Re: [projectvrm] Councils use 377,000 people's data in efforts to predict child abuse | Society | The Guardian

[Doc Searls < >]

On Sep 18, 2018, at 12:51 AM, Iain Henderson < " class=""> > wrote:

Yes, and through serendipity it looked worse in the paper version…..

It's by John Naughton, whom some of us know. Good guy, good journalist. He also knows our work here. This April column by John opens by sourcing one of my blog posts.

It's also not about the headline. It's about how the AI craze is thick with "magical thinking" and a lot like alchemy, a metaphor John got from this Ali Rahimi talk.

The real issue behind the headline isn't AI, but rather Big Data, the rude ways it's obtained, and the the tendentious conclusions that are usually reached.

I’ve not had time to ponder on this particular use case as yet, but i’m currently working on a big AI project in financial services, watching how the algorithms are being built; my prediction is that it is a data disaster waiting to happen.

Radical as it may sound, I think the steady state 10 years out will be the all data regarded as sensitive (in the legal sense) will be mandated to be held and managed by the individual or their agent. After that use for resource allocation is a good idea, until then no thanks.

Totally agree. What lies between now and then is a call to our work here.

Doc

Iain

On 18 Sep 2018, at 08:26, Tom Barnett < " class=""> > wrote:

https://www.theguardian.com/society/2018/sep/16/councils-use-377000-peoples-data-in-efforts-to-predict-child-abuse

While this is not about advertising it is about personal data, title and benefit. 

I'm intrigued to know what the group thinks of this use of data - where is the invasion of privacy (or not) in the process of service delivery?  Is this classic 'targeting' or is this a new form or prioritisation and allocation of resources?

<Screen Shot 2018-09-18 at 08.30.06.png>