Implying humans make consistent moral decisions?

Confirming TQ needs to mine data that it logs itself. It's not totally busy running a submarine sim.

Classifications tend to be quite fast once you've sorted the training set. In any case, I didn't say I thought it was appropriate, just that I took exception to the thought that computers are somehow inept in this regard.

Classifications tend to be quite fast once you've sorted the training set. In any case, I didn't say I thought it was appropriate, just that I took exception to the thought that computers are somehow inept in this regard.

You tried to use logoffsky to escape, they found your logoff location and were still there when you came back. Nothing wrong with that.

You also made an amazing statement: That data mining itself would mimic human judgement.

It won't.

It will just create metadata.

Yet, they are. I mean, you can keep saying "no, they're not" but you've misunderstood what you're asking them to define, and so do not understand why you're wrong on this.

Bumping with malicious intent is fine, so what use would a simulation that showed whether it was accidental or deliberate be?

Metadata for procedural classifications = judgement

matching human results with a high degree of accuracy = mimicking
Dunno what you're trying to get at

If the result is trivial then making the computer produce it is trivial.

No, metadata is just data about data. It's still two steps of refinement away from actionable knowledge.

the fact that the categorization of data is not the result you're trying to mimic.

You seem confused

Your understanding of classification also seems to be too narrow. You could, for example, consider the server logs over some time period the data set and harassment/not harassment the classification.

I, too, want a game that is barely functional with more than 2 people on grid because the server is logging every facet of every interaction

in the vain hope that one day heuristic analysis will be good enough to accurately determine human intent.

You're arguing something that is so removed from possibility that there's no logical objection someone can have to it.

The old addage is true here: If a computer is good at it, humans are bad at it. If a human is good at it, computers are bad at it"

No, my understanding of classification simply matches what we're talking about (the legitimacy of bumping). You could expand it, but then you'd be talking about something completely different and there would be even less reason to automate the decision-making. Nice attempt at moving the goalposts, though.

Judgement calls about harassment go beyond mere data. They're inherently case-by-case decisions, which are much better left to humans since they can judge intent and moderate their response as appropriate. That's where you went wrong from the very start: by assuming that it's a binary decision.

Hmmm it actually more interesting for me what he is using in the second video he posted here: https://forums.eveonline.com/default.aspx?g=posts&t=254190

Can someone explain me what happens after 2 minutes and 40 seconds?
If i see this correctly the autopilot says "initiate warp to gate"
and then seconds later his noobship lands at 0 on the gate....and jumps through...

I smell warp to 0 autopilot exploit....

Here the link if you are too lazy to click the post :P http://www.youtube.com/watch?v=aNZk7jBG7Ww

Jumping manually doesn't interrupt autopilot.

Really think about that.

If you gain suspect status by bumping someone...

- then every time you undock from a congested station (Jita 4-4) you will bump or be bumped. Everyone will gain suspect status and carnage will ensue.

Illegitimate bumping would presumably fall under harassment since the official stance is that bumping, in and of itself, is fine.

Implying that adding additional classes and associated responses makes the problem significantly harder?

Bumping is an exploit (wasn't designed | player used) , but one CCP overlooks (same goes for our ships bumping off of gates).

Can put a conditional in the code it's not rocket science.

- make an exception where ships won't go suspect if they are within a certain range of the station (in case of accidental bumping).
- make another exception where people within a certain range of the stargate won't go suspect (in case of accidental bumping).

WHOOPSIE-DAISY! Back to square one. People will be using the exception to bump people off of gates again (at least up to a point).

Jita already isn't like any other solar system in EvE. It's special, not equal.

Bumping is an exploit (wasn't designed | player used) , but one CCP overlooks (same goes for our ships bumping off of gates).

No. Implying that there is no classes, and (deliberately) no defined association to responses because both of those are better left to the non-binary adjudication of the human mind. Implying that action, reception and response are all subjective GÇö and that the devs explicitly wish to keep it that way so they can maintain a high degree of freedom, discretion, and not create any kind of ruleset that can will be gamed.

Read the rest of my post.

He landed on 0 at the gate. thats what im talking about. and the initiate warp sound came when he was tabbed out to another screen.
So there must have been some kind of bot program or something that warped him to gate at 0.

Humans have extremely specialized processing strengths that were evolutionarily driven; visual recognition is an example, creativity is an example, environment interaction is an example, reading over server logs is not an example.