It's a mistake to think that thought experiments are frivolous without a foot in reality. For example the EPR paradox, Einstein-Podolsky Rosen experiment, turned out to be completely verifiable,using Bell's inequality among other things.
Would Data pass the Turing Test?
- Thread starter Trekker4747
- Start date
It's a mistake to think that thought experiments are frivolous without a foot in reality. For example the EPR paradox, Einstein-Podolsky Rosen experiment, turned out to be completely verifiable,using Bell's inequality among other things.
Irrelevant. I've worked with engineers. I wrote a procedure once that took five pages to get across what I had to do in my job. An engineer rewrote it so it said what he wanted, and it became a twenty page procedure. And that is my point. Engineers overthink everything. That's how Scotty got to be known as a 'miracle worker', he overthought stuff, but didn't put it in writing.
You're confusing over-thinking with thinking over , you probably forgot safety precautions that would prevent someone from getting killed or having his arm torn off. I've heard of a guy who got his head torn off, because he didn't think it was useful to put his long hair in a hat before visiting an area with working lathes in there.
I recall a moment where Brent Spiner was talking about his character, and how he learned that Data was connecting with autistic people because of how he communicated.
I doubt the questions themselves would be relevant - the test could concern the conversation partner's opinions on chocolate-covered manhole covers for all we care. The test would just be looking for the sort of thing 'bots here do: mindless quoting of phrases borrowed from the conversation partner, with minor, systematic modifications such as turning a statement into a question or challenging an opinion with a generic "arguing class" insult. Statistics on word and phrase usage. That sort of stuff.
Timo Saloniemi
Timo Saloniemi
That doesn't surprise me one bit. I know a few autistic people and sometimes they sound just like Data. I don't know where the writers got the idea that that is how a robot was supposed to act, just like making Lal move ridiculously her arms around like a hand puppet would make her seem more robotic. I guess they thought that Marcel Marceau was a robotic expert... These people should have consulted someone with actual knowledge of the domain but I guess it was too much to ask...
What would be a (very) few typical questions in a Turing Test?
Might help the discussion.
Anything you want. The goal is to trick the putative person into revealing his/her non human nature. So, the more surprising and unconventional, the better.
Definitely!
A lot of things about Data are an insult to common sense, like why does he use a keyboard since he could communicate his request by a direct (wireless) connection, a billion times faster. Or why does he need to look at a monitor since he could get the data, also a billion times faster the same way. Data shouldn't have to talk to the computer for the same reason...
Does it occur to anyone that you need emotion to care about having emotion? This is seen everyday among patients that receive medication that cancels out their emotions, because they are suffering ignominiously because of them, they don't care about anything, they are literally careless. So why would Data care about having emotions, if he doesn't have them?
Except he could - for example when telling jokes - and I don't think it has ever been established that he couldn't do it in written form if needed.
You mean he could write the contractions but was unable to pronounce them? That's even more ridiculous a limitation than I thought.
Data can apparently speak French, but French also contains contractions, not the same as English but still there are words that can be pronounced by omitting letters and sometimes whole syllables. Is Data also incapable of doing that in French? If so, why? If not, Why?
If the story demanded that Data take the Turing test or have already taken the Turing test then I am sure that he would pass it with flying colors.
I am sure they'd have made him walk on water if it was needed... As matter of fact, they did.
I was never under the impression that Data had a ongoing direct connection to the ship's computer, if he did then why the need to "plug him in" on occasion?
Was it ever stated in an episode that he had such a wi-fi like connection?
When Data would say "accessing" and stare briefly off into the distance, my take is that he was searching through his own internal records, and not the ship's computer.
It was never definitively said that Data can't speak in contractions, in fact there are several occasions when he does, and it's not always clearly a writing or performance mistake. Rather, it's just not the way Data naturally talks.
This is a pretty normal type of characteristic even for humans. You have people who just prefer very precise, measured phrasing and others whose speech consists largely of slang and more "loose" phrasing. It's not that a person who normally uses a lot of slang and relaxed speech can't purposefully talk in a more formal way, it's just not as comfortable for them. It doesn't come as easily, and they may not be as good at it. The same goes for more formal speakers attempting to be more casual.
Contractions are not the mark of sloppy English or slang and as a matter of fact you find them in the highest form of literature and poetry. You find them in Shakespeare to give a famous example. I hope I have dispelled your misconception.
People (e.g. Riker and Guinan) are dumbfounded to hear Lal use contractions and that clearly means that Data is incapable of doing them.
In Future Imperfect Riker takes Data making a single contraction as proof positive that he isn't what he appears to be.
These examples prove that those rare instances when Data does a contraction and it goes unnoticed are clearly the result of a mistake. There's no way you can explain them otherwise.
It's not a matter of "sloppy English", or that they aren't used in "high forms" of literature. You sort of went your own way on that. Less formal speech can be just as, or more, effective in expressing oneself. All I'm saying is that different people feel more comfortable with different forms.
Riker was already having some serious doubts that things weren't quite adding up in Future Imperfect. Data's failure to perform a calculation that should have been easy for him was an even stronger clue. His use of a contraction was just yet another hint that pushed Riker to explode.
Riker was already having some serious doubts that things weren't quite adding up in Future Imperfect. Data's failure to perform a calculation that should have been easy for him was an even stronger clue. His use of a contraction was just yet another hint that pushed Riker to explode.
It shouldn't have if Riker had witnessed Data making a contraction even once. That's the "I've never seen him doing that" feeling, that pushed him over the edge, that and that he's not too subtle a character.
What would be a (very) few typical questions in a Turing Test?
Might help the discussion.
The questions don't matter. The answers don't even matter, what matters is HOW the question is answered and whether the answer can pass as one being provided by a person.
As for the "highest score" thing it could make sense in a "how much Data fooled the judge" sort of way. I.E. have multiple judges and score based on how many were fooled. Data of all tested A.I.'s fooled the most judges.
Last edited:
Similar threads
