One of the unique aspects about the Steem blockchain is its “Proof of Brain” based rewards mechanic concept, in which the community decides what content, creations are valuable and deserve to be rewarded.

As the blockchain is in its third year, and more blockchain based platforms with similar approach have appeared already or will soon be launched, the recent dip in the crypto market is a good time to consider whether Steem’s Proof of Brain has been a success so far.

<center>![651C552E-4A5E-437B-A484-F2EF683A03B4.png](https://cdn.steemitimages.com/DQmVzAtcZMtqNtXzow9hv94B12KyueU851baQj5ESpm68ZC/651C552E-4A5E-437B-A484-F2EF683A03B4.png)<br><em>Yet another Faketoshi lacking proof of brain</em></center>

That even more because in recent months STEEM, the token, has been a constant poor performer when comparing to other tokens in CMC’s Top 50 given that the Steem blockchain has one of the most active user bases, if not the most active.

Recently @crokkon published [a post](https://steemit.com/utopian-io/@crokkon/what-is-the-reward-pool-share-of-vote-sellers-bid-bots-and-autovotes-1535479520399) in which they analyzed how many of the daily upvotes were done by bots. To do so they checked the transaction signature of each upvote, using the [beem.py](https://github.com/holgern/beem) library.

In their analysis, more than 50% of the upvotes during the analysed period were done by automated services and/or bidbots/promotion bots. This equated to around 40% of the rewards distribution in that period. Steemauto was the leading service, by lightyears, with more than 40% of the votes (but less than 7% of the rewards). ~~Bidbots~~ Promotion services distributed around 33% of the rewards in that period with around 8% of the upvotes. 

<center>![95D48646-06E5-4E4D-91B4-44B8684F2FF1.png](https://cdn.steemitimages.com/DQmfMRsphQ4eCtTEoD4inZzux7E3EnB5zwS3upS8TRTsQLN/95D48646-06E5-4E4D-91B4-44B8684F2FF1.png)<br><em>Graph from @crokkon’s [post](https://steemit.com/utopian-io/@crokkon/what-is-the-reward-pool-share-of-vote-sellers-bid-bots-and-autovotes-1535479520399)</em></center>

You can check the actual numbers [here](https://steemit.com/utopian-io/@crokkon/what-is-the-reward-pool-share-of-vote-sellers-bid-bots-and-autovotes-1535479520399).

While interestingly enough Steem’s whitepaper doesn’t mention **Proof-of-brain** but instead refers to **Subjective proof of work**, the bluepaper opens pretty much immediately with Proof-of-Brain and the SMT whitepaper links to the bluepaper also pretty much immediately to explain Proof-of-brain.


## Have Automated Services Hurt The Value of $STEEM?

Given that both complexity and scarcity are generally considered elements of value, I have to wonder whether the lack of especially complexity has hurt the value of the $STEEM token. If one merely needs to join a curation project and end up on its trail’s autovoter, or pay for an automated not quality curated upvote... where’s the complexity? 

Where’s the scarcity?

<center>![16E9C5FD-FF93-4EFC-B7A6-E4CBA83DAD70.png](https://cdn.steemitimages.com/DQmcnxuwfcDRi7QX53Gkfs1mF5FyEvNFr74LL335gUR9amQ/16E9C5FD-FF93-4EFC-B7A6-E4CBA83DAD70.png)</center>

Have we so far built a Crypto-Keynesian paradise on the Steem blockchain? Have we treated the Steem blockchain as a Keynesian lunacy and hoped that the numbers and activity would de facto lead to value?

> <em>... makes clear that Crypto-Keynesians can’t tell the difference. This leads to the building products that add “activity” but fail in delivering any sort of value, in effect digging and filling ditches and pretending that means something. This is Crypto-Keynesian lunacy.</em>
> — [Jimmy Song](https://medium.com/@jimmysong), Bitcoin Educator, Developer and Entrepreneur

**Has automation killed (Steem’s) Proof of Brain and ruined the value of the token?** 

Chime in in the comments.

<hr>

#### Notes

This post is not meant to be yet another emotional “bots are bad/good” thread. Bots, and automation, have definitely contributed to distribution. After more than two years though, we have to ask ourselves the hard questions. No matter whether we like them or not. Without asking those questions we will not be able to design better and improved systems.

I am fully aware that token values depend on more than just one element. Just as they depend NOT ONLY on BTC’s dominance, especially not during dips and their recovery periods. If that were the case $STEEM would still be near the top 20. 

Yet, first impressions matter and complexity, or (subjective) proof of work AKA proof of brain may matter in those.

Yes, there are many more factors to be discussed but before we can possibly discuss “curation rewards”, it is primordial to actually look at what’s going on and assess what’s happening rather than potentially feeding automation even more without improving the core concepts and issues first. 

<center>![B88DA4DD-8235-4190-8C21-0820FE1A763F.jpeg](https://cdn.steemitimages.com/DQmTzDZ6kwk6uikKxdDz2MUF3PJFUDjfehKZUJvCwEUgb4B/B88DA4DD-8235-4190-8C21-0820FE1A763F.jpeg)</center>

To put the learn in fail where required. To #flearn.