[Foundation-l] On curiosity, cats and scapegoats

Kim Bruning kim at bruning.xs4all.nl
Fri Sep 9 11:54:26 UTC 2011


On Fri, Sep 09, 2011 at 09:24:36AM +0100, Tom Morris wrote:
> On Thursday, September 8, 2011, Kim Bruning wrote:
> 
> > That said, even a self controlled filter can be problematic qua bias
> > (especially if you're not sure entirely how to control it) [1]
> >
> > [1] http://www.thefilterbubble.com/ted-talk
> 
> 
> I'm not sure what I think about the image filter, but that's a pretty ropey
> comparison:
> 
> With the proposed image filter, the knowledge that a filter is in place
> would be quite obvious: there'd be a big gray box with "Image Removed" or
> something. And if you want to see them, you are only a click away from
> loading them.

This is true With the proposed image filter and only with the
proposed image filter -provided that the entire connection between
you and the wiki is transparent- (Which it need not be.)

If there are filters in between, it need not be true today. If your browser
is not under your control, it need not be true today.

That's today. A little while after the filter is introduced, we will
have 2 sets of effects, political and technical.

* Politically, people will see wikipedia as endorsing filters (it
  seems quite unlikely that they will take note of the subtle
  properties of our filter that make it tolerable to us) so more filters 
  will come into circulation, increasing the chance that there is a 
  filter not-under-your-control between you and the wiki. This
  will change the way the filter works.
* Technically, a new set of categories is created, or categories will
  be pressed into the new role. These categories are invaluable to
  filter makers. This means again there will be more filters (as
  above), but also that the little grey box will not be there to
  remind you that something is missing.

After that, we get back to the side effects of regular (non-wikipedia
kind) filters. This information is well documented all over the net.
You'll discover that not just images, but also the pages those images
are on will not be reachable. We've been told on this list that this
already happens to some people today. It seems pretty obvious that
the effect will be much multiplied once the categories are available
to third parties.

> And how is bias being introduced into my views by being able to go
> to [[Cock ring]] and not seeing a picture of a penis?  I fail to
> see how being able to opt-out of saucy sex pics actually moves us
> in any significant way closer to a world where we live in "filter
> bubbles".

I just provided you with 2 steps in that direction, pretty much the
first moves on the blue-team and red-team sides. 

On or around move 2 (6-12 months) we can start seeing people either
deliberately or accidentally start filtering things that are nothing
to do with sex or drugs at all, up to and including censoring of
civic information.

This has happened with filters in the past. I don't yet see why our
filters wouldn't follow the same playbook. So far, there is nothing
to differentiate our history from existing history.

"But ours will be different" is not an argument. ;-)

> Again, I'm not sure whether I support the image filter, but it's a
> rubbish argument to say that it creates filter bubble-type
> scenarios.

Seriously? That's pretty definitive. [citation needed] Show me a
filter scenario where it *hasn't* happened! 

sincerely,
	Kim Bruning





More information about the wikimedia-l mailing list