I believe that this post could be split into a few parts.
It’s deep, complicated, difficult and I have no idea what is the right way.
(what is the definition of the right way, if the right way exist, how do we evaluate the right way)
Incentives for coherent voting.
Coherence. Shelling Point.
I can imagine a scenario when a juror has some little known piece of expert information.
In a perfect world it should be possible for that juror to submit evidence and shout from the rooftops: “here is super duper evidence, it changes everything, look at it”
But in a reality, there is fake news media propaganda, many factors matter, including the presentation of the evidence. In this example: this juror is an expert but has only Comic Sans installed on their computing device, no one takes him seriously.
That juror may also find a dilemma:
- To vote for something that they know is true (a little known piece of evidence, buried among fake news)
- To vote for something that will earn them ETH and PNK (coherence reward, not losing PNK)
That leads to various voting strategies. I as a juror (2-3 times, real-life situation) couldn’t care less about the pixelated or blurry background, I wanted to have more crypto.
General policy: Fairness
I think Shelling Point will have to stay. Blockchain has no way of knowing what is true.
I want jurors to consider what is fair, not just coherent.
Aiming for coherency will get us: Keynesian beauty contest - Wikipedia
Aiming for fairness… 
So now we are in a definition of fairness discussion. I suggest: “best interest of the humanity, society, justice”
Example: a guy was caught red-handed, hands with blood, murder was just before midnight.
(some strange things happened, this is just an example)
- Murder before midnight.
- Capture after midnight
- 10 years statutory limitations.
- Prosecutors are a few hours too late. (I told you something strange happened)
True: according to the law from 1929 it was the date of the crime, not date of captured by the police.
Best interest of the humanity, society, justice is to charge the guy anyway, even if according to the law it not true.
Another example could be single mum smoking weed when kids went to sleep. Depending on the jurisdiction, it could be the death penalty.
“Code is the law” or fairness (definition subject to discussion) should prevail?
Moral. Ethical.
General Court Policy once again: Proposal for Kleros General Court Policy - Google Docs
Arbiter has an option:
0 = refuse to arbitrate
Check this out:
1. Kill A
2. Kill B
0. Refuse to arbitrate: kill both
_. Do nothing: kill both and someone else
What if:
- One guy is Vitalik, the founder of Ethereum
- One guy is Clement, the CTO of Kleros
- One guy is a pregnant lady
- One guy is a pregnant lady with Hitler
- One guy is fat and got a speeding ticket in 2009
- Any number of variations, options, fucking considerations…
Do we still aim for coherency, Shelling Point? Where (if at all) do we aim for fairness, moral and ethical standards?
Search for “collar bomb” using your favorite search and information consumption technology:
Kleros should establish a working relationship with https://cueims.soc.srcf.net/
My head is already exploding and I have more attainable goals to accomplish near-term, I’ll probably leave it here.
We have seen many examples in society where mathematicians have used their skills and training to inflict harm. Historically, the Manhattan Project is one of the most highly-debated cases of technological development, and it involved several of the finest mathematical minds of the time.
The diesel-gate scandal at Volkswagen, where cars were fitted with ingenious cheat devices to circumvent emissions tests, is another more recent example.
And less than a year ago, several operating system exploits developed by the NSA (such as EternalBlue) entered the public domain, where they were quickly used to develop crippling malware like WannaCry. Such examples are numerous, and have become more and more common in recent times.
