Finally come to fruition with 3 months ideation and 3 months planning! 25 participants in iceland for the next 2 weeks @juanbenet opened the retreat here in iceland on what impact evaluators are: a block rewards type mechanisms to coordinate actors iver the internet He broke it down into 3 steps; - measure relevant projects - evaluate or assign weights - a reward loop He gave the 2 powerful examples of bitcoin which gave rewards based on hash power, catalysing hardware & filecoin which did it on proof of storage & utilizes space capacity These rewards are both positive sum since the native currency of these networks can increase over time & also very frequent (less than 10 mins) providing quick feedback loop He also gave more real world examples like academic paper acceptance or even nobel prizes, which are less algorithmic & more subjective but have a longer time frame that make them less bankable for teams These types of retreats have been key to take forward the state of the art in various sectors, such as solvay for quantum or dartmouth AI He's found that having in person colocation but also deep work time is important. It's easy to simmer in ideas endlessly, so we need to start with exploration but eventually converge on specific solutions after a few days that can be shipped in the remaining week I also loved the specificity: improvements to mechanisms like hypercerts or rpgf are not in scope Working on evaluation frameworks & functions to create new block rewards type mechanisms are what we hope to see emerge over the next 2 weeks