[57north-discuss] IRC moderation

[Robert McWilliam]

Yesterday Iain (I see we now have a second Iain - this was irl) banned
hackerdeenbot from #57N.

This pissed me off a lot but I decided to hold off complaining about
it till I'd slept on it to see if I could still be bothered (again)
trying to fight for what I want the space to be or if I should just
take the hint and leave the rest of you to it...

Good news! My obstreperousness has won out and you get the following
rant.  

I think[0] he was objecting to a feature of the bot where it will give
a randomly filled in result from cards against hackspace (a variation
on cards against humanity that's a subset of the main pack plus some
hackspace related cards) if there has been no activity in the channel
for some randomly selected period of time between 10 minutes and 5
hours.

I'd categorise the CAH output as often nonsensical, regularly
offensive (sometimes spectacularly so) and occasionally funny. I think
that description also applies to the content in #57N that comes from
people.

If anyone hasn't seen the CAH output in #57N: this has (finally)
prompted me to put the code and lists up on github:
github.com/ormiret/cards-against-hackspace

You can get a feel for what is produced at idea.bodaegl.com/cah (you
get a new result each time you visit the page).

I added this feature to hackerdeenbot for a few reasons.

Firstly, I find a decent chunk of the CAH output amusing and thought
others might too.

Secondly, I thought inserting this when the channel was quiet might
stimulate discussion. This has happened a couple of times but most of
the CAH outbursts have happened during the night when no one is paying
attention to the channel (if hackerdeenbot is allowed back I might
rethink the scheduling).

Thirdly, I like that this output is achieved from a trivial
algorithm. The processing selects from a list of statements with
<blank>s in them and replaces each <blank> with a random selection
from the list of answers. It coming up with things that make sense is
a neat demonstration of the malleability of the English language. I
really like the cases where the output is grammatically correct but
doesn't make any sense and the inverse where it is technically wrong
but there is clearly some meaning to it anyway (and how sometimes the
mistakes look like the kind of mistakes people make).

Fourthly, I like the "offensive" output for the questions it raises
about the nature of offense (and because the things that can be
offensive in some combinations can be hilarious in others). Is there a
need for intent behind the statement, or can an algorithm be
offensive? Even an algorithm as simple as randomly shuffling phrases?
How much of the offense people find in the statements is from the
combination the bot has come up with or are there topics that cause
offense no matter what is said about them?

I added this feature after reading something from the government about
them wanting to police speech on the internet. My bot occasionally
saying things that could have been covered by the proposed limits was
my small protest in favour of free speech including the right to say
things that people might find offensive. It amused me to consider
someone trying to arrest hackerdeenbot. But we don't need to worry
about the government placing limits on free speech as Iain is way out
ahead of them.[1]

Setting aside any of the reasoning for why I think it is OK for the
bot to have this feature it's clear from the fact that I coded it up
that I think it's acceptable and even desirable for the bot to do
this. Iain banning the bot from the channel makes it clear he
disagrees. I am really pissed off that he thinks unilaterally banning
the bot is an appropriate way to deal with this disagreement.

I get that community moderation is a Hard Problem, and hackerdeenbot
may have stepped over the line of what we decide is acceptable.
Pushing around that limit is part of the point of the feature.

I strongly favour a light touch approach to moderation where whether
or not something is acceptable is discussed and if there is consensus
that a bot or person is over the line they are told that and allowed
to change. Banning should be a last resort. I am *very* opposed to
bans being handed out hypocritically[2] and arbitrarily[3] on the
whims of individuals.

TLDR: hackerdeenbot has been banned from #57N and I am not at all
happy about it. I think hackerdeenbot's behaviour was
acceptable, and pretty much normal for the channel. I see serious
problems with the way this ban was imposed.

      Robert

[0] From context it looked like that was the trigger but Iain didn't
actually give any explanation. 

[1] Just to be clear: this is hyperbole. I don't actually think any of
my or hackerdeenbot's rights have been trampled here - both I and the
bot are still free to say what we like elsewhere. But the unilateral
banning of disagreeable content is not how I want the space
communication channels to be operated.

[2] Iain's contributions to the channel are regularly offensive.

[3] As are many others.

________________________________________________________
Robert McWilliam     rmcw at allmail.net    www.ormiret.com

Cleverly disguised as a reponsible adult.
-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 836 bytes
Desc: Digital signature
URL: <http://lists.57north.co/pipermail/57north-discuss/attachments/20150405/76f3be61/attachment.sig>