Brothers, let me lay it out straight about a liquidation lesson. Last night, while Bitcoin was in that volatile range, I got greedy, opened a long position on 2 BTC hoping for a bounce, and bam! A needle drop hit my stop loss in the early hours. After getting liquidated, I actually calmed down and went back to review the RLHF feedback records for @OpenLedger . You know what I found out? Those making money by scoring models are getting way less than what the whitepaper promised.
The whitepaper sounds great, you score model outputs, provide quality feedback, and the system rewards you with $OPEN . But I pulled up the last thousand feedback records and their corresponding reward amounts, and I found a pattern. The addresses that actually snagged high rewards had super short feedback—just a few words like "bad," "correct," or "garbage," without even a decent reason. Meanwhile, those who wrote detailed analyses on why the model output was flawed barely got anything. The feedback mechanism at #OpenLedger doesn’t care about quality; it’s all about speed. After each model update, the first hundred feedback submissions get double weight, and everyone else, no matter how good their input is, gets diluted. This forces everyone to race against the clock instead of using their brains. You spend ten minutes analyzing an output thoroughly, and someone else just drops a "bad review" in three seconds and walks away, ending up with rewards that are a fraction of what you get.
What’s even crazier is that penalty mechanism. The whitepaper says malicious feedback gets slashed, but I checked the on-chain records, and there hasn’t been a single penalty executed to date. So, you can throw random scores around with no consequences, but if you take your time scoring seriously, you end up making less because you were too slow. This rule set has completely discouraged anyone wanting to do a good job, leaving only bot scripts and speed demons. I did the math; an average user spending fifteen minutes on a solid feedback gets about $OPEN , which is worth less than $0.5. That’s not even enough for a single Satoshi of Bitcoin; who’s going to want to bother?