I made a curation bot for steem
Hello steem community!
I know what most of you guys are thinking right now, that upvote #bots are bad for #steem. Sure, simple examples like this one are probably bad.
However, @curator is a bit different. Instead of blindly upvoting an author, not matter the content, @curator tries to identify the posts that are trending, based on a few metrics (payout, votes, comments, time posted and title), and scores all the most recent posts. It then proceeds to upvote those with a score it deems 'high enough'.
Well, this is an experiment, and I'm sure I am probably not the first one attempting to reap the curation rewards this way. The bot usually takes between 8 and 12 minutes to take his decision on a post, and is usually in the earliest 10 upvoters.
You can check all the upvotes of the bot on steemd.com/@curator
Or more visually, this is how the 'hot' section of steemit looks logged in as the bot:
Do you think a human could achieve the same precision?
Do you still think bots are useless?
Actually, I believe you are the first person to do this, at least publicly. I read either one of @Dan or @Ned 's threads, or it was in the White Paper maybe even, where it said that they do discourage the use of Bots, but that a Bot that could actually add something to the site wouldn't necessarily be a bad thing. And wherever I read that, the example they gave was that a closest a Bot would probably get is to upvote people who have already gotten upvotes on posts. This seems to be much more sophisticated than that. And eventually you could maybe even make a Bot that reads. Not saying that this is going to happen tomorrow or anything, but they are already making things like a program that runs on a robot that allows it to watch 600 hours of Situational Comedy (SitCom) Television, and then be able to start guessing what will happen next. And one robot, maybe even the same one, wrote a Sci-Fi story after watching a bunch of Sci-Fi, and some semi-well known people thought it was good enough to make into a real short film and star in it. So Bots aren't always "auto-upvote everything" type things, they can actually add value.
Totally agree with you. Another good example would be how Amazon is fighting fake reviews on their website with an AI.
@Curator could probably earn its own funding so that you can afford to make it better and better, and you could be developing AI in a few years, lol.
Where i can download your bot? Is it free or for sale?
Thanks for this information. I'm interested in building a curation bot.
Would love to connect with you offline. Fascinated by this. Reply to one of my articles or ping me on reddit. Cheers.
I sent you a message on reddit few minutes ago
This looks strangely like it was built with making money in mind, not helping the website or its users.
on Steemit, that may be the same thing. It may not. The incentive structure is designed to make people want to make money.
The difference between helping the website and money making is whether this bot is open sourced or not. I don't see any link to GitHub repo in this post.
I agree, but I believe if someone were able to come up with an "easy" way of gaining power/money on the site it could abused.
This is related to suggestion algorithms used by Netflix and other companies to whom accurate prediction of content that the users like is worth a lot of money. In 2009 Netflix gave a $1 million prize to a team who won their prediction algorithm competition:
https://en.wikipedia.org/wiki/Netflix_Prize
So if a bot is successfully voting for posts that the users enjoy, then it generates value. Steem is a complex system so it's difficult to say whether bots should be allowed or not. A serious problem would be if most of the upvotes come from shitty bots, resulting in noise.
I didn't even think about it that way. That means that this bot could be used eventually in the creation of Steemit based websites. People will be able to use Bots like this to find out what content gets what kind of leverage when and how. The bot might not be able to read it, but it can be used to generate databases of Meta-Data, and that Meta-Data can be used for Websites and Apps that people can access, or pages that can show different "Top" Categories.
I forgot about this example. I like the path you're thinking on. Interesting idea!
Exactly - if the bot's making money, one could say it's adding value by helping curate good content that others end up enjoying and upvoting. The more SP your bot gets the more it could influence post/trend performance though.
where is the bot
But it's not about precision. It's essentially voting on what the first small group of people think or feel about the post. It's a money making bot, not a curating bot.
Hello, @curator, It would be great if you supported my charity event: https://steemit.com/steemit/@jennamarbles/a-new-stage-in-the-development-of-steem-now-we-can-do-more I believe that we can help people with STEEM. I will be very grateful to you for your support!
Wow. Your bot is doing great, far better than me, but on other hand I im more interseted on content and communitation. You should step him down a bit to tend for his voting power some more as if its down you can't mine much because of his low steem power. Or if u were to boost his steem power you would do even better.
Yes I am aware. I am going to transfer some funds to it at one point ;)
Im trying to get the voting power to 50% atm
Interesting I need to make my bot now thanks :)
Here is an Archive of Cryptocurrency App building Code on Github for anyone creating a Steemit app
https://steemit.com/steem/@marsresident/github-cryptocurrency-app-creation-archive