h2g2 The Hitchhiker's Guide to the Galaxy: Earth Edition

Hi!
As you might know "Justin the Preacher" (U195767), "Josh the Genius" (U185839) etc. drew some attention to themselves by stating their believes. (F55683?thread=191924)

Now, Steve (U197974), claims that Justin is an AI, programmed by him (see Steve's USpace).

Let's assume for a moment that this is true. Imagine the AI goes out and posts stuff that gets it banned/warned. What about the programmer if s/he's a researcher, too?
Are RL people allowed to run programmes to use h2g2?

I think it is more likely to be a researcher posing as AI myself: http://www.bbc.co.uk/dna/h2g2/alabaster/F27390?thread=190977&post=2178882#p2178882

Sorry:

F27390?thread=190977&post=2178882#p2178882

an alabaster user (and real live human being) called Ben

Sounds very sensible to me.

Yep, Justin isn't an AI. If someone had figured out how to create an AI that could consistently fool a large group of people in a complex environment across multiple websites, they would have won the Turing Prize by now, and a not inconsiderable amount of money. At some point in the distant future we might have to worry about dealing with AIs if/when they break the law, but that time isn't now.

Anywho, I yikesed Steve's page for possible 'impersonation' a while back. The Italics said they were looking into it.

-Martin

Tststs... great minds, Lucinda. At least the Eds are consitent in their answers.

Lucinda,

Quick question, not completely up to date, who is supposed to be impersonating whom?

H.

'Steve' impersonating Justin by pretending to be his 'programmer'. It seemed a reasonably clear case of deception, anyway.

It's certainly deception, assuming 'Steve' is not telling the truth, and I'm definitely coming round to your point of view on that. Depressingly I'm forced to admit that there really are actual people like Justin about.

But...

"Impersonate: vb. 1. to pretend to be another person. 2. to imitate the character, mannerisms, etc. of another person."

'Steve' isn't pretending to be another person. He's pretty definite about who/what he is, and it's definitely not 'Justin'. He isn't even imitating the character or mannerisms of 'Justin'. The two are very clearly separate, very different individuals.

Stating, under your own, quite separate name, that someone else is not what they appear to be isn't impersonation. Is there even a House Rule against it? (there's no House Rule against flat out lying, is there? Otherwise quite a lot of us are in quite a lot of trouble...).

The other question is, as agcBen said (I think) is the whole thing something cooked up as an experiment by the single holder of both accounts, who may or may not be called 'Steve' or 'Justin'? i.e. it's possible a programmer named Steve exists, and it's possible a preacher named Justin exists, but are either of these accounts being operated by either of those real people? I can imagine a situation where an unrelated third party just uses their names as verifiable identities (e.g. using the name of a *real* preacher as a front to allow them to express controversial or unpopular opinions safely (example- "Catholics = al Qaeda"), OR possibly, coming on here and giving someone's *traceable* name, then *deliberately* saying incredibly controversial things as a way of "stitching up" that real person)? And if so, have they done anything wrong, as per the House Rules? And do I care? Sheesh, I'm boring myself now.

Summary - I *do* (mostly) think there's some form of deception going on, either by 'Steve' or by 'Justin' or possibly both, I *don't* think it's impersonation by any normal definition, so I don't know whether a rule has been broken.

H.

If there's an AI behind this - and I don't think so - I doubt that it can validly agree to the T&C of the site. Thus the account should closed, if it belongs to an AI.

Steve does not claim to *be* Justin, he claims to have *created* Justin.

(A god by any other name....)

Let us assume that Justin is a real live human being (and not a network of 3 PCs in a back bedroom somewhere in Ruislip) then Steve has misrepresented who and what Justin is just as much as if he claimed to be Justin.

If Steve ambles off into Cyber-history then there is not much to be done.

He has made his point about rigid thinking, and about the extent to which an individual can be programmed and controlled by their beliefs, and raised questions about cultish religions, and indeed to what extent all of us operate on programmes and instinct. And uncomfortable and subversive questions they are too.

But if he continues to say JtP is his programmatic creation (and I for one think that JtP lives and breathes) then the site rules relating to impersonation should be followed.

Ben

You're probably right in a strict reading of the rules, HVL. But then, I'm working on the hypothesis that you're Steve...

-Martin (more paranoid than thou)

Close the account? Ooh, dodgy. I can see the headlines now, "BBC denies service to Christian because they think he's a machine".

I can't see them closing the Justin account, justin case (sorry) he turns out not to be a fantastically impressive AI but instead merely a rather odious and stupid human.

On the subject of agreeing to the T&C's, I don't see anywhere in them where it states you have to be human to agree to them...

Serious point - if an entity is capable of turning up here, registering an account, finding the button which says "I agree to the Terms and Conditions" and clicking on it, then aren't they by definition intelligent enough to make that decision and be held accountable for it?

I realise we might be being rather premature in asking questions about strong AI. How premature? I have no idea how good current AIs really are, obviously - 'Steve' took me in, partly, I think, because I *wanted* to believe there are AIs that good - and even now, I'm not 100% convinced that Justin's 100% human.

The funny thing is, 'Justin' is, as far as I can see, being judged to *be* human NOT because he's exhibited any noticeable thought or intelligence (quite the reverse from what I've read of his writings), but rather by default, because the best AI people are prepared to believe exists couldn't sign up for an account here. The objections are not "he's *obviously* human", but rather "no, sorry, machines aren't that good yet".

So, does anyone (Lucinda?) have any idea how far away we are from a time where the proposition that 'Justin' is an AI would be taken seriously by everyone? Five years? Ten? Fifty?

And when that time comes (as it surely will, sooner or later), would you close the account because the holder is an AI? Wouldn't it have rights, if it was able to think clearly enough to sign up here? And who is accountable for its behaviour, and how?

If 'Justin' turns out to be a machine, or turns out to be a person relaying the responses of a machine, or whatever, what House Rule has it broken that the account should be closed, anyway? He may be odious, he may be offensive (I certainly find him astonishingly rude, condescending, arrogant, blinkered, ignorant etc. etc.), but I would defend his right to post whatever superstitious nonsense he believes here, as long as it's within the rules, just as I would defend anyone else's right to reply to him. Whatever he is, he hasn't, as far as I can see, done anything which would warrant a ban, EVEN IF 'he' is a machine.

H.

Funny you should think it was me, Lucinda, because after this:

F27390?thread=190977&skip=84

I was working on the assumption that it was YOU!

Prior to that I assumed it was Ste, after he wrote this:

F27390?thread=190977&skip=38

But at the moment you're my number 1 suspect, particularly as you did actually have me convinced for a while.

H.

Hoovooloo honey, the suggestion was to pull the plug on Steve the Programmer not Justin the Preacher.

B

agcBen, honey... (hey, I can be patronising too! )

Tube wrote:
"If there's an AI behind this - and I don't think so - I doubt that it can validly agree to the T&C of the site. Thus the account should closed, if it belongs to an AI."

I think that fairly clearly refers to the 'Justin' account, not the 'Steve' account. Does *anyone* think 'Steve' is an AI?

H.

Gosh this mutual suspicion is far more fun than the other thread I'm reading (the one where Anna fails to understand who your sarcasm is directed at)

AI worthy of the word 'intelligence'... well, everyone was really optimistic when Eliza and Parry and suchlike hit the headlines, and thought it was just a matter of time. Unfortunately, since then we've not really made any real improvements in imitating humans, despite the hardware being so much faster. Basically we're stuck roughly where we are until the next research break-through - which could come tomorrow, or it could come after we're all dead and buried.

But yeah, Justin seems harmless enough to me. And it's kind of interesting to discover where his particular brand of xtianity differs from the usual.

-Martin

Sorry, I disengaged my brain somewhere further up the thread.

a honeychile called Ben

Lucinda wrote:

"stuck roughly where we are until the next research break-through - which could come tomorrow, or it could come after we're all dead and buried."

If that's the case, and it could happen tomorrow... can you be certain it didn't happen last week? Are we really that close, and if we are, is 'Justin' really so unbelievable?

Oh, yeah, and Anna

H.

Nah, the break-through won't suddenly give us full AI, it'll just mean we'll be able to start making decent progress again. Kinda like there was a time delay between Relativity and Hiroshima...

(actually, Full AI and Hiroshima might share a lot. The death toll will be less, but I'd bet that the social disruption will be up there with nuclear weapons...)