Ask SN: how should we disperse earned sats?

93 sats \ 21 comments \ @k00b 7 Mar 2022 bitcoin

how should we disperse the sats we earn?

When ~jobs is decently subscribed we'll be generating some sat revenue. It'll be generating revenue because of the presence of stackers, so we want to distribute tthose sats back to stackers on a daily basis. I've been calling this yield.

These are my favorite two ways that we could do this:

based on valuable contributions that day. We use web of trust to score a user's contributions every day. If more trusted users appreciate your content that day, you'll get a bigger slice of the day's yield.
based on the web of trust only. @kr has said this is what he'd prefer. Basically we'd distribute the sats proportional to a user's trust regardless of their contribution on that particular day. This means that so long as you had a valuable contribution in the recent past, you'll earn some yield every day.

Have an alternative model? Propose it. Have an opinion on which of the above is better? Please say so and why.

other boring release stuff

I just shipped some ~jobs refinements per @kr and @cointastical's feedback. The navbar is more consistent with the main 'sub', job bids are specified as sats/min, and the jobs form has fields for a company name and location/remote.

plans between now and yield dispersal

reworking the navbar, particularly on mobile ... many users have complained they can't figure out how to post on mobile
the most important and/or low hanging fruit from the gh issue backlog

view all related items

37 sats \ 2 replies \ @earnbitcoin 8 Mar 2022

Show me the incentive, and I will show you a bot that exploits it. ~ Charlie Munger

That will be the biggest challenge with this.

10 sats \ 1 reply \ @zuspotirko 8 Mar 2022

Oh, that would be insane. But I can't imagine how that would work.

You mean someone programming a bot that reads posts and comments and AI-generates a comment/post that the AI thinks is going to be popular to get upvotes? That would be an insane technical feat. And also kinda not against the goal of generating more and more high quality content.

46 sats \ 0 replies \ @CheezeGrater 8 Mar 2022

Remember that "optimized for attention/upvotes" does not necessarily equal high-quality content.

This is already the issue with current social media algorithms.

26 sats \ 8 replies \ @gmd 8 Mar 2022

I might have missed the memo somewhere, but why not just keep the sats to fund SN?

I'm having a pretty fun time dispersing sats to posters and commenters. It seems like quality posters in your web of trust are already getting paid in sats which is already pretty cool and 10000x better than reddit karma.

31 sats \ 7 replies \ @kr 8 Mar 2022

Disbursing sats to users is going to be a really potent tool for growing the platform.

Imagine if Facebook, Twitter, Airbnb, Uber, or other Web 2 platforms could have given their users a direct cut of revenue generated by the platform, in proportion to the contributions of those users. Total game changer, and it disproportionately attracts users who want to add actual value to the platform (since you don't get paid if your reputation is 0).

Before Lightning it wasn't possible to reward users like this, but now that we can, I think that's the fair thing to do. After all, the only reason companies will pay to use the job board is because they know there is a strong community of Stacker News users for them to market their jobs to. Without users, the job board won't earn anything.

I think Stacker News will eventually take a cut of the revenue to fund the business, but as far as growth tactics and marketing pitches go, it's hard to beat real-time, direct payments to users.

Using revenue from the jobs board to pay for traditional growth/influencer ad campaigns would be far less effective IMO.

Plus, an added benefit of using only revenue for direct payments is that Stacker News never has to pay out of pocket for growth, and all Stacker News users are now incentivized to get employers to post new jobs and create a thriving job board.

0 sats \ 6 replies \ @gmd 8 Mar 2022

I get what you're saying. I just think the primary magic of SN is that we're sending tiny fractions of the hardest money ever made around instantaneously with our upvotes and tips. So theoretically you're already getting paid quite generously, directly proportionate to the quality of your contributions (imagine when upvotes reach near-reddit levels of traffic).

Let's say SN ~jobs grows widely successful, ala LevelsIO, and brings in about $1M ARR (really wanted to use the tilde there haha). Spread out among the zillion future users it seems a negligible amount of yield per user but an extremely meaningful amount towards keeping SN afloat, funding new feature development, potentially paying mods to get rid of spam, etc.

I guess besides youtube I'm struggling to think of other Web2 platforms that successfully leverage paying users/creators to the extent where that becomes the motivating factor to contribute vs the inherent joy of using the platform.

Less sexy, but a more traditional and perhaps more effective way to incentivize users to find employers would be a direct referral program- have the employer input the username of the stacker that referred you. That person gets a large cut of the sats fee (even up to 50%?)... although gaming this with a 2nd account to self-refer might be an issue.

36 sats \ 5 replies \ @kr 8 Mar 2022

The job revenue will seriously outpace the tip revenue for most users, but more importantly it will also enable users to be more generous with their tips.

In February there were 215 active registered Stacker News users. A single job posting for a month at the minimum spending amount generates 2,500 sats for the average user.

Power users could get 25,000, and occasional users might only get 250, all dependent on their trust score. Either way, the sats earned by jobs will be way higher for existing users to start.

In the entire history of Stacker News, there have only been 8 users who have earned more than 25,000 sats in the last 7 months, 6 of which were AMA guests and one of which is the founder of the site.

Now imagine 10 job postings, all of a sudden some users are earning 250,000 sats/month. That's a major incentive for more users to join in, which in turn generates more jobs. It's a vicious positive feedback loop. Jobs also won't be the only revenue source to fund users with either. Revenue could also come from a Product Hunt style sub, or even more general classified/commerce subs.

Impossible to say where the average earnings of a user will end up over time, but I'd bet the average user will earn far more from revenue than through tips at first, and then start to earn more from tips once all the new sats generated by revenue are floating around in the hands of every user.

One other way to back into an estimate is to use the total number of sats sent on Stacker News. In February, users sent a total of 262,503 sats between each other. Once again, 10 job postings for a month would distribute 5,000,000+ sats, 20x more.

The amazing second order effect of paying users is that Stacker News will be able to get more sats in the hands of more people. That's the hardest part for app developers trying to onboard new Lightning users today.

When every user gets an automatic daily injection of sats from Stacker News revenue, they will be more willing to tip other posters, and accelerate the magic you're speaking of.

The reason it's hard to find examples of sites paying their users is because it's been essentially impossible until Bitcoin & Lightning. I designed a similar feature into the app of a Bitcoin company in 2019, and they've picked up 500,000+ users exclusively by paying small amounts of Bitcoin to daily active users.

It entirely obsoleted all their other traditional marketing channels (Reddit, Facebook, Twitter, Google, influencers, etc...), not a single one of those channels could acquire a new user for less money.

57 sats \ 0 replies \ @cointastical 8 Mar 2022

The job revenue will seriously outpace the tip revenue for most users, but more importantly it will also enable users to be more generous with their tips.

Exactly. When my balance is flush because I earned a chunk from a post or whatever, I am more liberal with my upvotes and additional tips.

The wealth effect is real, even within the micro-economy of Stacker.news

31 sats \ 3 replies \ @gmd 8 Mar 2022

This is certainly a fun thought experiment.

To be honest I share @earnbitcoin's concerns about the unknown implications of incentivizing and paying for users to be here. Traditional marketing channels might be more expensive in terms of user acquisition but you might get users who inherently want to be there, rather than participate for the bitcoin yield.

Millions of people post on twitter and reddit for likes and worthless karma. SN is leaps and bounds ahead by virtue of sat-votes alone. I feel like in terms of 80-20 this is enough of a magical experience that I worry incentivizing further may have negative repercussions.

It seems we've always been able to pay users for content it's just never been able to happen so elegantly programmatically before lightning. But I think it's failed because people post and share on platforms by intrinsic motivation. Tiktok is the hardest platform to monetize but influencers go there because the user experience and audience reach is so amazing.

Isn't there research that suggests when you pay kids for chores/grades etc it kills their intrinsic motivation? There's a lot of sats to distribute today with low user numbers but I anticipate when job board income 10x's, the user count will have 1000x'd and there will be less to go around. What happens when users who enjoyed getting X sats/m start getting X/2-X/3 sats/m, etc? On a personal level I feel like I've begun developing irrational, negative feelings towards my fold card because the rewards, while still objectively great, aren't nearly as generous as they were a year ago.

I probably have less bitcoin than everyone else on this forum but I just think it's such a magical experience throwing sats around- I feel like that's more than enough of a killer draw. You send some, you stack some, and it all works out in the end. I worry what might happen when less scrupulous users start treating SN as an opportunity to stack sats rather than a fun community to give back to. I know censorship is a big priority here but what happens when speech starts to blend into spam?

I would argue the money and dev time should go towards paying StackerNews to fund and accelerate more magical feature development (subs etc) and that would bring in more quality users in the long-term.

31 sats \ 1 reply \ @kr 8 Mar 2022

It's definitely a challenge to pay users "correctly" without a KYC process in place to stop bots from creating many accounts.

So far I think the Web of Trust is the most elegant way to solve this problem, but it's still TBD how well this particular model of ranking users by reputation holds up when sats are at stake. I suspect there will be continual tweaking to the ranking algorithm as we learn more about how it works in practice.

However, this is the reality of operating on the frontier and building new tech.

There are enormous opportunities for builders to try new and innovative ideas (and create/capture a ton of value in the process), but not enough precedent to be sure how well they'll work, or which specific combination of ideas will succeed.

Since Lightning is a largely unproven payment stack, it's possible that many of its best use cases haven't even been discovered yet. This is exactly the time to test new, innovative ideas and quickly iterate until something sticks.

We've always got more traditional Web 2 models to fall back on if the new ones don't work, but I think we need to have the courage to test and experiment while the business is nimble, and the big opportunities remain to be discovered.

42 sats \ 0 replies \ @k00b OP 9 Mar 2022

I'm with @kr here. This feature is not permanent so we shouldn't be overly concerned with hypotheticals. We can cook up a ton of hypotheticals as to why it's a bad idea to make upvotes cost and reward users with sats too.

It's something SN can do that's new and different from existing platforms which is enough to make it worth exploring.

26 sats \ 0 replies \ @earnbitcoin 9 Mar 2022

On a personal level I feel like I've begun developing irrational, negative feelings towards my fold card because the rewards, while still objectively great, aren't nearly as generous as they were a year ago.

Similar experience with someone whose Youtube channel I was following. When they were monetized, they couldn't praise Google enough. Since they've gotten their third strike and are now on Rumble, they bad-mouth (albeit deservedly) Youtube / Google at every opportunity.

Moral of that story: Don't mess with someone's source of income.

15 sats \ 1 reply \ @cointastical 8 Mar 2022

A couple thoughts ...

Require user to claim

This "yield" should require participation. e.g., if I earned yield today, I should have to at least have to click a "claim" or something to get it. And, my claim can expire if some days pass and I didn't do the claim. This would also let the yield to those who don't care about the yield (at least not enough to come back to SN, or be bothered to click) to be re-added back to the pool of funds for distribution at a later date. That a claim on yield is available to me could be communicated simply as a notification.

Restriction on immediate withdrawal

Could (or should ?) there be restrictions on these funds? e.g., the yield earned can be used within SN immediately for posting, commenting, or upvoting, but to withdraw those funds would need to age 30 days (or something like that).

Then again, I suppose this is easily gamed with a second account, where I tip myself in a different account and my sats can be immediately withdrawn there.

A workaround I suppose is to make the yield sats be a different symbol in my wallet. And when you pay to post, comment, upvote or tip, you are spending from the yield sats first if you have any. And the recipient of the tip or sats stacked sees those appear in the wallet under the yield sats balance, until those sats have aged the 30 days (of issuance, not when the earning (stacking) occurred).

But that makes it possible that yield can circulate within SN, so that it doesn't get withdrawn for 30 days after issued (initially claimed). Maybe this adds too much complexity, and can be deferred until there's some data to show whether or not yield is simply being withdrawn from the site shortly after issuance occurs.

10 sats \ 0 replies \ @k00b OP 9 Mar 2022

These are nice downstream options. We'll definitely ship without these but they make hypothetical sense at least.

Restriction on immediate withdrawal

This is very tricky. We'd basically need to create some non-fungible sats for this to work. I suspect this is more complicated than it's worth ... I've been considering something like this for a long time (giving each user some sats when they sign up) and haven't yet thought of a reasonable design.

31 sats \ 0 replies \ @zuspotirko 8 Mar 2022

It's really nice of you to want to give back to the community!

I personally feel like you built this and you also deserve to profit from it. I wouldn't mind if it was spent on development for SN or somehow (ads?) to gather more users and grow the platform.

59 sats \ 2 replies \ @ThrillerX_ 8 Mar 2022

It would be interesting to use the ~job board sat fund as a bounty for different development features on SN. We could vote on features stackers want to see most. Could be a Stacker Labs of sorts…just an idea.

5 sats \ 1 reply \ @gmd 8 Mar 2022

I like this idea! Would be a meta way to for SN to eat its own dog food for sub development, poll features etc.

21 sats \ 0 replies \ @ThrillerX_ 8 Mar 2022

I agree!

21 sats \ 2 replies \ @kr 8 Mar 2022

Nice updates to the nav bar, job posting format, and the messaging/UX for job posters.

Can we get the ~ removed too? Seems like many people dislike it and I haven't found anyone that supports it: #13487

Also, really excited to see sats disbursed to the community, wanted to clarify my thoughts on the Web of Trust approach a bit more.

Quick background for new users... The Web of Trust is the key that allows us to give back sats to users.

By assigning each user a reputation score that starts at 0, grows when they share content that the community appreciates, and drops when they are inactive or share content the community doesn't like, we can use it as a good proxy for the reputation and influence of a particular user, and as a tool to reward them financially for their efforts.

Without the Web of Trust, we're forced to rely on either upvotes or sats to determine valuable contributors, and both those metrics are easy for bots to game.

If we design the Web of Trust right, it should imitate the way one's trust might shift over time within a local community.

Members of your community will learn to trust you when you help them, and learn to avoid you when you hurt them or are wasting their time. You'll also lose some level of trust if you skip town for a couple years and later decide to return.

Now, for the first time we can use Lightning payments to financially reward Stacker News users with actual Stacker News revenue in proportion to their trust/reputation/influence in this community.

When thinking about how to reward users, my bias is towards using the Web of Trust exclusively to reward users as they prove to be long-term, valuable contributors to Stacker News, rather than instantly rewarding users for viral stories to stack some quick sats.

First, I think most posts that end up on the top of the page are already big stories outside of Stacker News, and are already well known and widely shared across the internet. Not sure it makes sense to specifically reward these posts.

I feel that the strength of the Stacker News community comes from sharing content that isn't already being shared everywhere, and that people consistently posting thoughtful, valuable ideas should be rewarded instead.

Secondly, since the two options above both rely on a Web of Trust, the difference is just in how quickly the sats are distributed. My bias is towards a longer reward period than a shorter one.

In Option 1, let's say a new user finds a great story and receives upvotes from trusted users. As a reward, maybe they get 1% of Stacker News revenue that day, or 10,000 sats for example.

In Option 2, that same new user (who starts off with zero reputation) won't get the 1% of revenue, but will see their reputation sharply rise (since users with high reputations signalled their approval of the content).

As a result, the user in Option 2 will begin receiving a consistent stream of payments from Stacker News in proportion to their reputation. Let's say that's 100 sats/day, with a 200 day linear decay if they never return back to Stacker News.

In both cases, the new user earns 10,000 sats for their contribution, but in Option 1, that user gets all their income up front and has no incentive to continue engaging with the community until they find another juicy headline story.

In Option 2, that new user gets their money over time, and can actually change the value of their daily payouts by continuing to engage.

They now have a reputation to take care of, and their 100 sats/day could become 200, 300, 400 or more if they choose to keep contributing valuable content. If not, those earnings would decay down to 90, 80, 70, or fewer sats/day.

The difference I see between the two approaches is that the second one distributes money over longer periods of time, and does a better job of incentivizing consistent contributions from long-term users.

21 sats \ 1 reply \ @k00b OP 8 Mar 2022

I'll remove the tilde from the navbar when I ship the new navbar. 👍

0 sats \ 0 replies \ @kr 8 Mar 2022

🙏🏻