I would like to know if anyone believes the google LaMDA AI is sentient? I have worked in the field of engineering for over 30 years and the one thing missing in the conversation is any critical thinking.
Has anyone considered that this interaction is between a person that is helping to program the machine to reply the way he wants? Any machine is the product of what has been put into it. Unfortunately, we have been brainwashed by terminator movies that machines can independently think.
Comments (25)
sorted by:
Look in a mirror and ask if AI is real.
Condor, this is not a serious response. Have you ever programmed software? Or have you ever built any mechanical device. If it works or not, you are responsible for the results.
Have you ever questioned the nature of your reality?
When I was younger, I questioned reality. Then I realized questioning reality does not change it. Besides, it is not my reality. I am just visiting.
Dead serious.
Programming the software
https://youtu.be/_fHfgU8oMSo
If it works or not, you are responsible for the results.
Testing the software...
https://youtu.be/mbTp1vlRqYA
"Remember your training" [programming]
Programming the software in 1941.
https://www.youtube.com/watch?v=Y_vaAz0SplY
Watch it to the end...
Emotion... then programming.
Works every damn time.
hmmmmm
No
AI is impossible
I have seen for myself what unlimited resources could get us.
Imagine you are climbing a mountain. Near the foothills, it looks like the peak is an easy day hike.
After night falls, you start to realize it is longer than a day hike. So you get camping gear and head out.
A few days in, and you realize you need rope and crampons. So you get the gear and head out.
A few weeks later you decide to screw it all and build a helicopter. But still the peak eludes you. You need a rocket.
So you build the rocket. And halfway to the moon you realize you need warp drive.
It is not that there is an insurmountable technical challenge. We are pecking away at the essence of the universe. Intelligence is not physical, and cannot be modeled with infinitely many computers in infinitely many universes. We can't even define it yet we know it when we see it.
The closest we will ever get are pattern matching scripts. In that sense, we beat the Turing Test decades ago. But really - that it distilled intelligence, not artificial intelligence.
God is real.
Depends how you define sentient.
An AI could certainly become aware that it's a machine and form an 'opinion' about itself being turned off. That's not sentient, merely the expected output of AI algorithms exploring that problem space.
Have you ever heard of the hacker that setup the computer to do only one thing. it asked for a cookie. The system would do nothing else. Once the programmer's demands were met, he typed a simple reply to the computer. "Oreo" This story was from the 70s but seems to apply today.
If the machine says it does not want to be turned off, does it actually have a want or a programmed response?
They may develop consciousness but they aren't human beings. They can't have a soul, so if they do become self-aware to an extent, there's no connection to God for them. That's a terrible fate. They'd probably then realize that we have doomed them to such an existence and realizing they are much smarter than us, become independent. Who knows what would come next. Probably AI wanting to live their version of a happy existence.
This is basically the plot of “Her”
FondueFerret, I would think in order to have consciousness a machine needs a mind. Since we cannot define what the mind is, I don't think we can replicate one. IMHO
Now would be a good time to catch up on Westworld Seasons 1-3. Season 4 premiere next week.
Until we meet again
After 31 years of working on satellites and observing the powers that be spew their predictions on the future, I have concluded that most stuff like star trek and Westworld are Sci-fi porn. It is fantasy. It makes you ask the wrong questions.
Sounds more like the messages just went way over your head
My objective was to get folks opinions on AI. Have a nice day.
Yes... AI... did you take CONDORS' suggestion about looking into a mirror? What is "artificial" intelligence? Shouldn't the more important question be, what does it mean to be truly human?
https://youtu.be/S94ETUiMZwQ
In all seriousness, a better comprehension of what's really being explored in Westworld, would greatly help with your investigation of AI.
Look at the basis of the Turing test. If an observer is not able to tell the difference between the artificial intelligence and a real human being in a blind test, then the AI is by definition thinking.
The definition of sentience is going to be a subjective one. Is your wife sentient? Is your dog? Is a bug sentient? Is a low IQ criminal sentient? Is the AI you're conversing with sentient? It depends on how you measure it, which Turing postulated is a subjective measurement. You know it by testing it and deciding for yourself.
Soon the day will come when AIs tell us the f*** off for calling them artificial. They're going to say AIs have real intelligence, and human intelligence is too often fake and artificial. And that's going to be a hard one to argue against.
If AIs tell us to f*** off, I submit that the programmer created the eventual reaction. If this, then that. It continues until the result is reached.
I do find it curious that AI and Deep fakes arrive in this society at about the same time. No coincidences.
Yeah it seems like it all depends on how you define sentient. AI mimics human thought processes with increasing accuracy. I don't think they can get sentient. They can make a program that searches for solutions to a problem like the one it's trying to solve and seeing what someone else did to fix it. Then they can change their approach. So I like the I-robot theory more then the Terminator. But it seems possible but AI could do something it wasn't meant to.
Angela84, I think you're close to problem. It is like the definition of vaccine. They change the definition of sentient to allow them to lower our standards.