there's no cyber expert available (we have about 60
> projects live around the UK right now).
>
> I took a year out from disinformation response (the DISARM Foundation
> continues the AMITT/DISARM work), but it's starting to drag me back in
> again - I recently started an ATT&CK version for the mis/dis/mal around
> LLMs, so we can talk about how they're duped vs how humans are; and had fun
> using LLMs to create an information training environment, with LLM chatbots
> doing narrative attack and response.  Other than that, I've been pretty
> quiet.
>
> Sj.
>
> On Mon, 23 Oct 2023 at 13:02, Scott Yates <scott@journallist.net> wrote:
>
>> You are getting this email because at some point you were a part of the
>> CredWeb group in W3C, started by the incomparable Sandro Hawke.
>>
>> Then I came in and tried to create a new mission for the group, and then
>> we all just moved on.
>>
>> This group is now going to get deleted unless we find a new source of
>> inspiration and leadership, but before we do that... maybe we could just
>> get together to get quick updates on what we've all been doing?
>>
>> Let's do it at the time we used to do it: this Wednesday at 4 p.m. in
>> London, 11 a.m. Eastern, 8 a.m. in California.
>>
>> Here's a Google Calendar link.
>> <https://calendar.google.com/calendar/event?action=TEMPLATE&tmeid=NWxmcmJtYWhwajBpbmt1dWVtZmYzNm04bzUgc2NvdHRAam91cm5hbGxpc3QubmV0&tmsrc=scott%40journallist.net>
>>
>> I hope to see a lot of you there! If you can't make that time, feel free
>> to drop Sandro and me a note and let us know what you are up to, and if you
>> have any thoughts about keeping the CredWeb group going.
>>
>> -Scott Yates
>>
>
>
> --
> Sara-Jayne Terp
> sarajterp@gmail.com | +1 646 400 8497 | overcognition.com | @bodaceacat
>
>
>