[Osmf-talk] Paid Mapping / WikiPR like issues in OSM?

[Serge Wroclawski]

Jonathan,

I think you've touched upon the key issue here, which isn't paid vs
non-paid, it's a question of motivation and incentive. and how this
interacts with the issue of agency.

Imagine I had a contest and said "Whoever maps the most supermarkets
wins $1000. There would be motivation to go out and find supermarkets.
But there would also be motivation to copy other maps, or possibly
invent supermarkets out of whole cloth.

Now what if instead of a contest, these were paid employees? This
changes the equation because now we are talking about the issue of
"agency". If an employee does something wrong on the job, you do not go
after the employee individually, you go after the company. And the
company cannot say that they are not responsible for the actions of a
single employee, because that employee was acting as an agent of the
company.

So how does this intersect OSM?

Luckily, it hasn't so far AFAIK, but imagine if we had an interested
party who wanted to add features to OSM such as the example of the
supermarkets. They tell their employees "Go map supermarkets" and the
employees each do so, under their own accounts. So we have users A, B, C,
and D all mapping supermarkets. User A goes out and maps real
supermarkets that they've surveyed themselves. Users B, C, on the other
hand, copy supermarkets from other maps, maybe one of them uses Google
and the other uses Here Maps. And user D doesn't bother copying, they
just make up supermarkets. And so all over the world, we have
supermarkets being mapped, maybe for months.

No one really notices a few "wrong" supermarkets until one day then we
see that user D's contributions are untrue.

So we now have to go through the process of reverting D's contributions,
maybe blocking D's account (because D continues to have an incentive to
map!), etc.

We may never even notice B, C's contributions are copies until much
later, but such edits, whether they're copied or made up are bad.

Now imagine there is employee E. E does something even more interesting.
E is a clever programmer and screen scrapes supermarket locations from
websites, then geocodes them from Nominatim into OSM. The problem is
that these locations from Nominatim are just estimates (as they are
for many places)

In all of the cases, we have bad edits, with harm to the community
because the community has to clean up the mess. Cleaning the mess has a
cost- it takes time for users to revert changesets, to do the
investigations, etc.

If it's a single vandal user, we take these problems as rogue "bad
actors", but if it turned out that all these editors were acting on
behalf of a company, the question in my mind is "What responsibility
does the company have to the OSM community for the actions of its
employees?"

This hasn't happened yet- thankfully, and I hope it never does, but I
think Frederik brings up a really interesting question and it's
certainly one worth thinking about.

- Serge