algorithm   45882

« earlier    

In conversation: Penn Jillette • Vulture
David Marchese with a fantastic interview with the conjuror/juggler/magician, who has a fascinating sweep of insights:
<p><strong>Q: You’ve talked in the past about how the antidote to bad ideas is more ideas. But doesn’t the way things are shaking out online suggest that actually what we need are better ideas and not just more of them?</strong>

PJ: I believe in the marketplace of ideas but you’re right, we now have algorithms that push people crazy. YouTube is set up to push you crazy. If I search for vegan recipes, I’ll end up with 9/11 truthers. But it’s like the first time people saw movies, and the train on the screen was coming toward themThe 1895 film The Arrival of a Train at La Ciotat Station is a 50-second silent film showing a train pulling into a Paris station, and is an early document of cinematic technique: forced perspective, long shots, close-ups. There’s a myth about the film that in its premiere, audiences were terrified by the train coming at them, and ran from their seats. …

<strong>Q: And everyone jumped out of the way.</strong>

PJ: That’s right. They were screaming and yelling, but then it only took a millisecond for people to realize what was going on from that point forward. So even with all this bad stuff happening, yes, I still think people are overwhelmingly good, ideas are overwhelmingly good, and if you have Nazis being able to reach 10 million people, those same 10 million people will also be reached by Martin Luther King.

<strong>Q: Why do you think that? Isn’t the marketplace of ideas as it now exists online intentionally designed to send people further down a given rabbit hole rather than towards contrary ideas?</strong>

PJ: Yes, algorithms are weighted in favor of that, but that’s not the problem. If you’re worried about craziness in the next ten years, I don’t have any hope for you. Fifty years? No problem. It’s like when we first saw advertisements: they worked entirely. But now I can show you a TV ad and you don’t even reach for the phone. The words didn’t change, but you learned to tell that it was bullshit. We’re going to see that happening with the internet. People will learn to separate the good from the bad. But that whole idea that everybody else is going crazy on the internet sickens me. I can tell when something is garbage. You can tell. Who are all these mysterious people that can’t?</p>

Stay too for the bit where he discusses his experience on US Celebrity Apprentice with Donald Trump, and why he won't talk about what he heard. (He does, in passing, dispel any doubts about whether a tape of Trump being racist exists.)
Magic  trump  algorithm  socialwarming 
yesterday by charlesarthur

« earlier    

related tags

&  3d  adamfinkelstein  adaptive  ai  algorithms  angle  anki  anomaly-detection  anonymous  approximation  architecture  arrangement  art  article  artificial.intelligence  artificialintelligence  bacteria  bandit  bellman-ford  bezier  bfs  big-o  bigo  bing  bioinformatics  biology  blockchain  blog  bookmarks_bar  building  buildings  buterin  c++  caching  chart  cheatsheet  choice  choose  chords  christiantheobalt  circle  class  classroom  coding  compare  compiler  compilers  compilersparsing  complexity  composition  compression  computational  computer-science  computer  computers  concurrency  consensus  corner  course  coursera  cover  create  cryptocurrency  cs  cube  curated  curve  data-policy  data-regulation  data-science  data.ethics  database  deep-learning  deep  design  detect  detection  digitalrights  dijkstra  dimension  dimensions  distance  distributed  doodle  edge  education  electrical  elementary  elixir  estimator  ethereum  ethics  evol  evolution  evolutionary  explain  facebook  facialrecognition  fake  feature  feed  fibonacci  fill  filling  floorplan  flow  flowchart  fluid  formal  functional  fusion  g.  game-design  game  games  generative  genetic  geo  geometry  go  golang  google  google’s  graph  graphics  grid  guide  guitar  hallway  hash  health  healthcare  healthtech  hft  ia  image  imaging  implementation  inspiration  introduction  iw  java  javascript  keras  kyleolszewski  lalr-parsing  language  latest  learn  learning  legal  lessig  line  linear  local  lr-parsing  machine-learning  machine  magic  mapping  mariatemming  mary  math  mathematics  maxplanckinstitute  mechanics  memory  method  mit  ml  modular  more  motion  movement  music  myths  n  natural  near  netflix  newsletter  not  of  optimisation  organism  paper  parser  parsing  particles  pdf  performance  photon  pi  pl-design  planning  poetry  point  points  poly  polygon  popular  pratt-parsers  pratt-parsing  pratt  princeton  privacy  processing  programming-languages  programming  proof  puzzle  python  quantum  radius  random  recommendation  recursive-descent  reference  research  room  ross  says  school  science  sciencenews  scikit  search  searchcap:  security  sensor  seo  separation  shift-reduce  shortest-path  siggraph  single  slasher  slope  socialwarming  software  space  spacefilling  spatial  specification  speed  square  stake  strategy  surveillance  survey  switchgear  syllabus  terms  text  the  theory  top-down-parsing  trading  tree  trump  tutorial  twitter  update  vector  video  videos  vision  voronoi  work  yacc  zstd 

Copy this bookmark: