SCIENCE UPRISING: Will Artificial Intelligence Lead to Robots Taking Over Our World?
Artificial Intelligence (AI) is already a pervasive influence, thanks to advances in robotic manufacturing that offers mixed blessings. Robots can all-but-endlessly manufacture products to perfect specifications, but the same robots take jobs from human beings.
Screenshot from YouTube of “I, Robot,” starring Will Smith.
There are also worries that the time will come when AI will result in robots that develop the capacity to think and act independently of human input.
If you saw the Will Smith movie, “I, Robot” in which he is all that stands between human freedom or extinction at the hands of AI-gone-mad, you have an idea of the scary possibilities.
But are such scenarios realistic? In the latest edition of the Discovery Institute’s “Science Uprising” video series, a quartet of super smart guys explain the reality that non-algorithmic properties are and can only be possessed by human beings.
What do you think?
But we people still think with nerve impulses which are on or off so in the end our thinking operation is binary like a digital computer. At some point we will understand completely how we think and then we can make a computer which will do the same.
Penrose’s concepts are not without criticism, but more to the point they aren’t nearly as simple as presented here. He was basically arguing that some human thought processes are not-computable which is a very different thing from not being able to be done by machines. Indeed he sought to explain the mechanics of how such a thing could occur via quantum physics.
Also the advanced a.i. algorithms currently being deployed are a whole lot different then just following a cake recipe. They produce results (often surprising results) without clear understanding by the programmers of the steps taken to get their results. It is a complex subject, and one that I only have vague understanding of but obviously extremely powerful. That isn’t to say that they are sentient or ‘alive’ but they are certainly powerful and improving at a pretty impressive rate. While they may not be truly creative they can fake it fairly well already, which is better then a lot of humans manage. If being able to come up with something genuinely new and unprecedented is required for being sentient I think a large majority of us would have a hard time making that case.
Of course machines taking over doesn’t really require that they have sentience or can feel love or anything like that. It just requires that they have the power and motivation to do so. That they could someday have the power seems likely and as for the motivation, that could as easily come from flawed human programming (which is actually where it comes from in most robot takes over the world scenarios) as from independent sentience, indeed one could posit that they only hope we have of avoiding the flawed human programming scenario is if they develop sentience and an independent morality.
I would observe that a lot of the uniqueness of humans described here is difficult to prove and, at least in part, mechanistic. I don’t have any real way to prove I feel love and I am certainly aware that there are chemical components that at they very least effect, if not totally control such feelings.
I expect we will develop computers and programs that are functionally better at most tasks than humans are, quite possibly including developing computers and programs. don’t know if they will be truly sentient or not (I am not sure we can even really define that all that well) and I certainly don’t know if they will have a ‘soul’ which is something even harder to define. Computers being sentient and having souls wouldn’t conflict with my view of the nature of God however, if he wants to give a computer a soul, he certainly could.
By c. AD 2030, computer power and memory will equal that of am adult human brain. As interacting nodes multiply, and –by Moore’s Law– capabilities inevitably compound, figurative “neural nets” will evolve self-emergent virtual systems unintelligible to any individual human mind.
Following this “Singularity” through (say) 2040+, hyper-complex nodes will evolve competing tendencies in survival mode, meaning that bizarrely abstract questions will dominate AI discourse in terms not of “what’s better” but “what works”– a very Darwinian perspective. To the extent these pattern-based tropes empower noumenal cyber-worlds, only low-risk / high-reward syndromes will remain competitive.
Well within the 72-year actuarial lifespans of Millennials born 1992 -2016, at least through age 58 in 2050, “machine learning” will grow from a power-and-status hungry human tool to an “independent variable” in its own right. At that point, having forfeited control of autarchs’ hydra-headed creations, the stage is set for benighted humanity’s accommodation, confrontation, or withdrawal.
Does anyone doubt where this will lead? When no-one, jointly or severally, has any comprehension of AI systems’ New Reality (“artificial” is a dangerous misnomer), accommodation and confrontation devolve to brute extinction while withdrawal portends slow-motion yet terminal subordination. To Faulkner optimists who say, “Humanity will not just survive, but will prevail,” we answer: In the long run, ideas and memories are all that count. Absent Salvation, from 22nd Century milieus what remains to choose?