Tokenised attention has 3 key flaws. Here’s how InfoFi can reward real value.
Tokenising attention doesn’t guarantee that InfoFi rewards value.
@shmula exposed 3 critical flaws of InfoFi today, and here are my key takeaways:
The algo prioritises attention from smart followers
In its current state, everything is based on your social graph, or who follows you.
Smart followers have more power, and this is a problem when some (not all) choose to abuse the system:
Some accounts farm engagement by claiming you’ll get Yaps + higher LB rankings if they interact with your posts.
We have higher ranks based on who we know rather than how good our content is.
Which is not ideal for InfoFi at all:
InfoFi is not a good measure of value
Yaps are tokenised attention, not value.
It’s easier to quantify attention instead of value, because it’s so subjective.
A valuable piece of content to me could be useless to someone else.
Different people will have different perceptions of a piece of content, where some may think it’s valuable while others don’t think so.
Which is probably why InfoFi currently quantifies attention instead of value.
But anyone can do anything for attention:
There are no repercussions now if you engagement farm or write everything completely with AI.
The algo feeds on drama and hype, so it’s possible to gain LB rankings by spreading wild rumours.
Which is why attention should not be use as the sole criterion:
Attention alone is not enough
Attention optimises for vanity metrics, while signal optimises for value.
This doesn’t mean that all pieces of content with high attention are bad.
It’s possible to get a lot of attention on valuable pieces of content.
But it’s also possible to get attention with non-valuable content.
Attention can be a metric, but it can’t be the only one.
There needs to be other ways to measure the impact of a piece of content beyond metrics that can be easily gamed or inflated.
I agree with the variables suggested by @shmula, and relevance likely plays a key role in determining the value of a piece of content.
I see it evolving towards this cycle:
Projects identifying who their target audience is to convert into users
Creators creating content for that specific audience
Rewards are distributed to creators who create the most relevant content that converts the target audience
The current state of InfoFi rewards noise instead of real value to anyone (projects, creators, and audience).
I shared more of my thoughts on this here:
The problem with InfoFi is noise. Here’s how it can be fixed.
Public Yapper leaderboards are killing InfoFi.