[rfk-dev] NKI update 2012

Peter A. H. Peterson pedro at tastytronic.net
Wed Oct 31 08:43:08 PDT 2012

I was thinking it might be good to do an editorial pass over the
existing NKI, and encourage devs (anyone on rfk-dev) to submit NKI for
the new version. If people would prefer, I can set up an nki-editors
mailing list, but it seems to me like rfk-dev is low traffic enough we
can probably just handle it here.

Anyway, the motivation for editing came up for me because I was
playing the other day and hit

"Robot may not injure kitten, or, through inaction, ..."

... and then about 2 NKIs later hit

It's Asimov's Laws of Robotics.  You feel a strange affinity for them.

The first one made me laugh. The second one made me disappointed, like
someone retelling you a joke 5 minutes after you laughed the first
time. It's gilding the lily a bit.

Once, when I was at the robotfindskitten monastery meditating, it
occurred to me that NKIs are ACK (as opposed to NAK) when they provide
a unique experience (or opportunity for an experience) to robot, that
is not part of any dominant NKI theme or flavor. Obviously that's a
subjective definition, but I think that's unavoidable.

For example, both the laws of robotics NKI offer what I would call
good experience. But the experiences are similar, and I find the first
one more effective. When both happen to be in the same game, it
actually breaks the spell of rfk a bit.

Similarly, a number of the NKI are sort of brain teasers and/or
sciencey jokes, like

This is a Lagrange point. Don't come too close now.
Thar's Mobius Dick, the convoluted whale. Arrr! 
It's the horizon. Now THAT'S weird.

They're great because they also kind of doubly tickle your brain --
first, they're good NKIs, and second, they can't exist in reality so
there's this pleasurable experience of imagining them somehow being
"objects" on the plane robot experiences. 

But the game would be poorer if it seemed like the majority of NKI
were of this type, or Nethack jokes, or puns on kitten, etc. So
there's a balance between the number of arguably similar NKI and the
total number of NKI, and how NAK it is if those NKI appear together.
(As an aside, I would argue that brain teasers are less similar as a
class than say, Nethack references.)

This also explains why current pop culture references and name
dropping are less effective in general, because future robots won't
understand them because they are obviously missing a piece of
short-lived cultural context, or it will seem like the point is to see
somebody's name in lights, and so you can see the "man behind the
curtain" per se. 

Some of these, like the ones that mention Leonard and I, have been
left in as artifacts (but I would be happy to remove them if people
want to). Others that might be considered a kind of pop culture (like
Dmitry) get a bit of a pass because of an expectation of a shared geek
cultural context. But there aren't any references to Britney Spears or
Twilight. Thank gord.

In some cases, I think that there is a positive meta-experience
between NKI, such as with 

"Free Dmitry Sklyarov!" and 
"It's a free Dmitry Sklyarov!" 

Or, with 

"Yes!" says the bit. and
"No!" says the bit.

(The latter being a collective TRON reference, but one which stands on
its own binary leg.)

These pairs or trios obviously need to stand on their own merits, but
can work together to offer a larger ACKier experience -- when they
happen to appear together, the overall experience is greater... as
long as they don't become too winky of a theme.

Again, it's subjective. But that's kind of how I think about NKIs.

What do you folks think? I know this isn't the only way to think about
it, but having some kind of shared understanding should help us when
making editorial decisions.

I'll post my proposed changes in a followup email.


Peter A. H. Peterson
Graduate Student Researcher
Laboratory for Advanced Systems Research
University of California, Los Angeles

More information about the rfk-dev mailing list