Sustainability Principle is anticompetitive #66
Comments
|
I disagree with most of the comment above, and would like to make one point in particular
We mainly work on new technologies. Improving the principles we use to evaluate them is a good thing. Over time, we have applied greater emphasis on privacy considerations – new technologies have to pass more and better privacy checks than before. And this has led to better privacy overall. It happens quite often when we discuss the fingerprinting surface of some new tech, that someone points to some fingerprinting risk that already exists. “We allowed this before, why not continue to do so now?” This is a nihilistic argument that should be rejected. We should strive to do better with new technologies, and improve what we can in the old (but that work is subject to much higher compatibility restraints). |
|
@astearns Even if there are good intentions in the principle itself, that doesn't excuse the ethical dilemma. When W3C members rely on the sustainability principle as a strawman critique against a nascent technology, and that technology just happens to pose a threat to the business models of the few W3C members asserting the principle, it makes the principle an extremely convenient weapon against competition. For example, a company recently argued that the largest W3C members have weaponized the sustainability principle against them in "Does the W3C Still Believe in Tim Berners-Lee’s Vision of Decentralization?". Take note of the appearance of conflicts of interest listed in the section titled: "If these objections do not hold up, then what are the real reasons Mozilla, Google, and Apple are opposed?" The concern outlined in the links, above, is that the sustainability principle is being wielded as an anticompetitive weapon against competitors that may or may not rely on an emerging technology that isn't even required in the specification that's being proposed. |
Do you have data on this? Video only initially launched with no hardware acceleration; ever since it received hardware acceleration energy usage hasn't been a serious issue. An average consumer-grade laptop with HW decoding can easily get 6+ hours of usage. |
[…]
The purpose of the W3C TAG Ethical Web Principles is literally to "inform TAG review of new specifications" (emphasis mine). That it caters to new web technologies is by design and isn't specific to that particular principle. That said, there are potential tensions between principles, as mentioned in #62 and tentatively addressed in #64. Frankly, ethics wouldn't be a whole field of study if that wasn't the case. Of course, decisions will have to be made when principles conflict with one another. But even then, competitiveness itself isn't such a principle. And I don't see how it would meet the bar of becoming one. However, thinking about the web as a vector of undiscriminating economical opportunities would lean in the direction of requiring more competition but would anchor it in a humanist perspective. That's a discussion I'd be interested to have (in a separate issue). |
|
@cynthia Your question highlights the problem. Your instinct, in defense of
If that seems unfair and unrelated to |
|
@tobie Correct me if I'm wrong, but the Sustainability Principle appears to be the only principle that isn't tied to end user experience. Every other principle is directly tied to objectively making the macro and micro web experience better. User-centric principles make sense. But having a principle that has nothing to do with the user experience, and frankly nothing to do with the Web itself, makes it too easy for any member who dislikes the business implications of a proposed spec to make the loose environmental associations described, above. I would argue that TAG Ethical Web Principles should be limited to the Web, and W3C members should not be in the business of promoting degrowth on competitors or guestimating what energy mix a particular loosely associated technology may or may not use today or in the future. |
This comment has been minimized.
This comment has been minimized.
|
As a reminder, this repository (and everything else in the |
This comment has been minimized.
This comment has been minimized.
☝️ Exactly. That is entirely different from claiming 👇
One is a thorough review of highly specific technical implications, for Web users, directly related to a spec. The other is searching for distant abstractions and associations that can be based on exaggerated media reports. The history of open competition has always been dependent on allowing access to competing or increased levels of energy as civilizations grow. It doesn't make sense for the W3C members to be scrutinizing or policing energy consumption of competing technologies—particularly technologies that aren't even required in a proposed spec—while simultaneously enjoying unfettered access to energy. It's anticompetitive. |
|
I'll avoid the question of whether advocating for an ordering / priority of web principles is or is not anticompetitive, but regarding sustainability as it is related to power consumption, there can be data driven arguments over some aspects of that principle...
The last link in the list is most relevant to this thread... if we consider power saving / sustainability as the "most important principle" ... there is a clear ordering of recommendations for users with respect to selecting a particular browser... I don't think facts are "anti competitive" but... maybe this issue will help us decide. If W3C were to make the case that sustainability is the most important ethical web principle (i don't think anyone has proposed this concretely btw), it would not be hard to make the case that everyone should switch to Vivaldi Browser (if you believe those numbers...not sure I do). I don't think this is what folks are saying... clearly a balance of principles is needed, and we should consider the past present and future when making statements about principles. I do think the W3C should be cautious in selecting principles and defining a ranked order of importance because, values change over time, and sometimes what seems like a really smart thing to say... turns out to be not so smart years later...
Given the relationship between power consumption and memory and cpu especially as they relate to machine learning or other "valuable" computations... I think its safe to say, we would all love if the valuable computations could be achieved for less power... We could just as easily rank hardware providers with respect to power consumption:
Just because "sustainability" is a principle does not mean the W3C is advocating against particular browser or device vendors. That only happens if the W3C started pointing to those browser or device manufacturers by name, and ordering them according to "sustainability"... which I sorta did by linking these articles... I don't think that is the same thing as saying "try not to 'waste' electricity or try to be sustainable"... but also, we should be careful not to use the word "waste" incorrectly... being able to view a gif, or run javascript, or do machine learning is not a waste.... If power consumption provides value to users... I would call that a "cost".... and note that it subjectively... might not be "worth it"... and if you really care about power consumption, you can use your favorite search engine to find the devices and browsers that perform best with respect to power consumption. Another thing to consider is the total cost of the things in question... its not just how much power does a new cpu use... its how much power / carbon went into making it... the same is true of GPT-3 or other really large machine learning models.... you can't just say "well it only costs a little energy to get a classification"... you have to consider how much power it took to train that model, and how much carbon was used to produce all the cpu's that were required to crunch that data... modeling the "cost" of computations is a really interesting and complicated science... I'm not sure how "fair or biased" this Forbes article is, but it highlights my point pretty well:
I don't think making sustainability a principle means W3C is telling folks who to buy from... or that data centers powered by diesel and used to "improve user experience" is incompatible with sustainability. In the hopes of keeping arguments regarding web principles data driven I will end with this old and therefore probably incorrect nature article:
Yes, Google, Facebook, and Bitcoin all have power costs associated with their use.... It's a shame Bitcoin isn't a W3C Member to defend itself.... but I don't think its "anticompetitive" to debate the cost of sustainability, or keep it as an EWP. |
My reading is that several of the principles in the current draft extend beyond the end user experience. For example, harm to society and offline harassment are other impacts (like sustainability and carbon emissions) that extend beyond the current end user. Misinformation is another where the impacts are beyond the end user, although the principle as written focuses on addressing that by enabling each end user to understand provenance. Privacy and security are certainly relevant to the end user, but we should also consider privacy as a potential societal good and there are privacy interests beyond the user of a particular web page (sharing other peoples' data, for example). I think these are all very relevant to the Web and considering its impacts; just noting that environmental sustainability is not unique in that respect. |
@OR13 This entirely misses the concerns raised in the articles and posts I linked, above. I agree that the principle has noble intentions and, in theory, could be discussed with data driven arguments. But that's not at all how the sustainability principle has been used in practice. The only time I've personally seen the sustainability principle invoked publicly, it was used in the following manner:
• https://lists.w3.org/Archives/Public/public-new-work/2021Sep/0000.html I can't stress this enough, the proposed specification that was being objected to in that particular discussion stated nothing about using the technology that member was concerned about. This was an example of a key W3C member selectively attacking a proposed spec because something far removed from the actual spec/code might be optionally used by some methods. And the spec in question just happened to challenge their existing business models. That's not a data-driven argument. That's association fallacy, submitted with presenter bias, to protect business interests. If key members can feel justified citing the Sustainability principle based on exaggerated media reports on technologies that are far removed from the spec in question, whenever its convenient for their balance sheets, how does that enable fair and open competition from new technologies? It clearly doesn't. |
|
In the interest of making this conversation productive, I think there's value in breaking it down into different actionable parts. I've heard the following points be made in this discussion:
So unless I've missed additional points, I would suggest continuing these conversations in the relevant places and closing this issue. |
|
I think its currently acceptable for an AC Member to pick and choose the principles that they want to apply or feel are most important when reviewing a proposed REC and they are not required to be consistent in which they choose or in how they argue in favor or against the principle being supported... In fact, they might even be frustratingly strategic in their inconsistencies.
I don't think AC reps are required to disclose why they think one principal is more important than another for a specific standard, or if they or their business would benefit or suffer materially from a "new tracking / fingerprinting API" or "potential movement in a crypto currency" or "end to end encryption without backdoors"... I guess the toughest part of your quote for me to address is the "key members" part.... this seems to imply that some AC folks are more important than others, or that they might have more authority when responding to a Call for Review... I don't think that is the case, but even if it were, as I mentioned above, they are still human beings. There is a certain political inevitability to standards... I think @tobie covered it best here:
I would also add this:
@frabaghe maybe you feel that the Sustainability Principle is being used in a way that feels like bullying? Maybe some of the sustainability debate has impeded work, but at the same time it has generated a lot of discussion:
Also, I don't think CEPC applies to activities the AC performs w.r.t. specific working groups. I agree with @tobie unless there is a specific text change you think should be made to address the impact "sustainability" has on "competitiveness" this issue should be closed, and probably its better to just take that debate to #68 |
|
@tobie Thank you. That seems like a good start. 👍 I would suggest that it would be worth considering some changes to the language to encourage new and advanced technologies. While I'm sure we all agree that net emissions should be minimized, the current wording gives the impression that higher power consumption is inherently bad. I do not believe that is a viewpoint everyone would agree with. In 1964, Soviet astronomer Nikolai Kardashev proposed that a civilization's level of technological advancement could be measured by the amount of energy it is able to harvest and use. Kardashev placed civilizations capable of using all the energy that falls on their planet, from its parent star, at Level 1. We need to advance ourselves as a species to be better at energy production and transportation. Energy is the means by which humans have transformed the basic state of things. It began with the discovery and mastery of fire, which led us to the ability to cook which created the energy surplus that allowed us to evolve our large brains. Human awareness itself sprang from energy, and the ingenuity which came with it has enabled us to create everything that underpins modern life. Mankind figured out basic hydropower and windmills, steam, nuclear, and perhaps fusion will be next. And we will keep advancing in terms of our sources and uses of energy. Today a single person in the West enjoys the energetic equivalent of 600 humans working for them. Every single increase in energy efficiency brings with it an advancement. From a fair competition standpoint, the current wording of the Sustainability principle implies new technologies cannot be allowed to utilize more power than existing technologies. This is concerning given that every advancement in technology has led to increased power consumption that came with significant leaps in efficiency and benefits to society. If we believe, as Kardashev did, that civilizations advance as they increase their power consumption, then the Sustainability principle appears to worded in a way that would stifle human progress and competition. My suggestion is to change the language so as not to put forth the idea that the W3C believes in degrowth, which would come at the expense of human progress. Degrowth is anticompetitive as it means that new and advanced technologies should not be allowed to flourish and have access to more power. Additionally, the wording lacks nuance. If that increased power would otherwise be wasted or dissipated, an emerging technology that utilized that wasted power, and dematerialized other polluting industries in the process, might very well be considered a good use of increased power consumption. To imply that any significant increase in broad power consumption is bad would only serve to stifle human progress while not solving any real climate issue. Therefore, I would recommend considering new wording, such as:
My hope is that the Sustainability principle both promotes new competition and enhances human progress without demonizing increased power consumption. |
|
As noted above, this document is primarily about informing how the TAG reviews designs of new technologies and in doing so to help create a better web. We are also working to advance it towards becoming a W3C Statement at which time it may take on more formal significance, however it intentionally does not contain actionable guidance. That actionable guidance sits in the TAG's design principles, the Security & Privacy Questionnaire and other documents that the TAG is currently developing. Also as also noted above, all TAG repos do operate under the W3C code of conduct which we take very seriously. We can and do block aggressively when troll-like behaviour surfaces in our issues. Some of the above is getting close to this threshold. Having said that, whilst I do not accept the “anticompetitive” framing of this issue, there may be merit to the new text suggested by @frabaghe. Can we re-focus the discussion in this thread on that narrow issue - the possible reformulation of the principle? |
I second that. Here’s a diff of @frabaghe’s proposal against the current wording, for reference (updated as per @frabaghe's below comment):
|
Agreed.
Corrected:
|
|
Just jumping in to say: I, too, find merit to what @frabaghe is saying, having been on the receiving end of the Sustainability principle being used in an inappropriate way (assertions w/o citations wrt. sustainability as they relate to Decentralized Identifiers). I can also see the anti-competitive argument that @frabaghe is making, but don't expect those opposed to it to understand why any time soon. I find @frabaghe's original proposal and @tobie's rewording compelling. Is there a way we can insist that "sustainability justifications need to be data-driven"? Sure, you can read data like tea leaves, but in the case of Decentralized Identifiers -- zero data analysis was provided to demonstrate "clear harm with no advantage" in the formal objection by the objectors. |
|
We (@torgo, @hadleybeeman, @cynthia and I) discussed this in a breakout at the TAG face-to-face today. I want to acknowledge that "power consumption" and "emissions" are broad generalisations that are not actually that helpful when thinking about environmental impact of technology. All power generation is not created equal, so talking about "more" or "less" of it in a broad sense doesn't mean a lot. However it isn't in scope to turn this section into a primer on power generation. It is in scope to remind the reader to think beyond their immediate world view about these issues, which is perhaps where a lot of the knee-jerk reactions for/against certain technologies come from. Terms like "civilization" are extremely subjective and vary across cultures (and time). What is "civilized" to one group of people may be extremely objectionable to another. I think perhaps what is meant here is more like "society" in the sense of "everyone in the world". I want to be careful with using "net" (emissions, efficiencies, benefits, etc) because for the most part this is extremely difficult if not impossible to measure objectively, even if you constrain it within a particular time frame. Weighing the perceived benefits against perceived costs of a particular technology can result in completely different judgements depending on a person's priorities (and ability to predict the future). It also feels like there's a risk of an assumption of utilitarian ethics here, but it is also out of scope for this document to suggest a particular way of weighing up ethical considerations (thousands of years of philosophy study hasn't figured out yet afaik). Following our discussion today and this and related threads, I'll rewrite the sustainability principle to make it say more what we mean, and hopefully make it less weaponisable. |
|
The current rewording still considers sustainability in isolation, when its intent is to allow advancing EWP even though sometimes using more energy. I would add the following changes (redlined version):
(complete version)
If sustainability is considered in isolation, without regard to advancing other EWP, then it never allows plugging another computer into the Internet for the sake of learning, communication, or advancing digital human rights. |
|
"increased power consumption can beneficially advance civilization forward, " |
|
@rhiaro Thank you. That sounds like a good plan. 👍
Indeed it is difficult. Therein lies the conundrum of this principle. If we don't consider the net benefits, then it's quite easy to believe that—as @rxgrant points out—we can never allow the plugging in of another computer into the Web. For example, video streaming services caused the demise of the inefficient brick and mortar VHS/DVD rental & sales industry, which had its own environmental impact. Looking at the environmental impact of modern streaming services, in isolation, misses the point. So, my intent with the rewording was to suggest that the thinking that goes into evaluating this principle should not be particularly narrow. If a new technology can dematerialize the environmental impacts of legacy industries, that should be considered.
@masinter Kardashev proposed that the amount of energy a civilization is able to use, is a proxy for its technological advancement. (See the Kardashev scale) My point is that increased power consumption/expenditure is not inherently bad. There are two competing schools of thought. Neo-Malthusians (“resources are finite”; degrowth; scarcity mindset; zero-sum, finite games) and Prometheans (“human imagination is the most valuable natural resource”; daringly creative; abundance mindset; positive sum, infinite games). The W3C emerged from Promethean ideals that required increased power consumption and the early stage participants were once young, weak and risked failure. If the Sustainability Principle retains only Neo-Malthusian ideals, we risk stifling the young (Promethean) startups and technology of tomorrow. Neo-Malthusians can very easily justify resource-restriction to suit biases (or even anti-competitive business or atrocities), yet don't recognize that the materials needed to make silicon chips had little value before the computer was conceived. Imagination is the true resource. Indeed, we cannot predict the future. However, we need to make sure the Sustainability Principle doesn't hold us back from advancing as a species. Human progress will come from advancing up the Kardashev Scale, not degrowth. |
|
I think considering power consumption and hardware longevity are valuable pointers in a high-level principle and that we should work in a separate sustainability group to give more detailed insights into sustainability impacts that might come from any particular W3C specification. Removing any mention of power consumption or carbon emissions, or only promoting power consumption as inherently advancing civilization, would be extremely confusing in a high-level principle about sustainability. I do not see any evidence that the practice of considering power consumption will inherently require commitment to Malthusianism. As @tobie noted above, there is interest in explicitly noting tensions between different principles and how to balance them. I don't think singling out sustainability as the only principle that must be qualified based on other goals is productive. |
|
@masinter wrote:
That's not what's being said, by anyone, AFAICT. What is being said is that the consumption of power has to be weighed against what that power is being used for (the overall cost/benefit). Clothes washers and dryers are notoriously energy hungry (water distribution, heating water, heating air to dry) -- massive energy costs. Those massive energy costs also enabled women to liberate themselves and enter the workforce in ways that were not happening before that particular type of energy consumption: https://www.sciencedaily.com/releases/2009/03/090312150735.htm I cringe a bit to use that example (because it's a bit charged), and I wish I could use another one in the future... but hope those in this thread understand the gist of what is being proposed. You can't just say: "That thing is using lots of energy, so it's clearly bad!" -- you have to weigh the energy usage against what is being achieved with the energy usage. |
The technology born from W3C itself is a good example:
|
This comment has been minimized.
This comment has been minimized.
|
@csuwildcat, you realize I was agreeing with your arguments, right? :) @msporny: "And that's why I think @csuwildcat, @frabaghe, and @rxgrant's arguments cannot be easily dismissed." I said "I cringe a bit to use that example" partly because of this sort of knee-jerk reaction -- people won't read the article, or the scientific papers the article was based on, or understand that the study covered life in the United States between the years of 1900 to 2000... where things were sexist and there was a gross imbalance in household duties and the workforce due to traditional gender roles... and how burning more energy to automate work made things better (overall). |
This comment has been minimized.
This comment has been minimized.
This comment has been minimized.
This comment has been minimized.
This comment has been minimized.
This comment has been minimized.
|
Posted January 10, 2019. https://masinter.blogspot.com/2019/01/internet-is-a-wmd.html |
This comment has been minimized.
This comment has been minimized.
|
I’m sorry, @csuwildcat, calling people whose sources you disagree with “superficial NYTistas” has no place in this community, nor has your subsequent comment which is just gratuitously insulting the work of members of this community. Disagreeing is fine. This is neither. To be fair, I don’t think @masinter’s comment is conducive to a productive conversation either. If y’all were willing to edit your comments to help keep this conversation on topic and friendly, that would be very much appreciated. |
This comment has been minimized.
This comment has been minimized.
As the above quotes show, Kardashev's scale has been brought up multiple times in this conversation in order to argue that an increase in energy consumption was not only correlated to progress, but even necessary for progress. This argument is incorrect for a variety of reasons. First, Kardashev's scale (described in a 1964 paper) was designed with a very specific field of application in mind: the search for extraterrestrial intelligence (SETI) and in particular through the detection of radio waves. It never was intended as a useful model beyond that field. Secondly, Kardashev's paper doesn't correlate energy consumption with progress. It merely hypothesizes a regular 3% to 4% yearly increase in energy consumption based on data from Palmer Cosset Putnam's 1953 book Energy in the Future. Thirdly, Kardashev's explanation as to why energy consumption should stay exponential (rather than, for example, follow a logistic curve) is entirely lacking. He sees "no reason why the tempo of increase in energy consumption should fall substantially than predicted"—IEA data actually shows a more modest average yearly increase of 1.8% since 1971—and offers that "the availability of a large amount of information forthcoming from other and more highly developed civilizations might contribute to a staggering increase in energy consumption." To his credit, Kardashev does acknowledge the tentative nature of his estimates in the conclusion. Fourthly, more recent perspectives on SETI (such as the one expressed in Qualitative classification of extraterrestrial civilizations), come to very different conclusions as to the relationship between progress and energy consumption: "available energy is not an unique measure of [extraterrestrial civilizations]’ progress, it may not even correlate with the quality of use of that energy." And lastly, even if Kardashev's predictions about increased energy consumption turned out true, and even if a causal relationship was established that made an increase in energy consumption an inevitable byproduct of progress, implying—as was done above ("Human progress will come from advancing up the Kardashev Scale")—that increasing energy consumption yields progress would still be a logical fallacy. New feature proposals may be energy hungry. Whether their energy consumption is an acceptable tradeoff needs to be assessed on a case per case basis and should depend on what the feature enables. We shouldn't outright dismiss a feature because of its energy cost nor should we hold energy consumption itself as a progressist virtue. |
Thank you. |
|
@tobie wrote:
Yes, exactly this. This point should be what people walk away with when reading the Sustainability principle. |
I believe the subsequent comment, which I agree is gratuitously insulting to the work of members of this community and does not contribute to this issue discussion, remains in the thread. (And I'm not sure removing the suffix "ista" from one comment is addressing that concern either.) Thanks @hober @torgo and @tobie for repeatedly calling out disrespectful and unproductive behavior in this thread. |
This comment has been minimized.
This comment has been minimized.
@csuwildcat I personally don’t understand whether you’re targeting the EWP document, the sustainability principle, the formal objection process, the broader W3C process, a combination of the above, or something else. I continue to find your comments off topic and in violation of our community norms and I will be disengaging from this thread until this is resolved. |
|
I thought we were on a good track with discussion of a potential rewording of our sustainability issue. Unfortunately this issue has devolved into name calling. I'm calling time out on this discussion, freezing the issue, and putting a temporary ban on certain individuals. |
|
@torgo @hadleybeeman and I worked on the sustainability principle in our Hybrid face-to-face today. We're closing this issue on the basis of these changes. |
|
Sorry for the pre-emptive closing of this issue. Further discussion on the changes to the sustainability principle is welcome, and closing the issue wasn't meant to imply otherwise. If there are specific text changes or focussed related topics, we encourage new issues so that conversation is easier to follow than in a single large thread with many topics. |
|
No additional comments have come in so we're closing this. |
The specification of "new web technologies" in the Sustainability Principle is anticompetitive.
This Op-Ed argues that the TAG Ethical Web Sustainability Principle is both anticompetitive and is being weaponized in a way that inadvertently undermines every other W3C ethical web principle.
Another complaint on the anticompetitive nature of the sustainability principle was discussed here:
The
<video>tag in the HTML spec is overwhelmingly used for entertainment purposes and directly responsible for an order of magnitude (and more) more energy usage than the issue mentioned in the Op-Ed, above.Having a sustainability principle that only applies to new technologies is obviously anticompetitive. The sustainability principle protects W3C incumbents which in turn violates the Priority of Constituencies. It is concerning that the W3C would grant an ethical pass to their own existing technologies, that their businesses are built upon, while using the principle to deny emerging technologies based on a completely different sustainability standard.
While the sustainability principle has good intentions, it should either be applied to all technologies equally or removed from the list of principles to avoid the conflict of interests (or even the appearance of a conflict of interest) mentioned in the op-ed above.
The text was updated successfully, but these errors were encountered: