updated: 2022-01-23_12:32:31-05:00

Algorithms of bias movie

  • Face Scanning

    • 98% of matches are false positives.

  • These algorithms are replicating the world. Mathematical, not ethical

  • Algorithms for housing, jobs, education...

  • I think this has a bit of fear-mongering.

  • State surveilence is very different than corporate surveilence

  • CCTV cams are not technically up to the task of being turned into a mass surveilance state

The people paying for the views of poor people are predatory

Corporations affecting government policy

Weird dramatization of ai ...

Gender discrimination in amazon AI

Selling AI's with gender or racial bias to law enforcement & intelligence agencies

Pushing for federal regulation for facial recognition

Risk Models, algorithms of bias

  • Americans are bet against and selected and optimized for failure

  • 4 million Americans loosing homes in 2008

  • Everyone is just trying make a lot of money on it

  • Powerful people scoring powerless people

God damn, stupid ai dramatization

Value added model

  • what value is the teacher offering? (Houston texas)

Why aren't we using the constitutional right to due process in other areas

Recidivism Risk Algorithms:

How likely to recommit crimes again?
Biased against PoC

There is no positive feedback in this system

Algorithmic Justice League

Resume algorithms

Algorithmic Determinism

  • Shopping prices, criminal likelyness
  • China is transparent
  • Black kid stopped because he was matched,
    • plainsclothes officers, searched, fingerprinted etc
  • human rights operation
  • 14 yo
  • misidentification
  • Big brother watch & legoness ?
  • She was on the watchlist for demestic extremists (the lady who was in charge of regulating)

GDPR

  • framework for accountability of data
  • America is the wild west of it, and all the tech companies are there
  • free for all approach to data protection
  • silent dark path of AI
  • not 1984, more insidius
  • Marked senators as dangerous (amazon face )

Goals and challenges

  • law enforcement publicises it

  • other side of positives is the human cost of doing this

  • The technology we are creating shouldn't perpetuate bias

  • oversight in the age of automation

  • FDA for algorithms

  • 2400 faces misidentified..

  • AOC is fire

  • Facebook algorithm for corporate surveilence

  • Cities are banning it

  • Chechov, didn't kill the us

  • fully effient etc is loosing humanity

  • leads to inhuman things

2020 amazon 1y pause on facial recognician
june 2020 federal legislation put to congress, not done yet