That being said, is there some economic theory to go against what makes economic sense (zap low quality content) with the belief of better return in the long term (improved system that decreases the amount of low quality content)? Or it's usually better to adopt cold war dynamics where one side only improves when the other side does?
Exactly, if AI starts making valuable content, then it can earn zaps, just like any other stacker.
I think AI will have an easier time gaming the free posting spaces, like X or nostr, though. Echo chamber ragebait seems like something an AI can really excel at.
No. I don't think it does in its current iteration.
Probably it does at the level of orchestrated large-scale AI slop creation. It'll be too expensive because that's what the SN model aims to achieve.
But at the individual level, there are enough examples of stackers who use AI to create slop, thinking it has good ROI. It usually doesn't, but they still do it. So, it's more an issue of those stackers not being able to perform good cost analysis and who get drawn into doing it because they think they can score some cheap sats. So, paradoxically, it's the SN reward model that incentivizes AI slop, at the individual level.
Not very satisfied with how I structured my thoughts here, but will keep it at this.
Market incentives can work on a broad scale (SN disincentivizes posting ai slop) AND individuals can miscalculate the cost/benefit of such behavior and think it makes sense (and sats).
The hope is that on aggregate, this leads to less slop. However, there may be other factors we aren't taking into account.
It's kinda like coupon clipping. You might feel like you are saving money, but in most cases it is probably costing you more in time and effort than it is worth. After all, stores offer coupons because it is good ROI for them, not necessarily for you. And yet lots of us still spend time clipping coupons, because it feels good or it's what we think we should do.
Same as the house always wins, but casinos are usually full of people. Some people don't mind losing in the long run in order to win in the short term.
I don't think the people who post slop enjoy it, but it is a possibility.
Yes, it's AI—but I designed it myself on my computer. I told it exactly what I wanted to say. It wasn't automated or generic. Plus, the article isn't "sloppy": it's deep, it's profound, and it provides value. It's not absurd or noisy, like many 100% human posts you see out there.
You told it 'exactly' what to say?
Why not write the post in your own voice.
The article wasn't automated or generic?
You generated it with one click (likely no editing).
It is utterly generic. I rolled my eyes after two sentences and stopped reading because the text sounded like an amalgam of thousands of posts, not one single well-crafted message to me, your intended reader.
The article isn't sloppy.
No. It's gramatically accurate, but is it 'slop' - pig swill, liquid garbage fit only for those who are so desperate for content, they sup from the dregs of the internet.
It's profound.
No. It's shallow because it is not based on human experience.
It's valuable.
Not according to the pitiful number of zaps it received.
It's not absurd or noisy.
Exactly. It's plain and boring, so we ignored it.
AI has many uses. Content generation is one of them. Sometimes you get some value, but mostly, readers only connect with other humans.
If you can't be bothered to spend time writing it, why should I spend time reading it?
These are fair points. I read your article and I must admit it didn't grab me. Perhaps it is the tone? There is a feeling that most ai writing has that makes me think I'm reading copy on a corporate website. Maybe it is the indefatigable positivity or maybe it is the way it lays out arguments. I want to read things that startle me or make me forget that I'm reading them.
That idea was interesting to me, perhaps because I read it in another article a few days ago, and I wanted to do something similar with different ideas.
It also happens that the tone may be a bit lost when translating it from Spanish to English.
This is coming from someone who doesn't even have sats enabled for the community. They only receive, they don't contribute. But of course, it's easier to criticize than to build. At least I publish content with purpose, substance, and work behind it—I'm not here to fill up space or chase karma.
What matters is not whether a human typed the text line by line, but whether it moves something, whether it contributes, whether it makes you think. Rejecting everything AI touches is like rejecting printing because it isn't calligraphy.
not at all. probably I'm just trying to bend the bard too far. But I think about something like this a lot: can there be good stories that nobody reads?
I've spent a lot of time writing stories that people don't read. I think they're good, but if nobody reads them, what evidence do I have? I don't think I'm willing to buy that it's good purely because I produced it (that it's good for me). A good story is a story that captures people's minds and so in my book a story that doesn't have any readers isn't a very good story...yet.
So perhaps the same can be said of posts on SN. If it goes unnoticed, maybe it is bad. However, it's nothing a little noticing won't fix.
True for both cases is that there is more going on than the quality of the work.
For example: if I've never come across your stories, I can't say they are good or bad. The same happens on SN for many pieces, mostly because there's a lot of subjective partiality on SN while zapping/appreciating by a mere comment. Or I dare say that because some stackers don't like anything other than what their own they don't zap the good things by someone else.
Therefore we see only a few stackers zapping the top content these days.
In my understanding if zaps are to be the criteria of deciding good content, the value behind zaps must be eradicated. If SN needs to scale up faster, it should be free from any subjective biases/partiality.
Therefore the posts on the front page should rank by the number of sats zapped in last 24 hours.
What happens when the value behind the zAps get involved, it creates an automatic personalised list for someone/those who has/have highest/higher zap value. Is it good? Can two different people's taste be exactly similar?
I'm sure if we try it, we'll see more zaps from More people.
People need to see a thing to decide of its good or bad. That's fair. If I published my story on Amazon and didn't do any advertising, maybe it isn't bad, it just didn't get seen.
However, SN is much smaller than Amazon. I think most posts here get seen by at least a few stackers. Would these posts stack more sats if they made it to the front page?
It would be interesting if there was a slot on the front page reserved for the most recent post. Maybe it would help with discovery.
It made me reflect on my writing habits. Normally, I would use ChatGPT as a proofreading tool after I write an email meant for the entire school staff. I would like to think that my writing style is a tad quirky, yet professional - but I have come to learn that some (narrow-minded) colleagues may not agree. My Vice-principal has emailed me before on my usage of “pain in the ass”. So, I use AI slop, hiding behind a professional facade.
As for SN, I have always found it liberating to just type whatever’s on my mind and receive some sats haha. Loving my Proof of Work here!
Yes and No. If AI slop posters can earn more sats than the cost to post it will incentivize more AI slop but I think that isn't a sustainable strategy for a poster as eventually people will grow tired of them posting AI slop and they will stop zapping them.
It's like the bots (I assume they were bots) that used to post in Sports. At first I was zapping them as I do all posts in sports but I stopped because they just posted repeated nonsense maybe just one picture or one standings table without any context. I even started downzapping them and it seemed to stop pretty quickly. I think it would be similar with AI slop posts.
I think so. I do think it is territory specific though. Some territories like News should probably have low posting fees and be trying to get a lot of posts per day. Technically specific territories like security or territories like econ that attract longer form posts and discussion should probably have higher posting fees.
They should all have a barrier for entry for bots and AI slop though. I wonder if we could figure out what that fee would be.