---
date: '2022-12-19T14:09:11'
hypothesis-meta:
  created: '2022-12-19T14:09:11.863238+00:00'
  document:
    title:
    - My AI Safety Lecture for UT Effective Altruism
  flagged: false
  group: __world__
  hidden: false
  id: tmH8RH-mEe27ArstPwKXEA
  links:
    html: https://hypothes.is/a/tmH8RH-mEe27ArstPwKXEA
    incontext: https://hyp.is/tmH8RH-mEe27ArstPwKXEA/scottaaronson.blog/?p=6823
    json: https://hypothes.is/api/annotations/tmH8RH-mEe27ArstPwKXEA
  permissions:
    admin:
    - acct:ravenscroftj@hypothes.is
    delete:
    - acct:ravenscroftj@hypothes.is
    read:
    - group:__world__
    update:
    - acct:ravenscroftj@hypothes.is
  tags:
  - nlproc
  target:
  - selector:
    - endContainer: /div[2]/div[2]/div[2]/div[1]/p[43]
      endOffset: 779
      startContainer: /div[2]/div[2]/div[2]/div[1]/p[43]
      startOffset: 174
      type: RangeSelector
    - end: 16443
      start: 15838
      type: TextPositionSelector
    - exact: " And famously, self-driving cars have taken a lot longer than many people\
        \ expected a decade ago.  This is partly because of regulatory barriers and\
        \ public relations: even if a self-driving car actually crashes less than\
        \ a human does, that\u2019s still not good enough, because when it does crash\
        \ the circumstances are too weird.  So, the AI is actually held to a higher\
        \ standard.  But it\u2019s also partly just that there was a long tail of\
        \ really weird events.  A deer crosses the road, or you have some crazy lighting\
        \ conditions\u2014such things are really hard to get right, and of course\
        \ 99% isn\u2019t good enough here."
      prefix: ' the last jobs to be automated. '
      suffix: '




        We can maybe fuzzily see ahe'
      type: TextQuoteSelector
    source: https://scottaaronson.blog/?p=6823
  text: I think the emphasis is wrong here. The regulation is secondary. The long
    tail of weird events is the more important thing.
  updated: '2022-12-19T14:09:11.863238+00:00'
  uri: https://scottaaronson.blog/?p=6823
  user: acct:ravenscroftj@hypothes.is
  user_info:
    display_name: James Ravenscroft
in-reply-to: https://scottaaronson.blog/?p=6823
tags:
- nlproc
- hypothesis
type: annotation
url: /annotations/2022/12/19/1671458951

---



 <blockquote> And famously, self-driving cars have taken a lot longer than many people expected a decade ago.  This is partly because of regulatory barriers and public relations: even if a self-driving car actually crashes less than a human does, that’s still not good enough, because when it does crash the circumstances are too weird.  So, the AI is actually held to a higher standard.  But it’s also partly just that there was a long tail of really weird events.  A deer crosses the road, or you have some crazy lighting conditions—such things are really hard to get right, and of course 99% isn’t good enough here.</blockquote>I think the emphasis is wrong here. The regulation is secondary. The long tail of weird events is the more important thing.