Nearly everything that matters is a side effect.

[maradydd]

(The following, plus a few side anecdotes, was delivered at SIGINT 13, Cologne, Germany, July 5th, 2013. Here's the video.About a year and a half ago I was in Brussels for a workshop that Google and Privacy International hosted. The goal of this workshop was to develop policy language around privacy that Google could use in negotiating with

Comments

[whswhs]

Internet privacy policy is being developed by Ayn Rand villains? No, seriously, you've got your rent-seeking, you've got your belief in the primacy of state goals, you've got your denial of objective reality. . . .

[steer]

I'm puzzled by your "break the internet" story because I just don't understand what you're saying. On the face of it it's probably a bad idea to just break the existing internet. On the other hand, it's a very good idea to design something which can replace it or change it in such a way that it's no longer the internet.
I don't know if you think the "fuckup" guy is right or wrong and you described the debate in such strange terms I don't know whether you sypathise with the room (who seem to want to reengineer the internet, which is great) or the guy (who seems to be cautious abou breaking it (which is great).

that mathematics had to be subordinated to national sovereignty.

Already is to some extent -- I know of at least one set of physics equations which (at least as of the mid 90s and at least in the UK and I think US) were not public but instead you could send parameters to people and get a response. You were not to send too many parameters lest you reverse engineer the equations.

[maradydd]

The "(ahem)" is relevant. Hint: it was not a guy.

More later, conference stuff going on.

[steer]

I think you misunderstand the nature of my confusion. I'm not trying to guess who it was because I doubt it would mean anything to me as I don't move in those circles. You told the story so circumspectly I didn't actually know what the argument was about or which side you were on.

[steer]

Doh... sorry, being slow.

[moof]

Sometimes I wonder if it's possible to know one's own or someone else's Goedel number - or if it's possible, that there are only certain classes of minds (or Minds) that can grasp the sufficiently higher-order logic to do so. (Smullyan probably has something interesting to say about it.)

As a corollary, I wonder how close my emulated social sense comes to the real thing.

[maradydd]

I don't think mine is very much like the real thing at all, but I have some rudimentary processes for improving it. I don't know whether that necessarily implies that I'm iterating toward the real thing, and in fact I'm starting to suspect it doesn't, but I'm actually okay with that.

[whswhs]

I don't think even neurotypical people have anything describable as accurate social modeling. Our brains evolved when we lived in social groups of no more than a few hundred; research on memory thresholds suggests that a normal memory space can contain ~500 distinguishable entities (for example, primary taxonomic terms such as "dog" and "eel" and "fly"), and I've seen anthropological speculation that this comes from the size of social group where we can know everyone by sight. But we don't live in a society with 500 people, but in one with over 10,000,000 times that many; the number of sovereign states is a substantial fraction of 500, in fact. So what our primary social modeling system is doing is not actual social modeling, but folk social modeling, in the sense in which impetus theory is folk physics or the concept of the mind as an engine for taking attitudes toward propositions is folk psychology

[maradydd]

I'd like to think that a dearth of initial biases is an asset in this, yes.