New Database Showcases How Algorithms Are Rewriting Government Policies Around the US (muckrock.com) 19
v3rgEz writes: Every day government decisions from bus routes to policing used to be based on limited information and human judgment. Governments now use the ability to collect and analyze hundreds of data points everyday to automate many of their decisions.
The non-profit MuckRock, in partnership with Rutgers Institute for Information Policy and Law, has a database detailing how local governments across the U.S. are adopting algorithmic decision making, as well as an open collection of contracts, manuals, and other primary source documents detailing how these programs are implemented and overseen. "Automation and artificial intelligence could improve the notorious inefficiencies of government," argues one page at Muckrock, "and it could exacerbate existing errors in the data being used to power it..."
"Does handing government decisions over to algorithms save time and money? Can algorithms be fairer or less biased than human decision making? Do they make us safer?"
The non-profit MuckRock, in partnership with Rutgers Institute for Information Policy and Law, has a database detailing how local governments across the U.S. are adopting algorithmic decision making, as well as an open collection of contracts, manuals, and other primary source documents detailing how these programs are implemented and overseen. "Automation and artificial intelligence could improve the notorious inefficiencies of government," argues one page at Muckrock, "and it could exacerbate existing errors in the data being used to power it..."
"Does handing government decisions over to algorithms save time and money? Can algorithms be fairer or less biased than human decision making? Do they make us safer?"
convenient for some politicians (Score:2)
This can be a convenient way of ignoring certain underrepresented social groups. An algorithm may not understand the special needs of a neighborhood of immigrants and it may not recognize that another neighborhood is 'special' and populated with wealthy people. An algorithm allocating school funding may not understand that one community has a much higher birth rate than another. The people affected by the algorithm will never know how it came to various conclusions. An algorithm may be a convenient way to r
Re:convenient for some politicians (Score:4, Insightful)
Re: (Score:1)
Re: (Score:2)
Use an algorithm for voting districts (Score:4, Interesting)
Note what you don't feed the algorithm. Race. Party preference. Income. Just people and natural dividing lines.
Re: (Score:2)
people tend to live around like groups. But that doesnt mean bias, and that is a little bit of what you want when voting anyways.
What we the people want, is to stop the redistributing of one party to be water down amounts the other party to insure forever grantee election of one party.
Dems would gerrymander 7 predominate dem voting low neighborhoods with 3 neighborhoods. And Reps would do just the opposite. Of course sprinkle a few all of the apposing in one district to keep their party dominate overall.
S
Re: (Score:2)
Gov't by Computer (Score:2)
Waste, fraud and abuse of Govt cant be eliminated. (Score:2)
Fact 2: If a corporation is wasting money like this, people benefiting by that waste would not be able to nominate the board of directors to that company.
Fact 3: Supreme Court has repeatedly ruled that corporations are people and they have all the rights of flesh and blood citizens including free speech and money is speech too,
Fact 4: Beneficiaries of government waste, fraud and abuse actively participate in the ele
Re: (Score:1)
Who checks the Algorithms? (Score:3)
What if they are deliberately skewed towards the policies that the current POTUS wants?
What happens is there is a change of POTUS in November?
How long will it take for the [cough][cough] algorithms to be adjusted to fit the aims of the new administration?
etc
etc
etc
This is a slippery slope.
Same song different verse (Score:2)
This is just a more elaborate version of that bane of the mid '70s "Well THE COMPUTER says...". (Yes, you could actually hear the bold italic all caps). And it's subject to the same failings.
Algorithms are only as good as their developer. They may be far from optimal and they may fail utterly on corner cases. They also leave no human with a sense of responsibility for the outcome, and too often result in nobody having the needed authority to change an outcome no matter how obviously wrong. It's everything
Algorithms are not AI (Score:1)
Good decisions are based on insight ... (Score:1)
... not on numbers. Seems making bad decisions got even easier this way. Explains quite a lot though.
the one useful feature this should have (Score:2)
these things should be designed so as to prevent corruption happening.
this is the one thing this could bring to government
Re: (Score:3)
Are there humans reviewing all this? (Score:2)