Thanks for your question. My opinion of A.I. is that it will become more and more powerful and bring computing power to a very high level to more and more people. But I don’t believe it can ever become sentient as many people fear (Like Elon Musk for example).
Also, I think SIRI and Alexa are currently laughably far away from anything remotely resembling real intelligence or sentience. Though such programs may get much closer to “feeling” like real people, I don’t believe they ever could be.
As far as potential dangers involved in A.I., I believe they mostly involve prematurely entrusting too many processes, or life or social functions to the dictates of a program, no matter how sophisticated.
The only reason we might launch ourselves into such a dangerous future is because A.I. will probably create increasing credibility in areas where we specifically take control out of the hands of people with greater and greater success. For example: with self driving cars. Or in flight controls. In some applications, the amount of data that a human operator has to sort through in a moment is just too great, and our machines can and will do a better job of it.
The Boeing Osprey one example – a hybrid helicopter/plane, the idea was brilliant, but it crashed a lot early on for various reasons, but often because no pilot could manage the changing flight characteristics between rotary and winged flight, especially in difficult conditions. The Osprey was about to be trashed before engineers developed an A.I. to manage flight control and now the aircraft is viable again.
However, we shouldn’t forget what a computer is. It runs algorithms. An algorithm is a program, a set of rules that intelligent agents provide, followed by problem solving operations. A computer is a really complex “if this, then this” machine. When attached to all sorts of inputs, like images, temperature, air speed, language, musical notes, math laws, or sound waves, programs can crunch numbers a lot faster and enhance many human functions that approximate intelligence or even personhood.
But a computer just runs programs, it doesn’t think for itself. Or, if you insist that future computers could have “thoughts” it never can have thoughts about its thoughts. If you fear a computer could somehow become conscious, imagine a computer inside a room, then ask yourself where its mind is located. You can’t do it. What will forever separate Mind from “A.I.”? Free will, and creativity. Innovation. There is an autonomy in real intelligence that A.I. can’t imitate.
Part of the fears of A.I. come from people buying into a materialistic worldview ideologically. Because I don’t believe consciousness is an emergent property of matter, I don’t believe it can be an emergent property of A.I. If you believe that all there is, is matter then you must believe that mind emerged from matter over long millennia. That’s just an article of faith – since no observation or science says it can. And if you buy that, as many do, then you will believe much more easily that the collated matter inside computers could spontaneously evolve into minds as well.
The glaring problem with this is that we become super confident in materialism at preciously the point where it’s the weakest. Evolution can tell us how genomes change over time due to shifting frequencies of certain genes due to survival differences in offspring. That’s it. Meanwhile, the facts show the gene itself is a book, it’s data! The gene is written in a 4 letter chemical alphabet, but it’s “meaning” transcends the chemical components it’s written on. That’s what information is. And right now, we don’t have a clue how information could arise spontaneously from non-living, random chemicals.
Nevertheless, if you are told enough times that information can come from random bits, and that free thinking entities came from non-living matter, we become susceptible to the idea that mind and consciousness could come from a Turing Machine!
Not possible.
We see this exact same assumption-set in our overblown fears about the Alien question. At preciously the point in the theory of evolution where it’s the weakest (a self-reproducing cell arising spontaneously from non-living chemicals), we imagine that this unsupported theory is in operation all over the universe producing life randomly. We don’t have any working model for how it happened here yet, but that doesn’t stop of us from postulating that it has to be happening all over the place. Come up with a viable way non-living chemicals self-assembling into self-replicating complex biological machines could happen here first, then I’ll believe the same process is making little green men all over the galaxy.
Science is now pointing to the fact that Mind underlays the universe, from its finely tuned laws to its irreducibly complex machines, to its code at the center of life. Atheist Thomas Nagel surprisingly makes this very point in his book, Mind and Cosmos. So it seems, Mind is fundamental, matter an emergent property of mind. To turn all that observation around and begin to think matter could generate mind is not logical.
I believe A.I. could kill us only if we prematurely entrust too much responsibility to our machines, but make no mistake, A.I. is just a machine. Machines that enhance preexisting human intelligence and abilities, using human language, bound by rules given by human developers and having all its goals set by free thinking, creative humans. The result is an obedient machine – like a car or a calculator – a really complex awesome machine, but a machine nevertheless. You can’t get a person out of these machines, even if they can produce music (which they can) or can outwit a Jeopardy champion (which they can).
The Discovery Institute challenges the prospect of “strong A.I.” here: https://evolutionnews.org/2018/01/robert-marks-on-the-lovelace-test/ and here: https://evolutionnews.org/2018/01/is-aivas-genesis-genius/