R. A. Dehi @radehi@qoto.org

@conradhackett 8% of Americans are utterly delusional

@chadloder You are suggesting that Elon Musk bought Twitter to altruistically boost the profits of Starbucks and Amazon?

@mattl Just bulked, time to cut.

@bcrypt engagement intensifies

@kenshirriff Easy mistake to make. Interrupt handling is tricky.

@smari Can be hard to outsource things you'd write in an evening; glad you made it work.

@smari congratulations!

@shriramk @VirginiaEubanks @epicprivacy Perhaps you are correct; I will have to consider it. Thank you.

@shriramk @VirginiaEubanks @epicprivacy What I said above is slightly wrong: I should clarify that disproportionately placing surveillance cameras in Black neighborhoods is an example of how greater efficiency can make algorithmic law enforcement more harmful than human law enforcement even without secrecy; secrecy is a separate issue.

@shriramk @VirginiaEubanks @epicprivacy I'm criticizing the *title* of the article, and the way it uses the word "algorithms".

I agree that we should criticize harmful ways of using algorithms, of which there are of course many. But the things you're describing are a question of lack of transparency in government, not of use of algorithms. Keeping secret files on citizens on paper, or conducting secret hearings, would also be harmful, for the same reasons. Secret algorithms can be more harmful because they are more efficient, of course, and more extreme. The article's example of placing automated traffic cameras disproportionately in Black communities is an example of how this can produce injustice: heavily policing Black neighborhoods has been a known problem for decades, but automated cameras remove the cost, potentially enabling far deeper levels of oppression.

However, the article consistently uses the term "algorithm" in a way I am sure you will agree is nonsense:

'The nonprofit spent 14 months investigating the city’s use of algorithms and found they were used across 20 agencies, with more than a third deployed in policing or criminal justice. ... The project team concluded that the city is likely using still more algorithms that they were not able to uncover.'

If we interpret the term 'algorithm' according to its standard meaning, these sentences make no sense, because are so trivially, obviously true (except for the part about "more than a third"), but yet are presented as if they are news, the damning results of a 14-month investigation!

This is like the political candidate who accused his opponent of having a sister who was a known thespian and having himself matriculated in his youth. It's utterly indefensible, shameful scaremongering, and it leads to the kind of travesty mentioned at the end of the article:

'Last month, lawmakers in Pennsylvania...proposed an algorithm registry law.'

In less law-abiding places than the US, expect mob violence instead.

I haven't read EPIC's report, and so I don't know if this is entirely Wired's fault.

Dave Kilian wrote an introduction to lock convoying for total beginners, people who don't know what is a context switch.

https://davekilian.com/lock-convoys.html

@timbray She's @skdh@nerdculture.de

@VirginiaEubanks @epicprivacy @shriramk Nor am I accusing you of *being* either ignorant or shameful. I am saying that the people who will do the lynching are ignorant, and that the *act* of whipping them up like that is shameful.