W3C

Vision TF meeting

08 June 2023

Attendees

Present
AlanBird, amy, AvneeshSingh, Coralie, cpn, cwilso, Dingwei, dsinger, EricSiow, fantasai, Florian, gendler, hjrchung, igarashi, JamesRosewell, Jens, MichaelC, Ralph, seabass, Tantek, tzviya
Regrets
-
Chair
Tzviya
Scribe
amy, fantasai

Meeting minutes

Industry

<tzviya> w3c/AB-public#75

Tzviya:  There's been lots of discussion on this topic, including in the last few hours.
… Last week said we'd pick up today.
… Coralie created a PR
… which gets into a bunch of restructuring
… and that PR goes into a different direction.
… What I thought we'd discuss is Fabien's proposal
… for a slight rewording of what we have.

w3c/AB-public#75 (comment)

“where diverse voices from around the world and from different <ins>organizations and</ins> industries work together”

<Zakim> dsinger, you wanted to wonder if 'sectors' possibly preceded by an adjective would be better?

dsinger: I think it's trying to discuss different areas of interest, so sectors might be better.
… It's not just commercial interest.

AvneeshSingh: I agree with dsinger. The purpose was to highlight that we're worldwide
… and at the same time the different areas of technical work.
… I completely understand that governance etc might be missed out
… but if a University creates a browser, it becomes part of browser industry
… regardless of if it is nonprofit.
… So I would suggest to change the order, instead of saying “different organizations and industries”, flip the order to “industries and organizations”
… to put the emphasis on the technical diversity rather than organizational diversity.

cwilso: I would strongly argue against “sectors” or “multi-stakeholders.” It should be less specific.
… Otherwise it just says “we listen to lots of different voices.”
… [missed]
… I'm concerned about “civil society”, what exactly does that mean?
… I'm not opposed to simplifying or restructuring
… but would be concerned if we drop the word “industries.”
… And it's intentionally plural.
… There are many different industries, not just web browsers.

Florian:  I agree with cwilso.
… wrt AvneeshSingh's suggestion to flip the order, ok.
… I'm wondering if we could use dsinger's suggestion in addition, e.g. “industries and sectors”
… but concerned about excluding non-commercial interests, adding “organizations” maybe helps.

<amy> +1 to Florian's point about Industries and sectors

<Zakim> tantek, you wanted to suggest organizations & communities

Tantek: What AvneeshSingh said resonates.
… A huge portion of the Web was not built by for-profit corporation or industry.
… It was built by university-based browsers, community-based, open source web servers.
… It was built by folks not just trying to make a profit.
… At first I was questioning this issue, but now I'm swayed.
… We should de-emphasize for-profit aspects.
… Historically that has not been the focus of the Web
… and I don't think embodies the values we want to put forward.
… I would even extend it to include communities.
… There are informal communities that have contributed great things to the Web through W3C.
... So I suggest organizations & communities.

<gendler> +1 to Tantek's point on communities

<Zakim> seabass, you wanted to act as a living thesaurus

seabass: I would suggest “domain” and de-emphasizes what kind of organization.

Tzviya:  I would prefer to avoid spending our time bikeshedding.
… Let's put it in GH, and we can return to it later.

Florian: Tantek, you didn't want to emphasize industry. It emphasizes “industries”, i.e. different industries in contrast with each other.
… As opposed to removing in the plural, unless there's alternative equivalent wording.
.. So I don't want to skip it.
… It is different from “industry”, this makes the difference.

<dsinger> +1 to Florian

<tantekj> I think both singular & plural are a problem

Igarashi: I agree with the discussion.
… W3C is open for standardization, and doesn't matter what type of organization joins W3C.
… I'm opening a new issue of adding use case of industry
… because I have concern about web architecture designed for public interest sector.
… I'm not sure of the meaning
… but web tech is not used only for web-only internet, but also in various industries.
... Use cases should cover not just public interest, but also industries.

<tzviya> Igarshi's issue w3c/AB-public#84

Tzviya: Proposal here, if we can't agree, then we can go back to GH.

<tzviya> proposal: adopt "where diverse voices from around the world and from different organizations and industries work together"

<dsinger> sorry -1 on 'organizations'

<cwilso> -1

<dsinger> -1

<jrosewell> -1

<igarashi> 0

<gendler> 0

<tantekj> +0 I can live with it

<hjrchung> 0

<cpn> 0

<florian> +0

<fantasai> prefer to take AvneeshSingh's suggestion to flip the order

<koalie> +1

<Zakim> tzviya, you wanted to propose

Tzviya: OK, please comment in GitHub.

Malvertising

github: w3c/AB-public#71

Tzviya: Last comment, I'm suggesting we close this issue with no change
… since the Vision already lists [various things].
… Do we really need to list malvertising explicitly?
… There are many terrible things we don't list explicitly.

<amy> + to to not individually identifying terrible things

<Zakim> tantek, you wanted to discuss which problem we are solving with this

Tantek: Just to back up a level, I didn't think it would be controversial.
… It's the right thing to do, I should have taken it back to what problem are we solving.
… Watching the discussion of
… the higher-level problem that this is intended to improve in the Vision.
… I think we have reasonable alignment that the Vision needs to present a strong set of values that we can make decisions around
… and it does so in many ways.
… But it also provides, especially in the Intro, a list of harms that provide empirical motivation for making these values explicit.
… I want to point out that we are not just putting forth an academic abstract vision
… that we could agree to in theory but not know what it means in practice.
… By providing specific examples of harms for each value, it demonstrates more concretely what we're talking about.
… Merely mentioning an abstract value is insufficient.
… I haven't done a survey, but I think this is one part of listing specific harms.

Tzviya: We don't have any specific examples in the Values section
… other than [quotes sentence about centralization].
... I think you're making a very significant edit, to provide examples to each value.
… We need to give thought if that's improvement.
… Let's say we give an example for security.
… If someone points to that, and then people say “I didn't do *that*.”
… It's the same problem in CEPC.

<Zakim> tantek, you wanted to react to ta

<Zakim> tzviya, you wanted to respond

Tantek: It's not the only harm listed. Right there in the top of the intro, there's lists of unintended consequences.
… [quotes]
… These are all very specific. And misinformation follows those.
… Phishing is very specific, and malvertising is similar.
… It's not a giant change, we already have examples, this is just one more that fits into the kind of thing we want to avoid.
… I agree with your CEPC example, making list of examples seem exhaustive is the problem.
… We need to list examples as illustrative, not definitive.

jrosewell: Overall the intro as a whole should be rethought, and the examples Tantek mentioned removed entirely.
… The overall focus should be on safety
… and what we do as an organization to assist
… but not take on role of regulators etc.
... I hesitate until this issue, I raised another issue to rewrite intro to that end.
… The overall emphasis should be on safety and positive language.

<gendler> +1 to Tantek's comment on illustrative not definitive/enumerative

<Zakim> dsinger, you wanted to suggest a broadening of the sentence about misinformation

dsinger: I support broadening the sentence about misinformation,
... to the harm being manipulating and deceiving users

... of which misinformation and malvertising are examples.

<cwilso> +1 dsinger

<tzviya> +1 to dsinger

<hjrchung> +1 dsinger

<seabass> +1 to dsinger

<amy> +1 to expanding language re: misinformation, re: not deceived or manipulated

<fantasai> +1 dsinger

<tantekj> +1 dsinger with keeping those two as parentheticals with citations to Wikipedia

<tantekj> +1 cwilso

<Zakim> cwilso, you wanted to suggest "This has divided societies..." could be "These patterns and others have divided societies..."

cwilso:  I was also going to suggest we could introduce more expansivity
... continually calling out one more thing.
… I'm also concerned about "malvertising", which is not a word everyone knows.
... At very least, we need to call it "malicious advertising", people shouldn't need to look up terms
... but maybe we should [missed] and move on.

<fantasai>"This has divided societies..." could be "These patterns and others have
divided societies..."

<amy> + 1 to not closing the list of harms

<tzviya> +1 to it not being a common term - I had never heard it before

<tantekj> +1 to using "malicious advertising" instead of "malvertising" — and still cite wikipedia

<Zakim> seabass, you wanted to suggest de-emphasising mention of misinformation

seabass: When I think of Web, I don't blame misinformation on the Web
… any more that I blame misinformation on paper on the paper.
… I think the harm being "manipulating and deceiving users" is good.
… the Web needs to be itself trustworthy
… If a website indicates it was written by a particular identity, but is written by someone else
... this is a harm, misleading users to believe valid in a way that it isn't
... It doesn't seem to be a problem that W3C should be concerned with.

<cwilso> -1 to deemphasizing misinformation

<Zakim> tzviya, you wanted to respond to seabass

tzviya: The fundamental difference between paper and web is there's no interactivity on paper.
... Not talking about something where W3C provides standards, do enable misinformation etc.
... We need to be cognizant of these unintended consequences.
… But some of these features also make the web great.
… There's a balance to achieve.
... We have to recognize that there are unintended consequences, and do something about them.

<jrosewell> +1 to seabass

<hjrchung> +1 tzviya

<gendler> +1 tzviya

<tantek> +1 tzviya fundamental difference of physics of the web vs paper

<Ralph> [paper doesn't have the capability to physically alter the reader's environment -- modulo finger cuts and fire]

<Zakim> tantekj, you wanted to build on seabass examples of false identities on the web used in fake reviews to deceive consumers

Tantek:  I strongly agree with Tzviya.
… The physics of the Web are fundamentlly different from physics of print.
… Things spread at a speed on the Web that is far greater than any print media
… therefore it has a much different impact on people and society, and politically.
… Presenting info with falsified identities etc., yes that's deceiving the user.
… This happens all the time with sites with posts that show reviews and ratings
… that basically propagate false info about identities provided to inflate reviews on the Web.
… This harms users and consumers every single day.
… We have active harm to commerce, it's essentially a form of fraud
… and that's a new harm amplified by Web, above and beyond print media.

Tzviya:  Going back to paper, remember the article about vaccines and autism did a lot of harm too.

jrosewell: Since the dawn of time, tech has been used for good and bad.
… We should recognize that it's our job to where we can provide the tools for people to make choices
… e.g. many ppl choose media org that they trust.
… We should simply help them make smarter choices
… but we are not in the business of limiting harm.
… How do you decide what is a harm, what is free speech, etc?
… We should focus on knitting of technical standards.

<Zakim> seabass, you wanted to respond to tantek re. speed of transmission and identities

seabass: With regards to whether speed is an element,
… I don't think the speed of the Web is a factor of making culpable for misinformation.
… That would apply to any Internet protocol.
… One area W3C has responsibility in is Verifiable Credentials.
… If people are more likely to trust because of using this standard
… but if it's falsified, it's W3C's responsibility to fix that.

Florian: I want to disagree with jrosewell. Yes anything can be used for all sorts of purposes
… and we can't fix that you can say evil things with written language
… but tech isn't neutral.
... *How* you do something can make Web more or less susceptible to misuse

... from the get-go. The Web has been opinionated, tried to uphold values
...we cannot prevent all harms.
… But shouldn't pretend we have no influence on misuses.
.... We should identify these misuses and do waht we can to steer clear.

<tzviya> +1 technology is not neutral

<tantek> +1 technology is not neutral

<cwilso> +1000 technology is not neutral

<amy> +1 to Florian re: technology not neutral and misuse of tech

<hjrchung> +1 florian

<gendler> +1 to Florian on multiple sentences

<Dingwei> +1 to Florian

<dsinger> +1 we can't 'fix' it but we can, when possible, make good easier and harm harder (and thus more and less likely)

<gendler> Koalie, I wanted to respect Tzviya's ask for lack of over drilling :)

<Zakim> tzviya, you wanted to suggest we edit the sentence as dsinger suggested

<Ralph> [ https://mitsloan.mit.edu/ideas-made-to-matter/study-false-news-spreads-faster-truth ]

<seabass> +1 florian. "Does a W3C standard make misuse easier?" is the question

<jrosewell> -1 technology standards need to be neutral

Tzviya: I have a proposal to adopt dsinger's suggestion.

<tzviya> proposal: Rapid global information sharing has allowed misinformation to flourish and be exploited for political or commercial gain. This has divided societies and incited hate.

<fantasai>w3c/AB-public#71 (comment)

<tzviya> +1

<Alan> +1

<jrosewell> -1

<gendler> +1

<cwilso> +1

<amy> +1

<hjrchung> +1

<fantasai>+1

<cpn> +1

<seabass> abstain, as preferring to replace sentence entirely with dsinger's improved version

<koalie> +1

<Ralph> +1

<fantasai> seabass, I think that's the suggestion

<tantek> -1 this has nothing to do with this issue, I don't disagree necessarily but this doesn't close the issue

<florian> 0

fantasai: Just to clarify the suggestion is to replace the quoted language with the suggested text
... if we agree we should be adopting this.

<igarashi> 0 I agree with the sentence, but I am not sure it should be included in the vision.

Tzviya:  Tantek, can you live with this temporarily?

Tantek:   It's orthogonal to this issue.
… I don't think it moves this specific issue forward.

<florian> 0; for the reason tantek said

<AvneeshSingh> should we maintain status quo for now?

cwilso: I'm happy for you to work with dsinger or me on that.
… I think at this point, we could take a poll
… but I think adding "malvertising" to the list would not pass muster with this group.

Tzviya: I'm going to suggest that cwilso, dsinger, and tantek take this to drafting.

<tantek> +1 to working offline on this

dsinger: Let's see if we can get something better.

<tantek> noted that there was a proposal to use "malicious advertising" instead of malvertising that was +1

Types of Web

Types of Web w3c/AB-public#85

github: w3c/AB-public#85

Tzviya: Suggests to add sentences: “W3C leads the Web forward, towards the future Web, the semantic Web, the Web of data, the Web of everything (humans, machines, things, processes, services and linked data).”
… A lot of people had concerns about this
… and didn't think we should make this change.
… So the question today is whether people support this.

<tantek> -1 on this, agree with mnot

cwilso: My concern is that it adds a bunch of buzzwords and connections into the Vision, without defining principles of why those things are important.
… [quotes]
… It seems like a collection of cookies that we're trying to lick, and I don't want that in the Vision and Principles.

<amy> +1 to not including this new text into the vision

<gendler> +1 Chris

<tantek> +1 cwilso

<hjrchung> +1 cwilso

Tzviya: I agree, and seems most commenters agreed on this.
... Also +1 to mnot's comment
... Proposed to close, won't fix.

POLL: Close as WONTFIX

<koalie> [Jens arrives]

<florian> +1 to closing

<seabass> +1

<Dingwei> +1

<koalie> +1 to closing as wontfix

<dsinger> +1

<igarashi> +1

<gendler> +1

<amy>> +1 to closing

<AvneeshSingh> +1

<cpn> +1 to close

<tzviya> +1

<Ralph> +1

<fantasai> +1

<koalie> +1

<cwilso> +1

<hjrchung> +1

<Alan> +1

<tantek> +1

<Eric_> +1

<jrosewell> +1

RESOLUTION: Closed WONTFIX

<tantek> has to run to another meeting but would like to contribute to discussion of issue 17 "next week" as stated in the agenda https://www.w3.org/wiki/AB/VisionTF/2023-06-08

<tzviya> w3c/AB-public#88

Unintended Consequences

github: w3c/AB-public#88

Tzviya: Fabien suggested breaking up this sentence.
… I'm hoping people have read this:

tzviya quotes from the issue

[[The Web has been a force for good, and has sparked major social changes. But the Web's amazing success has led to many unintended consequences that harm society. For instance, openness and anonymity have given rise to scams, phishing, and fraud. The ease of gathering personal information has led to business models that mine and sell detailed user data, without people's awareness or consent. Rapid global information sharing has allowed misinformation to flo

urish and be exploited for political or commercial gain.]]

Tzviya: We're not discussing content of sentence, just format.

<amy> +1 to easier to understand

<cwilso> POLL: option 1: enumeration. option 2: split sentences. option 3: leave as-is

cwilso: There were 3 options, so let's consider the three options.
… The first option was enumeration with semicolons,
… the second was split into sentences,
… the third is no change.

<Alan> 2

<fantasai>1

<tzviya> 2

<florian> 1 > 2 > 3

<AvneeshSingh> +1 to option 2

<koalie> 2

<cpn> option 2, suggest splitting into multiple sentences

<amy> 2

<seabass> 1

<cwilso> 1>2>3

<Alan> 2>1>3

dsinger: If you split it, it makes the first thing look like it's the only example.

Florian: I think both new options are better than original, and yes short sentences are nicer than long ones.
... But in the split sentence version, the two sentences afterward seem like they're not examples, because only the first sentence is introduced as an example.

<jrosewell> 0 - none of the options as the section needs to be replaced to be positive and focus on safety within the limits of neutral technical standards

<gendler> 2 > 1 > 3

<Dingwei> 1

dsinger: Can we leave this to the editors?

<amy> +1 to editors working on it

<hjrchung> +1 dsinger

fantasai: +1 to florian

Tzviya:  So  the conclusion is that this needs tweaking.
... We work on rough consensus here.

<cpn> +1 to editing so it's clear as well as splitting into sentences

PROPOSED: Improve the sentence, leave to editors as to how.

<tzviya> +1

<fantasai> +1

<gendler> +1

<amy> +1

<Dingwei> +1

<hjrchung> +1

<florian> +1

<koalie> +1

<Eric_> +1

<cpn> +1

<igarashi> 0

<Alan> +1

<jrosewell> abstain - see previous comment

<cwilso> +1

<dsinger> +1

<Ralph> +1

AvneeshSingh: Just want to mention, for non-native speakers, improving is important.
… I chose option 2, because for screenreaders it will be provide appropriate pauses.

<seabass> +1 AvneeshSingh. The bullet point is always an option

RESOLUTION: Improve the sentence structure, leave to editors as to how.

Integrity

github: w3c/AB-public#13

Tzviya:  This isssue is about being more explicit about improving the Web’s integrity.
… I believe various changes since have accomplished this, and we can close.
... mchampion suggested we give more info on security/privacy,
… and a suggestion on how to make how to make these things actual.
… Does this go into the Vision, or a Strategic Objectives document?

cwilso: I think this is a good draft of those Strategic Objectives.
… So I suggest we keep it open until we start work on that
… but don't want to add into this document at the moment.
… So I don't think anything to do on this issue for today.

<AvneeshSingh> +1 to do nothing on this issue today

<amy> +1 to marking under strategic objective and coming back later

Chris: Let's mark this under strategic objectives and come back later.

Administrative

Tzviya:  We have 5 mins left so let's not go further. Two weeks from now we'll address some of these issues.
… Coralie made a pull request on one and we'll address others then.
... The Vision TF is scheduled to go through end of June. The last call will be the last (so far) scheduled call. We'll figure out in the next AB who will Chair this.
… Some from of the Vision TF will likely continue.

fantasai: Chris, we have an outstanding AB resolution to publish as draft note. What's the plan?

Chris: I've been swamped with finishing something from my day job. I'll take a first stab at closing this.

Florian:  Would it help to explain scripts for building process? Look at the Process and if there are any issues, ask me.
… Put time in my calendar and we'll work it out.

David: I think it's time to start looking at this as something we'll put in a formal state.

Meeting closed.

[adjourned]

Summary of resolutions

  1. Closed WONTFIX
  2. Improve the sentence structure, leave to editors as to how.