Holloway: There are hard, practical reasons why computers cannot understand concepts like “infinity” and “truth” and therefore cannot be conscious.
The GPT-3 program can get through grammatical issues on which others stumble, says Robert J. Marks. It is a simulation because the AI can perform the task but does not “understand” what the concepts mean: The classic test for AI common sense is resolution of Winograd schema. Winograd schema contain vague, ambiguous pronouns. Common sense Read More…
Calling these stem cells “self-reproducing robots” is like saying that humans create catbots when a pet cat produces a litter of kittens.
The friend who forwarded this story notes, “Even though we didn’t know maybe half of what’s in our cells, we somehow knew that most of the genome is junk?” Darwinism did that, of course. It was the Darwinians who needed the idea that most of the genome is junk.
Rees and colleagues assume that artificial intelligences can be creative thinkers. But the evidence so far is against that view. It’s not necessarily a matter of just ramping up the technology. By their very nature, computers compute but creative thinking is largely non-computational. We should keep that in mind when we encounter breathless media releases claiming to have overcome the problem.
The problem with getting AI to understand causation, as opposed to mere correlation, has led to many spurious correlations in data driven papers.
The Turing test, and the Lovelace test, are attempts to determine if computers can show human-like intelligence. Holloway asks, what happens if researchers succeed in creating lifelike machines? in the sense of “wanting” things? “If we create an all-powerful artificial intelligence, we cannot assume it will be friendly. Thus, we need a Terminator test.”
The next iterations of science fraud will employ machine learning trained on enough of the internet to avoid obvious goofs. We will need better, more sophisticated methods.
This is just a note for record on what monism is (as opposed to dualism, Creation by a Supreme and maximally great and good being, etc). A useful point of departure is a diagram from Wikipedia on dualism (and they give only one type) vs monism: Wikipedia notes, next to this: Different types of monism Read More…
The most likely reason one can think of for the persistence of computer-generated gibberish in the science database is that many other papers sound like that — but are in fact authentic human creations — so no one really wants to go there.
Let’s see what happens in China. This could be an important test of human exceptionalism.
Angus Menuge: I don’t see any reason from these amazing enhancements of the complexity of these [computer] systems to think that the systems would move from not having subjective awareness to having it or from moving to true intentionality about anything beyond themselves.
The problems of replicating oneself are addressed in a funny sci-fi short on human selfhood: For one thing, the replicant doesn’t know that he is not the original. He has no reason to think so.
Berger: “Man is an imperfect image of God, as we all regularly demonstrate. Some images are more degraded than others. Similarly, any image man creates of himself will be a less than perfect image of himself. Hence, man can never make AI that is in the image of God, He can only make a degraded image of himself. “
Computers require complete data to come to a correct conclusion but humans often work very well with incomplete data.