Wednesday, 13 December 2017

Roger Stone

Read Jon Ronson - The Elephant in the Room

Alex Jones is paranoid.

Stone was introduced to Jones by Reeves - the grandson of the original superman
Stone had a business with Mannafort and they also worked with Lee Atwater.

Stone worked for Savimbi in Angola
Bob Dole

Stone knew Roger Ailes

Except Stone is part of the establishment - he was Nixon's Counsel.

They have a cult mentality - they will kill the GOP
This is a coup - they do not want to win hearts.

Alex Jones hijacked the Young Turks
Stone and Jones brought up Clinton rape accusations to the debates.

Manages media - creates false flag confrontations


This is a slightly puzzling book 

p22 the author is partially wrong. Hierarchies do work and Herbert Simon showed why, but this was not because of top down control. They can be non-directed and spontaneously arise.

p25 gatekeepers to the rich and powerful. Is this a good idea?

p27 - why did the author write the book?

  • potentially undermines her credibility.
  • makes people wary in talking to her.
  • obvious that she is a Soros fan.
p32 "Money is mostly created by banks offering loans" regulated by central banks interest rates and asset purchases (Gold etc.) At the minute with QE $17 trillion has been pumped into the markets and created huge asset bubble such as BitCoin. The intention was for the money to be used for investment and to kickstart growth but this has failed. It has remained in the markets and the banks and not been distributed to the wider economy. This is going to result in a very serious and drastic need for realignment. 

Fundamentally commodities are more important than other markets because we cannot live without them. We depend on them for:
  • Shelter
  • Warmth 
  • Food
As Apple share price rises the return per share has fallen because this is pure speculation and not investment. 

p53 power of the central banks is greater than the politicians. Brexit proves this wrong. You can get a populist vote in ignorance of how the central banks work and this can create a suicidal economic policy.

Creating Research Objects

There is a serious problem with scientific fraud and the reproducibility problem. We need to think about ways in which we can check the integrity of a study. 

This is also a way of encoding know how.

Metadata is too time consuming to create at the minute. It needs to be built into the planning and research process itself (GitHub?)

Want to create a knowledge exchange report
Open Research Data - Manyika
Rules for growth - Stodden

Data management plans are required by research councils

  • Integrity checking
  • Hashing
  • What are open file formats?

Blockchain for the trust layer
  • Politicians - regulations and policies
  • Qualifications
  • Medical Records
  • Passports
  • Forensics
  • Risk assessment and rating
Restoring trust is essential
(Byzantine General's Problem)

Money is a trust system representing work done
Reputation is also a trust system but this is only as strong as the weakest link.

Thinking about the bootstrap

  1. Bootstrap samples experimental units but in phylogenetics you sample the VARIABLES - sites.
  2. How should we treat sites?
    1. Remove totall variant?
    2. Remove sites where a row is missing?
  3. You cannot say that parametric and non-parametric are the same thing. They are correlated but not directly comparable.
    1. Carry out FastTree with H5N8, then H5 then N8
    2. Use the parametric and non-parametric bootstraps
    3. Use the CONSEL measures as well.
  4. Having more bootstraps than 100 makes NO difference to the bootstrap values. They converge quickly empirically.
    1. This is far below the theoretical numbers needed by Efron says that this is usual.
    2. Suggests that sites are linked and so there is less independent variability than it appears.
    3. Need to experiment with conserved sites.
    4. Need to experiment with the substitution models to look at sensitivity and also gamma.
  5. There is a lack of independence between sites in the evolutionary models but this is IGNORED in the bootstrap calculations. You should bootstrap codons and not individual bases.

Need to create synthetic data where the true tree is known. This can be used to test:
  1. Effects of sampling by censoring the data.
  2. Evaluate modeltest.
  3. Check trees from bad evolutionary models against the best models (probably the same!!!)

The process of learning

Genetic: Very slow learning and wasteful because it depends on selection. This works between generations.

Taught: Fast learning that sums up what happens in a community.

Exploration: Novel learning by experiences. This is learning through interaction

Distances in psychology.

not transitive
not symmetrical

Tversky 1977 - features of similarity
AI cannot make human decisions until it gets beyond clustering distances.

Undoing Project p 107-114.

Belief in the law of small numbers - The Undoing Project p157-163

Pundtits (illiterate "experts") p 168