On BidBots, altruism, and power

There is a fine line between promotion and spam.

I, for one, welcome our new robot overlords

Note, I often drop that line anywhere on the internet where it is even slightly appropriate. I look at it kind of like insurance, in the event an AI decides to sort people into friend or foe at some point in the future.

I have been trying to increase my understanding lately about the workings of upvote bots on STEEM. I don't think I know everything there is to know about the subject, by any means. I have run a couple of experiments and done a little bit of investigation, however, and I have some observations. Hopefully most of them are sensible, and I'm certainly happy to discuss whatever ways I may have gotten it wrong. That's how I learn.

There are a variety of different kinds of bots running around the STEEM blockchain, and I have to say that overall I'm a fan of the activity. Some of the bots are kind of spammy, some are kind of fun, and some of them may be causing problems, depending on your view of their various behaviors. That is a general statements about all bots. For this specific post I'm going to limit the scope to just the particular strain of bots that upvote (or downvote) content.

Even with the upvote/downvote constraint, there are still a wide variety of bots automating things. By far the voting bots that seem to draw the most attention and controversy are the bid bots. If you haven't already seen it, and I'm guessing most people have, you should take a look at steembottracker. This is a tool created by @yabapmatt to keep track of the action in the bidbot marketplace. It is fascinating and in some sense mesmerizing to watch and study the changes.

Sure it tastes good, but is it good for you?

In one very real sense, the bidbots are a free market in action. Even if this were not an openly functioning market, there would likely still be a black market in selling and buying of upvotes. This system makes it more efficient. There do seem to be a number of concerns that people have with bidbots generally. The loudest of these concerns boil down to a couple of ideas. First, spammers and opportunistic shitposters can and do use the bots to make a mess out of the trending page. Second, that this activity unfairly allocates rewards, also known as raping the reward pool.

Clean Trending

There are a number of efforts underway (or which have been battled about in the recent past) to combat the spammy behavior of people abusing bidbots to promote low quality content. It's a long list. There is plenty that has already been said about this, and I'm not going to attempt to recap it all here. One thing that I will say about trending, and the view that it presents to the world at large, seems to mostly be a UI display issue. If the trending page did not give credit to purchased upvotes, that part of the argument would vanish. A significantly weird thing to try to explain to anyone who isn't intimately familiar with the system and looks at steemit is the earned value story. "Yes, that post got $432.12 in rewards, but it cost $600 to get them" is a variant of something that I find myself saying, and it doesn't make any sense. Maybe it's a good thing that we present to the world a completely false sense of earnings and the effort required to succeed on this platform. Probably not.

Don't make me blow my whistle

The reward allocation issue is also a reasonably big deal. I'm not the guy to describe it all. It's an ongoing drama and one of the trickiest design problems with DPOS generally. Stake rules on this playground, it is not a democracy. There's a fair amount of game theory in play with this problem, since the overriding best interest of all STEEM stakeholders is an increase in the value of the system which will drive long term price appreciation. That general goal is true for the community as a whole, but on an individual basis some people may decide that it is in their best interest to be as greedy as possible and let everyone else worry about growing the system.

That, in a nutshell, is how the system works. Whales run the show. They directly influence behavior, and what they want to have happen usually happens. The only thing preventing or allowing a thing to transpire on STEEM is the community norms and the behaviors propigated by and among large stakeholders. I am very encouraged by the wealth of talent that has assembled in this community, and the enlightened self-interest that many people display in helping to build an ecosystem that drives exceptional long-term value. We need more of this, and also more of the attitudes and efforts that prevent people from grabbing everything they can get their hands on for the least amount of work.

An Experiment

With those thoughts as context for my recent activities, as I said at the beginning, I'm learning some of the ins and outs of upvote bots. Without passing judgement on whether we should have them or the ways in which they are good or bad, I've taken some notice of the idea that "good" content creators should be able to outcompete the "bad" creators and use the tools that exist to their advantage. Beat the spammers at their own game, as it were. I think this is not possible without help, because by its very nature creating useful content takes much more time and effort than creating crap. This is a huge part the efforts to clean up trending and fight abuse - there are a number of strategies in play. Blacklisting abusers, whitelisting honest creators, ruthlessly flagging bad actors, all of these things help, but they're also only as strong as the weakest link. I have to say that it's still a pretty difficult game to win, although the bad actor route keeps getting more difficult, as it should.

Now I'll describe my latest adventure in a bit more detail. It's a bit of an unusual use case, and just one of the many possible ways the bots and automation can be used. I tend to write really long things that take a while to create. Kind of the anti-shitpost. As a result, and with various other demands on my time, I only put out a post or two a week depending on what's going on in my life. I do have a number of people who support my efforts, for which I am super grateful. No huge payoff, but it's an encouraging affirmation. As part of my odyssey in learning how things work, I've got a bit of a trickle of STEEM coming in. I could get into all of the details, but suffice it to say that I'm mining some crypto on a very small scale and converting it to STEEM. That is yet another learning experiment. The cool thing about this is that it's a (very small) direct benefit to the ecosystem. I'm buying STEEM on the open market and powering it up.

Originally, I just directly powered up the STEEM I deposited. What I decided to try in the experiment was "indirectly" powering up by purchasing bot votes. The way I looked at it was this:
- I get to learn a bit about the bots and how that market functions
- It benefits the bot owners, since they get the STEEM
- I still power up, so I'm in as good (or better) position. It just takes a week. I can wait a week.
- Saved the best for last. All of the people who support me by curating/upvoting my stuff get a benefit.

That last point is significant. The curation model is quite a complex puzzle, of which I understand a lot but not everything. The key point being frontrunners benefit from large votes later in the cycle. So, what I've done is buy some votes on what I consider to be worthwhile posts. Hopefully this doesn't make me a bad person. I think everyone wins, in this scenario. STEEM as a whole benefits from the purchase of STEEM on the open market. The bot owners get some return on their stake, and their stake grows. I get a bit more STEEM than I bought. The curators get a bit more SP than they otherwise would have. Obviously, this is not the only way in which the bots can be used, but I thought it was honest enough that I was willing to give it a spin.

Market Efficiency

One of the most important things I learned by doing this, was just how ineffecient the current system is. The idea of an auction seems to make sense, until you deal with the way it works as a practical matter. With a normal auction, you place a bid and if you get outbid, you don't pay. With a vote buying auction on STEEM, you pay and hope nobody else bids. Theoretically this will work out because nobody with any sense would bid and make the payout return negative. Let me tell you, they do.

There are a lot of variables, and I could keep on typing and describe them at great length, but suffice it to say that there are inefficiencies on both sides of the market. For both the sellers and the buyers of upvotes, the system has some drawbacks. For now, I will just leave it at that. There are some initiatives underway that look to be addressing this problem, whether the @bid.bot system, or @smartsteem, it seems to be a goal to remove some of the confusion and uncertainty from the process. I'm sure there are others. I would love to hear what people think about this, or at least I think I would. Please be civil :)

footnote: as per usual, images are CC0 from pixabay.com