Is Artificial Intelligence Sentient?

a reply to: iamthevirus

What I meant was that there is an incompatibility between "thou shall not kill" and things like "nor shall be compelled in any criminal case to be a witness against himself, nor be deprived of life, liberty, or property, without due process of law", that implies that people can be "deprived of life".

originally posted by: TheRedneck
So the context must be programmed in, by (wait for it) a programmer... aka a human. Thus, any machine interpretation is ultimately the interpretation of the programmer who programmed it. There is no free will or conscious thought occurring inside those electronic components.

Isn't that what deep learning is supposed to do, to analyse large data sets and, in the case of language, "understand" their meaning and how to use it?

a reply to: neoholographic

Does it Contemplate " I Think , Therefore I Am " ? If Not , then No .

a reply to: Zanti Misfit

Probably not, but it knows how to say it so people believe it.

That's all it needs.

a reply to: ArMaP

Ah , Who Exactly Are these " People " that you Speak Of ? Names and Addresses Please.......

a reply to: Zanti Misfit

Those are identifiable data that cannot be shared because of the GDPR.

a reply to: neoholographic

The answer is still “No. AI is not sentient”.

Cross multiplying matricies is no more sentient than an actuary table is a horoscope telling you your future.

They are numbers. That is all.

The rest is widow dressing. Why is AI trained on fixed size data sets??

Because they can’t adapt to other sized data sets (think of mage software doing pattern matching at 600x800 pixels given a 4K picture of the same… the system hoarks!) That is all “AI” as of what we know about.

Seriously. We humans know nothing about what sentient behavior is let alone model it in code where it can “be” (or in the Biblical sense, abide)!!

Quantum computers don’t offer much support to the argument either because it is still matrix multiplication. Remember: matrix are set entities. The results are what are interpreted and we (humans) attribute qualities like “problem solving”, “intelligence” and “sentient”.

They are all anthropomorphisms of what we see in animals (including humans). And all that is because the brain is not a closes or set system.

Computers still are.

Some appeal to or tendencies to assign mental capabilities better than others but in the end, it is just a coded response.

Disclaimer: CS guy who gave up “AI” in 1990 because the brain was a black box (still is! But we have more clues. PS - It is now “cognitive computing” and spans several disciplines including coding, ANNs, and even brain wetware… the “how a brain functions”. And still, there are no answers that adequately explains any of it! But cool things have happened: they can design “eyes” that are tricked the same as humans; the “black box” has shrunk some; matrix math has gotten better; chip speed to simulate “AI” has gotten faster; data processing speeds are nearing milliseconds which is more human than the hours it took before. All taken together, we have more useful tools but that does not have anything to do with “sentience”).

Sorry bud!

If there was anybody rooting for a real AI then it would be I. But some 30 years later, I don’t see any progress.

Amalgamation paradox which also goes by several other names, is a phenomenon in probability and statistics in which a trend appears in several groups of data but disappears or reverses when the groups are combined. This result is often encountered in social-science and medical-science statistics.

en.wikipedia.org...'s_paradox

a reply to: ArMaP

In practical terms, these complexity results suggested that while Bayesian networks were rich representations for AI and machine learning applications, their use in large real-world applications would need to be tempered by either topological structural constraints, such as naïve Bayes networks, or by restrictions on the conditional probabilities.

en.m.wikipedia.org...

a reply to: iamthevirus

I was thinking more along the lines of the hypothetical quantum computing yet to come or whatever but I still wouldn't trust it to legislate that's for sure.

The mythology surrounding this mysterious "quantum computing" gets downright absurd at times.

Quantum computing differs from standard computing in two aspects: it is much faster than standard computing because it uses electron spin which is quantum entangled to communicate faster than light between components, and it has three possible digits instead of two. Binary can be represented as 1 and 0, but tertiary can be represented as 1, 0, or -1. That latter difference means a whole new software concept to take advantage, and we are just now, after 50+ years learning to imitate intelligence in a limited fashion using the old binary concept.

So I really don't put as much faith in quantum computing as many do. It's not that it won't work; it probably will. It's that the concept is so new it will take a good amount of time to compete with binary computing.

That's a huge problem with all of statistics and even climate modeling today, if it's programmed to find climate disaster of course that's what it's going to find, peeps don't seem to understand that.

Indeed; I have tried to explain that concept many times. Computer models are very successful in that they can perform calculations with extreme precision many multitudes faster than a human. But that's all they can do, and those equations must be programmed in by a human.

A good mathematician can infer much from an equation or a dataset just by looking at it. Even if said mathematician cannot specify exact numbers, he can see how different manipulations will affect the outcome. It is quite possible to examine an equation, compare the expected results to desired results, and use minor variations in the variables to determine the outcome. The computer then crunches the numbers, comes back with specifics, and those specifics show whatever trend the mathematician desired.

That's why any computer model used in an actual study must include the code used. If no code is specified, then no link exists between raw data and information obtained; it's just someone's opinion pulled from thin air.

TheRedneck

a reply to: ArMaP

Isn't that what deep learning is supposed to do, to analyse large data sets and, in the case of language, "understand" their meaning and how to use it?

Supposed to do? Yes.

Does do? Hardly.

TheRedneck

a reply to: iamthevirus

I thought that "thou shall not kill" would take precedence over the rest.

a reply to: TheRedneck

Ahh the quantum, overused and abused... everything quantum.

definitely cultish, I mean just look at YouTube, odd how science keeps looking for spirituality.

Maybe they'll eventually find some ha.

a reply to: iamthevirus

Science will eventually discover spirituality through the observed interactions between waves and particles, and the affects that electromagnetic frequencies, generated through conscious thought, imposes over matter.

They will discover a force of energy that causes all things to conform to the perfection of its design.

If you don’t believe in intelligent design then you are part of the dogma that enslaves the scientific community.

The full spectrum of the reality we live in is far beyond our concepts of programming or computing and will require a paradigm shift before any real progress can be made.

I read a lot of articles that compare computing with our physical and spiritual reality but human technology can’t even approach simulating the truth of our multi dimensional, eternal existence.

Computers follow code! nothing else! .. it would be simple to simulate sentience but sentience requires a soul and spirit.

The essential components of consciousness exist in several dimensions, not just sitting on some physical storage media that requires maintenance and an environmentally controlled enclosure.

Could you imagine AI created by Microsoft or Google .. patch after patch for eternity lol .. not to mention the risk of bad players attempting to substitute a fake, false God, in the shape of a simulated sentient AI in an attempt to hoodwink the Mass formation delusional, brainwashed trans-humans.

A simulation is a simulation is a simulation ..

.. AI will never become sentient!