Sorted output:
  [
    'Stanford University — Machine Learning Security Research Led by Dan Boneh and Florian Tramer',
    'Wilson Center — AI Policy Seminar Series',
    'UC Berkeley — AI Safety Research (2018)',
    'Machine Intelligence Research Institute — AI Safety Retraining Program',
    'Ought — General Support (2018)',
    'AI Impacts — General Support (2018)',
    'Open Phil AI Fellowship — 2018 Class',
    'Oxford University — Research on the Global Politics of AI',
    'Michael Cohen and Dmitrii Krasheninnikov — Scholarship Support (2018)'
  ]

Stanford University — Machine Learning Security Research Led by Dan Boneh and Florian Tramer
  Number of paths:  102
  Mean: 0.04340463625937361
  90% confidence interval: [
      3.102187977841675e-7,
      0.10636313508232977
  ]
  3.902 seconds needed for processing

Wilson Center — AI Policy Seminar Series
  Number of paths:  52
  Mean: 0.23923636834715892
  90% confidence interval: [
      0.0000012797707857581555,
      0.7647175734237982
  ]
  0.855 seconds needed for processing

UC Berkeley — AI Safety Research (2018)
  Number of paths:  37
  Mean: 0.6148765126821536
  90% confidence interval: [
      0.0000012857319777641759,
      2.4285898907477064
  ]
  0.484 seconds needed for processing

Machine Intelligence Research Institute — AI Safety Retraining Program
  Number of paths:  21
  Mean: 46.33576477417035
  90% confidence interval: [
      0.000022241280559831212,
      201.37264605399085
  ]
  0.173 seconds needed for processing

Ought — General Support (2018)
  Number of paths:  36
  Mean: 63.37826067838051
  90% confidence interval: [
      0.00004411196776834627,
      145.5050535108514
  ]
  0.412 seconds needed for processing

AI Impacts — General Support (2018)
  Number of paths:  47
  Mean: 9862.157135402398
  90% confidence interval: [
      0.018650934888088284,
      22307.090052203544
  ]
  0.635 seconds needed for processing

Open Phil AI Fellowship — 2018 Class
  Number of paths:  372
  Mean: 86301.01328794319
  90% confidence interval: [
      26.02179139839112,
      115734.18190287006
  ]
  35.203 seconds needed for processing

Oxford University — Research on the Global Politics of AI
  Number of paths:  2261
  Mean: 785779.6653748289
  90% confidence interval: [
      114.32195138741375,
      960250.7929788446
  ]
  1048.955 seconds needed for processing

Michael Cohen and Dmitrii Krasheninnikov — Scholarship Support (2018)
  Number of paths:  5243
  Mean: 1482642.2273006735
  90% confidence interval: [
      8.379381400123396,
      330071.1150485886
  ]
  5550.563 seconds needed for processing