CredWeb this week

Meeting this week, usual time, 8 April 2020 2pm ET 
<https://www.timeanddate.com/worldclock/fixedtime.html?msg=CredWeb&iso=20200408T14&p1=43&ah=1>usual 
place https://zoom.us/j/706868147.

Last week, we had a lively discussion of my straw proposal for a 
credibility network to help people figure out who they should trust. 
Coming out of that, I wanted to try running the network viewer 
<https://credweb.org/viewer/> on people's real credibility statements 
(so far we have some from me, Symeon, Subbu, and Annette), but I wasn't 
able to get to that this week, alas.

I did write up The “Probability of Net Benefit” PNB Credibility Score 
<https://docs.google.com/document/d/18tPpaQnxuGbgG7sOWEfPYWiA9D4vNMfxXR1SJRIvAoI/edit#>, 
in the hope of making that part of the discussion more concrete. I'd 
love to see other proposals for how we can combine or reason about 
credibility data, since this one has some clear flaws.

One interesting focus this week might be Paul Graham's new micro-essay, 
Coronavirus and Credibility <http://paulgraham.com/cred.html>, where he 
argues the public failures of people around Covid-19 ought to negatively 
impact their credibility.

As I tweeted <https://twitter.com/sandhawke/status/1247183232241672195> 
back to him: "Indeed. But humans aren't very good at remembering these 
details, and we don't currently have good systems for collaborating on 
this. Open solution in the works at@W3C <https://twitter.com/w3c> 
http://credweb.org <https://t.co/Sott21IZaX?amp=1>. Hoping to release 
demo in a matter of days."

We could perhaps do that with credibility statements like:

  * I consider the credibility of [  ] to be (increased | decreased) by
    [  ]
  * I consider the credibility of [  ] on the topic of covid-19 to be
    (increased | decreased) by [  ]
  * I consider [  ] to be not credible on the topic of covid-19 due to
    [  ] as seen at URL [  ]
  * ...

If you could spend a few minutes trying to make statements like that, to 
see what feels right, and make them available (like this 
<https://docs.google.com/document/d/1ShiK_Pkd46foPbWCayfkUh3UV5Bhd5KHI5SKYoUIkiI/edit>, 
same as last week, or just in email to the list), it will help inform a 
discussion tomorrow, and I will redouble my efforts to make a viewer for 
them.

     -- Sandro