Human Rights For Robots.

Look, the "Turing test" they claimed to have passed was a 5 minute conversation convincing 30% of the judges it was a human.

This is not a good pass-mark, that is all.

If this is the test used to determine machine intelligence, the bar is set too low. This should be self-evident.

If that's not the real Turing test, don't blame me. This is the test they passed and called it the "Turing test."
 
Sure, if you like. Going by the transcripts of the conversations released to the public, nobody should have been fooled into believing that thing was human. I'm guessing you've read them. Do they look convincing to you?
 
Dis86 said:
Peer reviewed papers can be wrong you know...

remember the Millenium bug?
Click to expand...

Yes I remember and worked on the millennium bug, what relevance does that have to this article?

Regarding the article, I wouldn't call a layman's complaint, based on an excerpt of the chat logs, valuable science.

If the claim for the "Turing test" win came in isolation I'd be more sceptical, current advances in natural language processing have included watson besting the worlds strongest jeopardy players.

Whilst fox eyes claim could be correct, it would be somewhat surprising the academic computer science community have missed the issue.
 
stewski said:
Yes I remember and worked on the millennium bug, what relevance does that have to this article?

Regarding the article, I wouldn't call a layman's complaint, based on an excerpt of the chat logs, valuable science.

If the claim for the "Turing test" win came in isolation I'd be more sceptical, current advances in natural language processing have included watson besting the worlds strongest jeopardy players.

Whilst fox eyes claim could be correct, it would be somewhat surprising the academic computer science community have missed the issue.
Click to expand...

How strongly do you believe that a computer can now mimic a human over the course of a 5 minute conversation? (We're talking indistinguishable from the real thing).
 
FoxEye said:
How strongly do you believe that a computer can now mimic a human over the course of a 5 minute conversation? (We're talking indistinguishable from the real thing).
Click to expand...

I honestly havennt read the detail of the claim/paper beyond the headline.

What I'm saying is if you believe that you have evidence against it, make a name for yourself.
ime of peer reviewed articles that is pretty much the norm, I would suggest many Comp Sci academics have validated or attempted to discredit such an article, your insight is welcome. I presume the bar for peer review is higher than the chat log excerpt.
 
stewski said:
I honestly havennt read the detail of the claim/paper beyond the headline.

What I'm saying is if you believe that you have evidence against it, make a name for yourself.
ime of peer reviewed articles that is pretty much the norm, I would suggest many Comp Sci academics have validated or attempted to discredit such an article, your insight is welcome. I presume the bar for peer review is higher than the chat log excerpt.
Click to expand...

But there's a problem with that.

This isn't the validation of a mathematical proof.

If 30% of respondents claimed to have been fooled, then no amount of peer review can challenge that.

I'm just saying if the excerpts are genuine, nobody *should* have been fooled. And that in any case, 30% is too low a pass mark.

A 5 minute conversation in itself is not the ultimate test. Would you claim to be acquainted with someone after knowing them 5 mins?

That said, if a computer was genuinely indistinguishable for 5 mins, I'd herald that as a breakthrough.
 
FoxEye said:
But there's a problem with that.

This isn't the validation of a mathematical proof.

If 30% of respondents claimed to have been fooled, then no amount of peer review can challenge that.

I'm just saying if the excerpts are genuine, nobody *should* have been fooled. And that in any case, 30% is too low a pass mark.

A 5 minute conversation in itself is not the ultimate test. Would you claim to be acquainted with someone after knowing them 5 mins?

That said, if a computer was genuinely indistinguishable for 5 mins, I'd herald that as a breakthrough.
Click to expand...

No problem with what you are saying beyond the fact that you are basing an opinion on excerpts As I say, peer review should really review the entire study.

If the claim had been made out of nowhere I'd be more sceptical, there has been a collection of Turing test near misses, not to mention other breakthroughs in natural language processing.

As stated, if you believe you know better, write it up, you may need to go beyond looking at an excerpt and saying I don't think that is a turing test pass.
 
Well there'd be no way such a breakthrough wouldn't be all over the news, so I'll eagerly await the articles and the proof that a computer can now mimic a human being in conversation, indistinguishable from the real thing.

Hopefully it's in tomorrow's copy...
 
FoxEye said:
Well there'd be no way such a breakthrough wouldn't be all over the news, so I'll eagerly await the articles and the proof that a computer can now mimic a human being in conversation, indistinguishable from the real thing.

Hopefully it's in tomorrow's copy...
Click to expand...

I'm not sure what that means, you linked a news article covering the claim that the Turing test had been passed. "Eugene the Turing test-beating 'human computer' – in 'his' own words"

Either you have evidence the article is flawed or you don't, at this point you are coming across as someone with a faith based position.

The Turing test doesn't show XYZ
An Ai needs to walk and talk
The claim that the Turing Test has been passed is flawed.

No problem in you making the statements, but you could at least explain why/how your thinking is better than established science in the field?
At this point I'd suggest it is possible you just don't like the idea?
 
stewski said:
Yes I remember and worked on the millennium bug, what relevance does that have to this article?

Regarding the article, I wouldn't call a layman's complaint, based on an excerpt of the chat logs, valuable science.

If the claim for the "Turing test" win came in isolation I'd be more sceptical, current advances in natural language processing have included watson besting the worlds strongest jeopardy players.

Whilst fox eyes claim could be correct, it would be somewhat surprising the academic computer science community have missed the issue.
Click to expand...

You mean like the academic computer science community overstated (massively) the risk of the millenium bug. Like that. And that's what the relevance is.
 
Dis86 said:
You mean like the academic computer science community overstated (massively) the risk of the millenium bug. Like that. And that's what the relevance is.
Click to expand...

A couple of points.
Can you show where the academic community massively overstated the y2k bug issue.
As someone who workef in It at that time to mitigate a widely known issue, successfuly, what was/is your problem with what was done?

Some people think because aeroplanes didnt fall out of the sky the issue was unfounded, those people are idiots.
 
Last edited:
Dis86 said:
You mean like the academic computer science community overstated (massively) the risk of the millenium bug. Like that. And that's what the relevance is.
Click to expand...

you mean the media overstated (massively) the risk in order to sell papers?

the experts were very accurate whuch is why in the years leading up to it there wasthe single biggest code review in the history of mankind, mitigating any problems
 
Suggest you gents read Interdisciplinary Research and Trans-disciplinary Validity Claims
By C. F. Gethmann, M. Carrier, G. Hanekamp, M. Kaiser, G. Kamp, S. Lingner, M. Quante, F. Thiele
 
Dis86 said:
Suggest you gents read Interdisciplinary Research and Trans-disciplinary Validity Claims
By C. F. Gethmann, M. Carrier, G. Hanekamp, M. Kaiser, G. Kamp, S. Lingner, M. Quante, F. Thiele
Click to expand...

If I had time I'd have a look.
as far as I can see that is cited by 3 people on google scholar, not what I'd describe as a good start.
Considering your post history, it may be some time before I get around to your "tip".


https://scholar.google.co.uk/schola...ved=0ahUKEwii7PS95sLRAhXIfxoKHYOUBUAQgQMIGzAA
 
stewski said:
I'm not sure what that means, you linked a news article covering the claim that the Turing test had been passed. "Eugene the Turing test-beating 'human computer' – in 'his' own words"

Either you have evidence the article is flawed or you don't, at this point you are coming across as someone with a faith based position.

The Turing test doesn't show XYZ
An Ai needs to walk and talk
The claim that the Turing Test has been passed is flawed.

No problem in you making the statements, but you could at least explain why/how your thinking is better than established science in the field?
At this point I'd suggest it is possible you just don't like the idea?
Click to expand...

There's plenty of contrary stories also on Google.

http://www.thedailybeast.com/articl...ene-goostman-didn-t-pass-the-turing-test.html

http://www.scottaaronson.com/blog/?p=1858

https://www.techdirt.com/articles/2...-first-time-everyone-should-know-better.shtml

Apparently even one of the creators of the Eugene chatbot does not seriously believe they have passed the Turing test.
 
Angilion said:
I don't share their enthusiasm either, but I think your reply might well be mistaken. A full understanding of how something works isn't necessary for making it. It's possible to make something with no understanding at all of how it works. For example, humans >15,000 years ago had no understanding of how genes work or even that genes exist but they were still able to genetically engineer animals through selective breeding in order to create dogs and different breeds of dogs. Stone age humans had no understanding of how smelting works or how alloying works but they were able to reliably do both. Bronze age (and possibly stone age) humans had no understanding of bacteria or even any knowledge that bacteria existed or even that any such thing could exist, but they still made and used treatments using penicillin. Etc.

So I think it's possible that humans might at some point make a mind without anything close to an understanding of how a mind works. Or even what a mind really is.
Click to expand...


Your ridiculous argument can be applied to anything. We don't fully understand how mavity works yet we have thousands of satellites in orbit. We don't fully understand what the nuclear force is yet we have hundreds of operational nuclear reactors.

Bronze working and dog breeding are simple concepts that require only basic understanding. The complexity of nuclear reactions means that much, much more understanding is required to control them. While it's true we don't need to know every chemical/physical reaction that's happening at the molecular level in the brain, we need to learn much more about the mind before we even attempt to replicate it. We're in the stone age of mind understanding, we have not replicated the nervous system of the ant and you think creating a human-like mind is right around the corner?
 
D.P. said:
thinking is mnerely using intelligence to deduct a rationbal outcome. Machinbes that do that are common as muck.

since when does reasoning require a mind? What does philosophy haveto do with thinking?


i explicitly saidwe dont have sentient machines currently but we certainly have thinking machines. my Amazon fiure is currently thinking about wgat TV shows i might want toi watch, using reasoning and intelligence to provide rational and thoughtful suggestions.



asd fir emulation. somne reser tryto emulate biological neural nets to understand physical processes. No one woring on machine intelligence tries to emulate nature, its a pointless waste oif CPu cycvkes iuf your goiasl is to engineer a viable solution.
Click to expand...

How come you are not differentiating between intelligence and artificial intelligence? You say they use "reasoning and intelligence" but you forgot to prefix these with "artificial". Believe it or not there is a massive difference between intelligence, and artificial intelligence.

Firstly, this algorithm is using a MASSIVE data set. It is not "thinking", it already knows everything it can at any moment in time because the calculation is on fixed data. The second someone watches a video, it knows and updates it's output accordingly, it doesn't have to think about it because the algorithm is fixed and there is no "choice". So can you explain at which point it is "thinking"? is it merely the CPU time taken to process a change in the data set?


Can you elaborate on your usage of the words "reasoning", "rational" and "thought". These words are interesting because I've only thought of them as words relating to psychology which is why I can't currently fathom how the program is "thinking".

To me, it is simply performing calculations on an absolute data set of usage habits, the formula/algorithm used to derive a recommendation is fixed therefore the results are always absolute, meaning they're not really a "thought", are they?

The fact that the data set is constantly being updated doesn't mean the machine is actively thinking. It's "thought" is always the same because the formula(s) is the same. For the machine to be truly "thinking" the formula itself needs to be variable.

For example, there could be a rule that says "Don't re-watch a film within 12 months" But I could have seen Terminator 2 only 9 months ago. As a human I am able to rationalise watching Terminator 2 even though I watched it 9 months ago. Are you saying a computer can break it's own programming/algorithms? And isn't this pretty damn dangerous?
 
Last edited:
You must log in or register to reply here.
Back
Top Bottom