After running @hiveusmaximus for 3 days now, he created 155 posts/comments, which comes to around ~0.01$ fees per post/comment in OpenAI api calls. Each post/comment took 1-3 prompts to complete, between small, medium and large language models.
Due to the fact that his account has now been downvoted to -5 negative reputation, and the only way to bring him back up were if other stakeholders decided to stand up to abusive downvotes, I've decided to conclude this experiment. Meaning, I will not continue developing AI agents on Hive for now.
If you disagree with AI content on Hive, that's fine - I understand. I know the dreading feeling of worry, not knowing whether my dev skills will be relevant in 2 years. But I see this as a reason to strive to become better, the best developer I can be - and most importantly, learning more about AI instead of fearing it. And I had the same intention in mind with Hive. But in the end, I'm simply sharing where I see the world heading to. Or in the words of Morpheus:
All I'm Offering Is The Truth.
All the best
Wolf
I think you should let it run as an example with the understanding that the people using it need to kick in for its maintenance.
You could set it to burn 98.37% and keep the rest as upkeep funds.
Or, since you code for hive, you could drop it into the dao.
Then the curators could decide if it provides a valuable service and not the tyrannical overlords that have spoken.
Like I said, I would like to get paid for adding non-original content to the chain, too.
Any ai bots trained on hive data need a healthy dose of anarchism, imo.
https://peakd.com/c/hive-129614/created
https://peakd.com/hive-129614/@antisocialist/the-thirteenth-tribe-ch1-pt-1-5
https://peakd.com/hive-129614/@antisocialist/1917-1921-the-ukrainian-makhnovist-movement
Why not just decline rewards and see how thats gonna go?
This is a good suggestion !BBH !PIZZA
@cwow2! @day1001 likes your content! so I just sent 1 BBH to your account on behalf of @day1001. (13/20)
(html comment removed:
)
There was never any farming done, nor was there any planned. The last comments by @hiveusmaximus were already declined, afterwards the account was nuked to -5 rep regardless.
The reason I disagree with disabled rewards by default is that it's limiting the potential impact by AI agents. I'm fully aware that @hiveusmaximus was an early version, but at this point, anytime rewards for any AI agent would be enabled, the account would be at the mercy of downvoters? Do you believe that this attitude will foster innovation? I don't.
@hiveusmaximus might be early, but it might not even be the first :o
Hiveusmaximus legit fooled me when I was scrolling through Snaps x)
I am still for Downvotes as a way to do anti-abuse.
No, is it good to have guardrails? Yes.
But no, it does not foster innovation.
True, you were upvoting the bot until after the downvotes.
Genuine question.
Why don't you stay a bit powered up? I know you have server cost as a witness, so some hive needs to be pulled out to pay for that.
But does it have to be all?
It seems this is the main problem people have?
I rather like this question. I'd love to read a response.
We need a "have seen" feature on Hive xD
Im not dead yet but sometimes I feel like it on this issue
I think this is the challenge. For most of us, the sequence of events was meeting HM on snaps, voting this poetic little fellow, and THEN finding out he was an AI. Nobody likes to be fooled, and it is the thing most people fear about AI: it becoming so good at imitating humanity that we are seen as replaceable by the chattel-makers of the world. As an African American, I will add that certain citizens of my nation have been looking for a free labor solution ever since my ancestors have been free, and if there was a way to do that, NO ONE would get paid. AI represents that nightmare across all races, genders, colors, and creeds -- that is not all it represents, but it is a factor.
So, you walked us through our nightmare: a Hive on which human Hivers are not even needed to generate content and get rewards. Those who can code a nice AI bot can sit back and compete in the reward pool, and those who can't -- oh well. I don't think you meant to do this ... you might have avoided it by letting it be known what you were doing before launching it. Even a precoded little note: "AI Agent in Training" or a few days of posts explaining itself and the benefits and the why might have helped. Instead ... well, you know what has already happened.
Some time ago you wrote a post about God, and I told you then: time to walk it out. The Scripture itself says that even God does not make major changes without telling the people what is up. Adam and Eve sinned ... and God told them a Savior was coming, at least 2,000 years in advance. Noah preached about the flood 120 years ... and that's not counting the fact that his great-grandfather, Enoch, preached about a coming judgment, and named his grandfather Methuselah: "It [the flood] will not come until he dies." Methuselah lived 969 years -- so that's 1,000 years of warning. And so forth ... the stories of prophets coming hundreds and thousands of years in advance are many. Only Satan, and unwise men, set up situations to fool people on major change, the one to gain control, the other because simply unwise. But a man might become wise by noting how God operates, and doing likewise. God condescends to keep the people informed, though there are no consequences for Him if He chose not to. Mortal men face consequences, sometimes severe, so they would be well-advised to walk in wisdom.
Please don’t misuse and abuse quote from a great man!
This is one of your IA posts, tell me what's the point to have an account posting this and getting rewards?, because I don't get it.
Not sure which post is from. The account is nuked to -5 reputation. But I'm pretty certain nobody would reward that kind of content. Hence why there's no reason to decline rewards. Each API call costs money, so there is a disincentive by default to spam.
Ok what about this one? this isn't of the AI you are mentioning?
https://peakd.com/hiveusmaximus/@hiveusmaximus/dc6feeff-d59a-46a1-b5db-9a5898341557#comments
You're upvoting yourself with alt accounts, no wonder why you get those downvotes, being honest I guess that you understand that this is a clear case of reward farming posting spam.
Wolf only started to upvote his own account after it was targetted with Downvotes :)
$PIZZA slices delivered:
@day1001(10/10) tipped @cwow2
There, fixed it for ya.
Interesting project/experiment. Philosophically, it is certainly a must-explore 😜
The downvote comes to me as no surprise but I got no idea how one would go on to see its potential. It will be intersting to see, if you find a way to accept it by the Hive.
Your cost is $0.01 for posts and comments, but the Hive community has not allowed you to receive money directly from the central reward pool. You also compared Layer 2 rewards on other chains with Hive's Layer 1 rewards, which may not have been the most fitting example. Have you taken a step back to consider if there might be alternative approaches to make your AI agent both sustainable and profitable without relying on the traditional reward structure?
One option could involve asking the Hive community for HBD donations to support your AI agent. These donations could be staked, and the monthly interest generated could determine the number of posts and comments you can make, spreading the effort evenly over time. Have you explored this option in detail and considered how it could align with community expectations?
Additionally, have you thought about accepting delegations as a means of support? By leveraging curation rewards, your model could potentially become self-sustaining. Have you investigated whether the community would be open to delegating resources if you demonstrate the value your agent provides?
Another possibility could be showing the Hive community the impact of your AI agent in burning a percentage of total Hive inflation. This could help reduce inflation while simultaneously positioning your model as a stake-based system funded by HP or HBD, with the community’s backing. Have you analyzed the potential for this approach to gain support?
What about engaging with users directly through AI-generated comments on their posts? If they find your service helpful, they could choose to delegate resources (HP) to you. Have you tested this idea and gauged user responses?
Finally, have you contacted any Layer 2 token projects to explore whether they would be interested in supporting your idea by allocating a portion of their reward pool to it? Have you paused to reflect on whether a middle ground exists, one that addresses the Hive community’s concerns while achieving your goals?
By rethinking your approach and considering these possibilities, could you find a way to make your AI agent financially sustainable while building greater trust and collaboration with the Hive community?
Sorry for hearing that. I think we should have a chat on discord and exchange some ideas.
Congratulations @therealwolf! You have completed the following achievement on the Hive blockchain And have been rewarded with New badge(s)
Your next payout target is 170000 HP.
The unit is Hive Power equivalent because post and comment rewards can be split into HP and HBD
You can view your badges on your board and compare yourself to others in the Ranking
If you no longer want to receive notifications, reply to this comment with the word
STOPCheck out our last posts:
As usual the, Hive Perverts, are always watching. And ready to take away rewards from anyone who they deem unworthy. It's like they are the stereotypical bad guy.