Deleted
Deleted Member
Posts: 0
|
Robots
Mar 29, 2020 6:32:28 GMT
Post by Deleted on Mar 29, 2020 6:32:28 GMT
|
|
|
Robots
Mar 29, 2020 9:10:41 GMT
Post by Retread-Retired-Cameron on Mar 29, 2020 9:10:41 GMT
I'm glad you're enjoying the book. I also hope the Texas Ranger is close enough to a regular police detective.
And thank you, it's always good to hear when a work is liked.
|
|
|
Robots
Mar 30, 2020 13:39:58 GMT
Post by And Kevin 2024 on Mar 30, 2020 13:39:58 GMT
Once the hurdle of creating the mechanics and software to enable a biped robot to stand up and walk (which Honda did many years ago at great expense) the other hurdle is creating interaction software that can pass the Turin Test. (Some actually cheat, because the voice is someone on a microphone somewhere). But a lot of research has discovered that people would not like to work alongside robots that seem human, or even just look human. But they would cost a fortune though, humans are cheaper and make themselves. Most modern day robotics is based around creating walking machines to do jobs too dangerous for humans, and of course in assembly in factories, which is nothing new. But do note that many of these intelligent appearing human-like robots, cannot also walk. www.digitaltrends.com/cool-tech/coolest-military-robots/
|
|
|
Robots
Mar 30, 2020 13:46:15 GMT
Post by And Kevin 2024 on Mar 30, 2020 13:46:15 GMT
|
|
|
Robots
Mar 30, 2020 18:31:43 GMT
Post by Retread-Retired-Cameron on Mar 30, 2020 18:31:43 GMT
Mr Lomas,
On one of my external drives I have a copy of an AI that has successfully passed the Turing Test. When someone told my copy they wanted it to die, the reply was "why do you want to kill me", and there was nothing in the data it had available to pull that reply from, it had to pick the words itself. That was twenty years ago.
One of the reasons I write is to offer the reader a chance to look at the world from an altered different perspective.
|
|
|
Robots
Mar 31, 2020 1:47:47 GMT
Post by And Kevin 2024 on Mar 31, 2020 1:47:47 GMT
On one of my external drives I have a copy of an AI that has successfully passed the Turing Test. When someone told my copy they wanted it to die, the reply was "why do you want to kill me", and there was nothing in the data it had available to pull that reply from, it had to pick the words itself. That was twenty years ago.
I would like to know what that is, because I have seen many recent ones tested on TV, and it's not long before AI is detected.
|
|
|
Robots
Mar 31, 2020 2:08:54 GMT
Post by And Kevin 2024 on Mar 31, 2020 2:08:54 GMT
|
|
|
Robots
Mar 31, 2020 9:38:48 GMT
Post by Retread-Retired-Cameron on Mar 31, 2020 9:38:48 GMT
On one of my external drives I have a copy of an AI that has successfully passed the Turing Test. When someone told my copy they wanted it to die, the reply was "why do you want to kill me", and there was nothing in the data it had available to pull that reply from, it had to pick the words itself. That was twenty years ago. I would like to know what that is, because I have seen many recent ones tested on TV, and it's not long before AI is detected.Well, for an AI to pass the Turing Test it has to have a really good cover story and natural sounding responses, which the old one did when it was tested. I believe it was posing as a 16-year-old girl.
The ones you've been seeing tested on TV either didn't have a very creative writer working on the cover story [personality, memories, and so on], had flawed programming, were intentionally set up to fail, or some variant combination of the preceding.
Now if you have the time on your hands, you can look up the ones you can play with at home, and with some time spent you can see how well you can get one to pass. As long as it isn't saying "Danger, Mr Lomas" you should have fun.
|
|
Deleted
Deleted Member
Posts: 0
|
Robots
Mar 31, 2020 10:30:31 GMT
Post by Deleted on Mar 31, 2020 10:30:31 GMT
|
|
|
Robots
Mar 31, 2020 11:00:13 GMT
Post by Retread-Retired-Cameron on Mar 31, 2020 11:00:13 GMT
|
|
|
Robots
Mar 31, 2020 15:11:16 GMT
Post by And Kevin 2024 on Mar 31, 2020 15:11:16 GMT
Well, for an AI to pass the Turing Test it has to have a really good cover story and natural sounding responses, which the old one did when it was tested. I believe it was posing as a 16-year-old girl.
A 16 year old girl of what intelligence? Many humans could possibly not even pass the test during general conversation!
The ones you've been seeing tested on TV either didn't have a very creative writer working on the cover story [personality, memories, and so on], had flawed programming, were intentionally set up to fail, or some variant combination of the preceding.
Not at all, are MIT incompetent, for example? But they were being tested by the likes of sociologists, linguists, and the like. It was concluded that eventually the AI can be caught out by a person of average intelligence. But "... The test results (Turing test) do not depend on the machine's ability to give correct answers to questions, only how closely its answers resemble those a human would give. …"
Now if you have the time on your hands, you can look up the ones you can play with at home, and with some time spent you can see how well you can get one to pass. As long as it isn't saying "Danger, Mr Lomas" you should have fun.
There are many Chatbots on line. It's not long before it's realised that they are. So what is the name of the software you have?
|
|
|
Robots
Mar 31, 2020 15:30:47 GMT
Post by Retread-Retired-Cameron on Mar 31, 2020 15:30:47 GMT
Mr Lomas,
Just call it AI software. I had the opportunity to check out the programmable version and a pre-programmed versions.
It actually took me a bit longer to break the pre-programmed version, and the "interrogation" lasted roughly 24 hours. The AI basically lasted about 6 hours longer than a living human would under similar conditions, before it got twitchy and eventually quit responding. After the program shut itself down it would no longer reboot.
As for the competence of people at MIT doing AI research, since I don't know any of them personally, I'm not qualified to speak as to their competence or lack thereof. I'm also not licensed to offer a professional opinion so I'll refrain.
|
|
|
Robots
Apr 2, 2020 16:07:22 GMT
Post by And Kevin 2024 on Apr 2, 2020 16:07:22 GMT
There's often times I think I may be chatting to AI, even face to face. As I said, I bet that non-face to face many humans would not pass the Turing test.
'The Turing test, developed by Alan Turing in 1950, is a test of a machine's ability to exhibit intelligent behaviour equivalent to, or indistinguishable from, that of a human. Turing proposed that a human evaluator would judge natural language conversations between a human and a machine designed to generate human-like responses. The evaluator would be aware that one of the two partners in conversation is a machine, and all participants would be separated from one another. The conversation would be limited to a text-only channel such as a computer keyboard and screen so the result would not depend on the machine's ability to render words as speech. If the evaluator cannot reliably tell the machine from the human, the machine is said to have passed the test. The test results do not depend on the machine's ability to give correct answers to questions, only how closely its answers resemble those a human would give.'
The bottom line is perhaps the most interesting one.
|
|
|
Robots
Apr 2, 2020 16:25:31 GMT
Post by Retread-Retired-Cameron on Apr 2, 2020 16:25:31 GMT
I am well aware of who Alan Turing was, what his profession was, as well as the reason why he committed suicide.
|
|
|
Robots
Apr 5, 2020 19:41:42 GMT
Post by ronmiller on Apr 5, 2020 19:41:42 GMT
Some interesting thoughts regarding the Turing test and some equally interesting results, all of which find both supporters and skeptics... isturingtestpassed.github.io/ I must point out that the author of this page is pretty skeptical about anyone who claims that the answer is "yes." One of the problems facing anyone taking the challenge is that Turing's conditions were not as rigorous as they probably ought to be...probably because the test was originally proposed as a kind of thought experiment. One problem, of course, is the issue of the evaluator him/herself and what subjective standards or bias they may have (for instance, just what would the evaluator consider a "human" answer?). Besides, since there is a 50/50 chance of merely guessing correctly which correspondent is the computer, there needs to be a more rigorous protocol established. For instance, told that one of the two people he or she is talking to is an AI, the evaluator will select one or the other as being the robot. But this selection would also occur if the evaluator were to be deliberately misled and both correspondents were in fact human. If something like that were to occur, it would certainly call into question the validity of the test and the effect of preconceptions on the part of the evaluator. www.theguardian.com/technology/2014/jun/08/super-computer-simulates-13-year-old-boy-passes-turing-testwww.theguardian.com/technology/2014/jun/08/super-computer-simulates-13-year-old-boy-passes-turing-testwww.zdnet.com/article/google-duplex-beat-the-turing-test-are-we-doomed/By the bye, one would naturally assume that a computer impersonating a 13-year-old boy or a 16-year-old girl would be impersonating one of average intelligence.
|
|