Conditional proof of stake hashcash

Applications
19

Here’s another variation using public voting and reviews.

Currently, we get bombarded with marketing trying to capture our attention and make want to buy things, and this is followed by various rating and review systems that are often created by the same platforms that are trying to sell us stuff in the first place. This is noisy, heavily biased, and subject to manipulation; everyone wants to say their thing is the best thing and they have no reason to tell you otherwise. The real information comes after people buy, but most people don’t have that information when making the purchase decision, so they either don’t do it or often have regrets. If you want to try something new you take the risk ¯\(ツ)/¯

What if we had a system where the producer AND the user rate the product, and if the user’s ratings are not in line with the producer’s ratings, users get a discount for current or future purchases? It’s a win/win because then people will be incentivized to try and buy more stuff, and producers will be incentivized to produce better stuff. It also greatly increases the cost/burden of purchased reviews (positive or negative).

For example: if you’re a band who does live performances, and you think your show is great and people will love it, rate your show X out of 10 stars. If fans disagree, offer them a refund proportional to the delta between your opinion and their opinion. This would incentivize artist/brands to make the best stuff possible because that way they get rewarded more in the form of money and positive marketing, but it also incentivizes the fans/consumers to try out new things even if there’s a risk they don’t like it. Only people who purchased a product can leave reviews, and they are incentivized to do so either by social pressure if reivews are public and/or by economic incentivizes in the form of a refund if they were dissatisfied or a discount on future purchases if they were satisfied.

I currently see this in 2 flavors, basic and keynesian, but I’m sure there are many more:

  • Basic: X is your self assessed star rating, if people feel like your show/thing was less than X give them a refund proportional to the difference (say 10% per star or something), and if they feel like the value received was greater than X they can leave a tip and/or get a coupon for a future purchase.
  • Keynesian: X is your self assessed star rating, if people think that it was less than X they would receive a refund proportional to the other people who also thought it was less than X, and if people thought something was greater than X they would be encouraged to tip an amount proportional to the amount of people who also felt that way.

Regarding attacks to this system, it’s not perfect and would not scale to something as important as elections or health care, but neither would Yelp or Amazon reviews. This is intended to reduce the surface area of crowdturfing attacks from 24/7 online everywhere to verified users who are socially and economically incentivized to create signal rather than noise. For may usecases that would be a great improvement. The system can still be attacked, and there are many many ways to do so within and beyond the protocol described, but I think the cost of doing so would be higher than the systems in use today. Happy to talk about said attacks though, so please let me know what you think! :slight_smile:

Inspired by ideas from these blog posts:

Initial (and evolving) sketches of the system can be found here:

1 Like