Donate to Remove ads

Got a credit card? use our Credit Card & Finance Calculators

Thanks to johnstevens77,Bhoddhisatva,scotia,Anonymous,Cornytiv34, for Donating to support the site

Passing the Turing test...

Scientific discovery and discussion
UncleEbenezer
The full Lemon
Posts: 10694
Joined: November 4th, 2016, 8:17 pm
Has thanked: 1460 times
Been thanked: 2965 times

Re: Passing the Turing test...

#511396

Postby UncleEbenezer » July 2nd, 2022, 6:18 pm

stevensfo wrote:I think that it would be more interesting to get people to chat with AI vs real humans, 'without' being told it was a Turing test.

That's precisely what happens in online encounters that aren't billed as a test.

Like the scenario I mentioned earlier in this thread.

This is a case of moving (and ill-defined) goalposts. The original Turing test has long since been overtaken by events.

9873210
Lemon Slice
Posts: 986
Joined: December 9th, 2016, 6:44 am
Has thanked: 226 times
Been thanked: 297 times

Re: Passing the Turing test...

#511418

Postby 9873210 » July 2nd, 2022, 8:09 pm

UncleEbenezer wrote:This is a case of moving (and ill-defined) goalposts. The original Turing test has long since been overtaken by events.

The Turing test was intended to sidestep the question of if a computer is thinking or just executing an algorithm or clicking relays. It is very much still relevant as shown by many of the responses to this thread.

No matter how ill-defined the Turing test is it is a paradigm of clarity when put against the arguments of people who believe in souls.

UncleEbenezer
The full Lemon
Posts: 10694
Joined: November 4th, 2016, 8:17 pm
Has thanked: 1460 times
Been thanked: 2965 times

Re: Passing the Turing test...

#511434

Postby UncleEbenezer » July 2nd, 2022, 10:45 pm

9873210 wrote:
UncleEbenezer wrote:This is a case of moving (and ill-defined) goalposts. The original Turing test has long since been overtaken by events.

The Turing test was intended to sidestep the question of if a computer is thinking or just executing an algorithm or clicking relays. It is very much still relevant as shown by many of the responses to this thread.


Indeedie. The test as originally stated is routinely passed by today's generation of bots. But people are uneasy with the implication that that makes them "intelligent", and so feel it necessary to move the goalposts.

The Turing test was of its time. It no longer solves anything if we resort to arguments such as seen in this thread and many other places.

No matter how ill-defined the Turing test is it is a paradigm of clarity when put against the arguments of people who believe in souls.


:lol:

Yeah. Bots are too dumb to be as stupid as humans. Or something.

odysseus2000
Lemon Half
Posts: 6366
Joined: November 8th, 2016, 11:33 pm
Has thanked: 1535 times
Been thanked: 959 times

Re: Passing the Turing test...

#511487

Postby odysseus2000 » July 3rd, 2022, 11:03 am

Is it possible, or will it ever be possible, to develop a test or series of tests that would objectively show that a machine is not sentient?

Or are we in the realms of subjectivity as applied e.g. to pornography where the legal framework seems to be that it is impossible to define but obvious yes or no in any particular case as viewed by many & a poll of responses tallied?

I am coming towards feeling that tests such as proposed by Turin & updated are no longer relevant & anyhow don’t address the issue of whether the ai is useful or not which in a practical sense is what matters.

I could be completely wrong & welcome comments.

Regards,

XFool
The full Lemon
Posts: 12636
Joined: November 8th, 2016, 7:21 pm
Been thanked: 2608 times

Re: Passing the Turing test...

#511501

Postby XFool » July 3rd, 2022, 12:12 pm

odysseus2000 wrote:Is it possible, or will it ever be possible, to develop a test or series of tests that would objectively show that a machine is not sentient?

Never mind a machine, how would you test that with another human being? (Or an animal?)

Obviously, in practice, we simply don't need to bring this up as an issue. The simplest assumption is that we share what we share - this was mentioned by Demis Hassabis in one of those Lex Fridman interviews.

odysseus2000 wrote:I am coming towards feeling that tests such as proposed by Turin & updated are no longer relevant & anyhow don’t address the issue of whether the ai is useful or not which in a practical sense is what matters.

I could be completely wrong & welcome comments.

As far as whether AI is "useful" etc. I suspect the Turing Test isn't really that important and in general it may be a bit of a distraction. After all, what is even meant in this context by "think" ?

I think I'd go with this:

Impracticality and irrelevance: the Turing test and AI research

https://en.wikipedia.org/wiki/Turing_test#Impracticality_and_irrelevance:_the_Turing_test_and_AI_research

odysseus2000
Lemon Half
Posts: 6366
Joined: November 8th, 2016, 11:33 pm
Has thanked: 1535 times
Been thanked: 959 times

Re: Passing the Turing test...

#511516

Postby odysseus2000 » July 3rd, 2022, 1:16 pm

Short 2 minute podcast where it is argued that consciousness & intelligence can be separate:

https://youtu.be/k-2t3WbRw8E

Regards,

odysseus2000
Lemon Half
Posts: 6366
Joined: November 8th, 2016, 11:33 pm
Has thanked: 1535 times
Been thanked: 959 times

Re: Passing the Turing test...

#511594

Postby odysseus2000 » July 4th, 2022, 8:48 am

Consciousness is computation (5 mins):

https://youtu.be/qgtUa7o0pKA

In disagreement with Sir Roger Penrose.

Regards,

UncleEbenezer
The full Lemon
Posts: 10694
Joined: November 4th, 2016, 8:17 pm
Has thanked: 1460 times
Been thanked: 2965 times

Re: Passing the Turing test...

#511684

Postby UncleEbenezer » July 4th, 2022, 4:48 pm

odysseus2000 wrote:Is it possible, or will it ever be possible, to develop a test or series of tests that would objectively show that a machine is not sentient?

Such tests would have to be devised around a definition of sentience. Which then leaves lots of questions for debate: how many species have sentience (and does it extend to plants where we can demonstrate reactions to stimuli having properties akin to our own reactions)? What about the "hive mind" of social insects? Etc.

The lesson of the Turing test is then that that definition will then survive only so long as it is unmet.

odysseus2000
Lemon Half
Posts: 6366
Joined: November 8th, 2016, 11:33 pm
Has thanked: 1535 times
Been thanked: 959 times

Re: Passing the Turing test...

#513423

Postby odysseus2000 » July 11th, 2022, 12:02 pm

Interesting short (5 mins ish) Joe Rogan podcast on whether Google’s AI is alive (strong language):

https://youtu.be/FEwt7XmDaSc

Regards,

XFool
The full Lemon
Posts: 12636
Joined: November 8th, 2016, 7:21 pm
Been thanked: 2608 times

Re: Passing the Turing test...

#516616

Postby XFool » July 23rd, 2022, 12:19 pm

Blake Lemoine: Google fires engineer who said AI tech has feelings

BBC News

Google has fired one of its engineers who said the company's artificial intelligence system has feelings.

XFool
The full Lemon
Posts: 12636
Joined: November 8th, 2016, 7:21 pm
Been thanked: 2608 times

Re: Passing the Turing test...

#516845

Postby XFool » July 24th, 2022, 5:55 pm

It's started! :shock:

Chess robot breaks seven-year-old boy's finger during Moscow Open

BBC News

A robot broke a seven-year-old boy's finger during a chess match in Moscow last week, Russian news outlets report.

"The robot broke the child's finger," Sergey Lazarev, Moscow Chess Federation President, told Tass news agency. "This is of course bad."

stevensfo
Lemon Quarter
Posts: 3438
Joined: November 5th, 2016, 8:43 am
Has thanked: 3810 times
Been thanked: 1398 times

Re: Passing the Turing test...

#517174

Postby stevensfo » July 25th, 2022, 9:17 pm

XFool wrote:It's started! :shock:

Chess robot breaks seven-year-old boy's finger during Moscow Open

BBC News

A robot broke a seven-year-old boy's finger during a chess match in Moscow last week, Russian news outlets report.

"The robot broke the child's finger," Sergey Lazarev, Moscow Chess Federation President, told Tass news agency. "This is of course bad."


The language and complete lack of detail in that BBC report is truly pathetic but after generations of dumbing-down in the media, I guess that people have become used to basing their opinions on 5 second sound bites.

The robot probably thought that the boy's finger was a chess piece and then the boy (seven-year olds do very silly things!) may have yanked his hand, causing the robot's clamps to tighten. The boy could have panicked and caused a small fracture. Remember that 'break' these days can mean anything from an almost invisible fracture to a true break of the bone. I should know. My little toes have a lot of experience. :(

Steve

PS Do the Russians have their own version of the Terminator? 8-)


Return to “Science”

Who is online

Users browsing this forum: No registered users and 8 guests