The social consensus layer: "proof of attention"

in #technology2 months ago

Alternate title: What should a curator reward?


Image by ChatGPT

About a month ago, @philhughes posted the article, Steemit: A Platform About Content, SP Delegation or Daily Diaries? | Part 1, discussing the fundamental nature of the Steem platform.

Later, @xpilar posted about that team's experiment with AI commenting in the article, The day that became a nightmare due to the use of AI to comment on posts to the users in steem.

If you haven't already, I encourage you to go read them.

Both of these articles and their reply threads are exploring the topic of purpose. What is the purpose of the Steem blockchain?

In Steem's early days, the concept of proof of brain was published to answer this question:

Steem provides a scalable blockchain protocol for publicly accessible and immutable content, along with a fast and fee-less digital token (called STEEM) which enables people to earn the currency by using their brain (what can be called “Proof-of-Brain”).

In today's article, I'm going to make the provocative argument that the "proof of brain" idea puts the brain on the wrong side of the ledger. Content creators should earn tokens for "drawing human attention", not for "using their brain". Why? Because the Steem blockchain is a blockchain play to draw value from the attention economy. Justapedia offers this definition of attention:

Attention is focused mental engagement on a particular item of information. Items come into our awareness, we attend to a particular item, and then we decide whether to act.

Justapedia also provides the following intangibles that can be provided by a supplier in the attention economy:

  • Immediacy - priority access, immediate delivery
  • Personalization - tailored just for you
  • Interpretation - support and guidance
  • Authenticity - how can you be sure it is the real thing?
  • Accessibility - wherever, whenever
  • Embodiment - books, live music
  • Patronage - "paying simply because it feels good"
  • Findability - "When there are millions of books, millions of songs, millions of films, millions of applications, millions of everything requesting our attention—and most of it free—being found is valuable."

In the end, if no one reads an article, it doesn't matter how much brain went into producing it. The value of a particular piece of content derives from the "focused mental engagement" that it receives.

I have argued before that the Steem curation system is - at heart - nothing more than a ranking system. The investor/voters spend seven days deciding on the rank of a post, relative to all of the posts that are out there, and when the voting window closes a consensus has been reached. The post gets paid according to its rank at that moment.

The question that @philhughes and @xpilar raised, however, is... "ranked according to what?"

I believe that this is the fundamental question that drives the value of the STEEM token. More attention means that more value is extracted from the attention economy.

If the ranking is random, then there's no incentive at all to post here. If the ranking is based purely on stakeholder investment, then there's nothing to distinguish STEEM from a proof of stake blockchain, and little reason for any non-investor to visit the web sites that display Steem content. If the ranking is based on some idea of "quality" that's not accompanied by reach, then growth is limited. To maximize Steem's value, the ideal ranking at payout time has to correlate with the total amount of attention that a piece of content receives.

And the attention that posts draw will be based on intangibles like the factors above: immediacy, authenticity, findability, etc...

So, the next question is, if ranking is based on attention and attention is based on intangibles, then how do we measure attention?

The naïve solution to this problem is to use a view counter. In fact, Steemit had one some years back. Unfortunately, though, this had problems. First, it could be gamed easily. Second, when payouts and views were out of alignment it was demotivational for authors. Third, a page-view is not necessarily true attention.

My thought about how to measure attention in the future is to pursue a partially formed idea called "proof of attention". A simple example is for the content producer to include questions in their post and upvote to reward answers in their replies. This basic scenario has been referenced as "proof of reading" by Steem authors in the past. Then, curators can estimate attention by reply activity. I think this is useful, but not complete.

What I'm wondering about for the long term is whether developers could build a "proof of attention" concept into their platforms, and if so, how?

Imagine that a post gets edited a few times during its lifecycle with multiple questions that appear and disappear at random times and instruct readers to respond in comments in exchange for upvotes. Then, the readers who reply are demonstrating attention on a certain post and also at a certain time.

The questions could even be embedded in image file so that the actual posts wouldn't need to be edited. For example, steemit could provide a library of image links with questions for authors to use in their posts and then just replace the image at the same link periodically to provide the "proof of attention" questions.

Another example would be posts like the word search articles that I used to do. That was actual attention, and it was easy to observe.

Obviously, and admittedly, this is not a fully formed concept, but I wanted to get the thoughts out here. If we want to maximize the token value for our investors and provide rewards to authors in proportion to their contributions, then I think it's important to understand our goals for the reward system.

What are your thoughts on the use of attention as a ranking standard and on ways that we could measure it?


Thank you for your time and attention.

As a general rule, I up-vote comments that demonstrate "proof of reading".




Steve Palmer is an IT professional with three decades of professional experience in data communications and information systems. He holds a bachelor's degree in mathematics, a master's degree in computer science, and a master's degree in information systems and technology management. He has been awarded 3 US patents.


image.png

Pixabay license, source

Reminder


Visit the /promoted page and #burnsteem25 to support the inflation-fighters who are helping to enable decentralized regulation of Steem token supply growth.

Sort:  

Very interesting, this would significantly improve the current way of distributing rewards.

As you have already written, the question is how attention can be measured in the most tamper-proof way possible. Answering questions would be one possibility, but off the top of my head I wouldn't want to do that.

Even if we could implement this technically, I suspect the major stakeholders would not be enthusiastic about it.

I'm going to make the provocative argument that the "proof of brain" idea puts the brain on the wrong side of the ledger. Content creators should earn tokens for "drawing human attention", not for "using their brain". Why? Because the Steem blockchain is a blockchain play to draw value from the attention economy.

I think the idea behind "proof of brain" is to go beyond what people do pay attention to and more thoughtfully assess underlying value, analogous to the way a philosopher might encourage someone to go beyond their initial thoughts on an issue and instead engage in "reflective equilibrium" before settling on their views.

In the end, if no one reads an article, it doesn't matter how much brain went into producing it. The value of a particular piece of content derives from the "focused mental engagement" that it receives.

Theories of value (like the Labor Theory of Value, etc.) are tricky. It might be the case that intrinsic value isn't something that things have, or if they do that may not be accessible to us, we may be limited to determining what things seem to be worth to us (similar to the way we can't directly access information about the world, we have to infer and deduce based on what we perceive). Ideally we'd have institutions that help us do things like find consensus, avoid over- and under-valuing things, etc. (but obviously that's easier said than done).

If the ranking is random, then there's no incentive at all to post here. If the ranking is based purely on stakeholder investment, then there's nothing to distinguish STEEM from a proof of stake blockchain, and little reason for any non-investor to visit the web sites that display Steem content. If the ranking is based on some idea of "quality" that's not accompanied by reach, then growth is limited. To maximize Steem's value, the ideal ranking at payout time has to correlate with the total amount of attention that a piece of content receives.

I think this is somewhat bidirectional. To the extent that the ranking is not valuable people will not pay attention to it. People for the most part already know what they do pay attention to, so reflecting that back to them is probably not too valuable. Being able to correctly tell people what would be beneficial for them to pay attention to is potentially valuable, and people might reward that with their attention.

A simple example is for the content producer to include questions in their post and upvote to reward answers in their replies.

As an aside, I wonder if generating good questions about articles is a task that the current generation of AI would be good at.

I think this is somewhat bidirectional.

Definitely true. I was thinking later that proof of attention is just the flip side of visibility as a service that I've written about before. And, of course, posts with more visibility will get more attention (all else being equal).

Theories of value (like the Labor Theory of Value, etc.) are tricky.

Yeah, I guess I was making the old argument that if a tree falls in the forest and nobody is there to hear it, then it doesn't make a sound. I think it's a solid argument with regards to a post's influence on the blockchain, though. Sure, there's probably some sort of intrinsic value, but it's mostly "potential value" until it gains someone's attention.

As an aside, I wonder if generating good questions about articles is a task that the current generation of AI would be good at.

I think this might be worth considering. I suggested something similar, here.

Instead of impersonating a human, suppose the bot commenter said something like this:

Hi, I'm an AI, and my analysis suggests that this post might be a good one for stimulating conversation. Here are some insightful topics and questions that emerge from your article. People might want to discuss these in the replies:

Append a list of suggested topics/questions

I totally agree. Proof-of-attention should be considered while deciding a post's worth. However, I have a few concerns regarding your idea...

1- Do you think readers would answer those questions? It's possible that people enthusiastically read an article, even engage but don't care to answer the questions.

2- What kind of questions there should be? Should be related to the content...

3- People can use AI for finding and answering questions and skip reading the post.

I agree on all of your points. Personally, I'm much more of a lurker than a commenter, so questions in an article wouldn't do a great job at measuring my attention. The concept definitely needs to be improved & extended. Of course, there's a whole area of research into attention tracking, so I'm sure there are tons of possibilities for web site owners that I'm not even aware of. (things come to mind like page view times, mouse movements, click tracking, etc...)

As to the type of questions, the first thought is that they should be related to the content, but there are other possibilities. For example, there could be generic questions about finding related information elsewhere on the Internet or even in other Steem posts. i.e. "Find a Steem post that relates to this one and post a reply to describe some things that the two posts have in common". I thought of some others earlier, but now I'm drawing a blank.

You're also right that people would use AI to find the questions, and try to game the system. That's another reason why I thought of embedding the questions as images. It's not impossible to read text from images, but that would make it harder.

Focusing on 'proof of attention' fits well with Steem's goal in the attention economy. Instead of just rewarding content, 'proof of attention' will encourages active engagement, which can lead to more meaningful interactions. You’ve suggested some ways to keep ranking standards and attention metrics strong. These can improve the Steemit's content curation and motivate real community participation. Also, using comment ratings and upvotes can help measure the quality of engagement. This could put more value on active participation, like thoughtful comments and replies. As a result, both content creators and readers would be rewarded for encouraging discussions and deeper involvement with posts.

Upvoted. Thank You for sending some of your rewards to @null. It will make Steem stronger.

Coin Marketplace

STEEM 0.30
TRX 0.32
JST 0.051
BTC 103090.04
ETH 3836.90
USDT 1.00
SBD 4.00