Tuesday, June 18, 2019

“AI” in Justice Systems



We heard about a new program that the San Francisco, California District Attorney is developing a program to attempt to reduce racial bias. That and another article on “AI Hype” and other AI development projects are in this week’s post.





---
San Francisco District Attorney's Project

An interview on NPR’s “All Things Considered” program (available as a podcast) of San Francisco District Attorney George Gascon describes his efforts to try to reduce racial bias in charging decisions.  The interview question leads that “the idea for this came from a 2017 study, which was commissioned by the district attorney, which found, quote, "substantial racial and ethnic disparities in criminal justice outcomes."

Subsequently, researchers from Stanford University were brought in and they, after study and I’m sure other recommendations, came up with what is essentially a redaction tool “using natural language search, goes through… and removes all the things that could be either racial identifiers or proxies for racial identifiers”.

An “AXIOS Future” article on the same system (scroll down to see item #2) does a nice job of breaking down how the system works and further notes that “It’s not fancy AI” and that the 2017 study “indicated that racial disparities in San Francisco's criminal justice system are largely not the result of the DA's charging decisions.”

Congratulations are sent to San Francisco for the effort but as all the articles noted, there is a lot more work to do.

AI Hype

Another take on “AI” comes via an article on “Medium.com” titled “The BS-Industrial Complex of Phony A.I.” that is subtitled “How hyping A.I. enriched investors, fooled the media, and confused the hell out of the rest of us”.  I enjoyed the article and appreciated the last section on one of my favorite test systems “Amy” of https://x.ai/

And yet Another AI Article

Last, we learn via this article from Forbes about “AI Breakthroughs” about four developments, only one (the first) that have applicability in our line of work. 

  • Speech to text 
  • Protein "folders"
  • Games
  • Self-driving cars (where I learned about Tesla’s built-in “AI chip”)

Enjoy.

No comments:

Post a Comment