Great post Kevin, no surprise that I entirely agree as most of this has been covered from our private discussions

I just want to dispel a few common misconceptions that always seem to distort the conversation

• MYTH - We have a problem of greed

Very few investors are in crypto to exercise altruism alone. Any successful economic system must assume all players wish to maximize their returns, then align behavior that add the greatest value with the highest rewards. Linear fails at this as it leads to content agnostic behavior gaining the highest rewards.

• MYTH - This is a problem of culture

We had a relatively functional voting economy around this time last year, and despite this, we find ourselves in a complete failure of a content discovery platform today. Broken Window Theory - If we failed at changing the culture for the better when it was much easier to do so because people back then were generally behaving better, what hope do we have of doing it now that vote rewards don't track content quality at all? This isn't a problem of culture, it's purely about misaligned economic incentives.

• MYTH - This problem is inherent to stake based voting

No it is not. Under certain conditions such as superlinear, the value of your vote can vary heavily depending on the popularity of the post on which you're voting. This means you can easily earn more curating potentially good content than voting your own posts up from 0. A combination of superlinear, additional downvote power and higher curation, as suggested, can likely allow content reflective voting behavior to out compete content indifferent behavior (eg. vote selling/farming).

• Vote farming/Self voting don't provide the highest returns anyway - Irrelevant

You may be talking about tracking bid bot payments and/or known high self voters and getting your vote in beforehand to extract some fat curation rewards. Firstly, larger whales would not be able to do this due to the size of their votes, and therefore, the bigger votes on this platform which determine overall rewards and exposure are still going to be content indifferent. Secondly, once the market matures enough, it's trivially easy to avoid - larger voters can just spam multiple posts (hundreds perhaps) and only vote up the ones after 15 mins with the least curation 'stolen'.

• SMTs and Good Person Tokens will solve this - Problematic View

Maybe. Don't get me wrong, I think SMTs are great. But they're 6 months away and in reality, it'll take far longer for any of them to garner sufficient market confidence to really play a role in the content discovery process. The use of Oracles also impose very high practical cost and a lot can go wrong in reality. Basically they're very far off and a lot needs to go right for them to work effectively. Something like n^1.3, 10% free downvotes and 50% curation is just a lot more direct and simpler and easier to implement. Ultimately we need the Steem base token to have a functional reward distribution too, not just SMTs, which are a big If.

• Superlinear/Higher Curation/Stronger Downvotes are bad - BUT NOT AS BAD AS A FAILED PLATFORM

Yes all three of these suggestions have a non trivial cost. But compared to a completely dysfunctional content discovery and rewards system and they are very much worth exploring. The idea is to use just enough of these measures to tilt the scales of profit maximization behavior from content agnostic to content reflective and leaving behind as much as we can to incentivize content creators.