Might want to fix your headline... Tom, not 'Ted' Gruber
Still the same by the time I got here. Proofreading is important.
also... artificial intelligence is a buzzword that still has yet to actually relate to actual intelligence. The best simulation of intelligence can't pass the most basic tests for intelligence. I wish there was a different term for this stuff so we could separate actual intelligence from cleverly made software tools that can't think.
The amount of information that gives rise to consciousness is so huge that silicon cannot handle this. An energy efficient biological machine is needed. All the rest is a cartoon of true consciousness.
Meanwhile we should not mix up intelligence and consciousness. There may be several types of intelligence that may come close to consciousness or not. What we need is "they" solve problems, not think.
Your first point was addressed above.
Your second point: How can intelligence exist without consciousness? How can intelligence exist without thought? Intelligence as pure data and no thought is not how i use the word. Your definition of intelligence does not match my own. Let me see what the dictionary says...
Number 5, which seems tacked on because of the pressure of the internet era, seems to defend your usage. At the very least, we are differing on the usage of "intelligence". I am using the definitions that involve thinking. I don't see how problems could be solved, or how systems can accommodate people without thought and thinking. I argue that the kind of intelligence people seem to be most fired up about involves thinking, which computers cannot do.
Might want to fix your headline... Tom, not 'Ted' Gruber
Still the same by the time I got here. Proofreading is important.
also... artificial intelligence is a buzzword that still has yet to actually relate to actual intelligence. The best simulation of intelligence can't pass the most basic tests for intelligence. I wish there was a different term for this stuff so we could separate actual intelligence from cleverly made software tools that can't think.
The amount of information that gives rise to conscience is so huge that silicon cannot handle this. An energy efficient biological machine is needed. All the rest is a cartoon of true conscience.
Meanwhile we should not mix up intelligence and conscience. There may be several types of intelligence that may come close to conscience of not. What we need is "they" solve problems, not think.
There's absolutely no proof that supports your assertion that silicon-based consciousness is impossible.
I believe he is right about the energy efficiency part, a problem that will be solved when the next paradigm shift in computation happens. We have a dizzying and growing array of new meta materials which will supplant silicon. That combined with the "new" electronics of quantum spins and optical pulses will advance computational power by an order of magnitude, so who knows how far the simulated intelligence might take us.
Metamaterials, huh? The more exotic the materials involved, the more extreme the conditions required for them to behave as desired, and the more difficult it is to actually MAKE those materials... the less likely we are to see the related technologies developed into any usable state. There seems to be a bell curve involved with the speed of advancements when it comes to electronics (and other areas of discovery) and we have reached the point where the rate of increase/improvement is seriously leveling off. All the "easily discovered" things have been discovered and physics shows us that there are hard limits to "more/faster".
Someone posted an article about quantum computing and artificial intelligence on facebook months ago. I argued back that the article was spurious, or the claims were, at least. The poster referenced Roger Penrose's quantum consciousness assertion, which i investigated and found to be just as problematic as the article. Here're a few excerpts from my responses: The article they link to for the "quantum memory problem" is called "Scientists Have Built a Functional ‘Hybrid’ Logic Gate for Use in Quantum Computers". Take a look at the "computer" here. It's a huge pile of machinery to manipulate the atomic states of certain elements. This is not a computer. In terms of quantum computing, this is the equal to when the first vacuum tube was being designed, NOT equal to when thousands of vacuum tubes were actually being used in a real computer. Computers like Colossus and ENIAC were actually programmable and functional computers using the vacuum tube to do counting. While taking up a room, they were actual practical applications of vacuum tubes to do math. The quantum computer isn't even at vacuum tube status. The difference being that manufacturing a vacuum tube is orders of magnitude simpler than manufacturing a Pentium CPU... which is orders of magnitude simpler than manufacturing a quantum logic gate.
Be careful what "science" articles you read. The headlines (and often the articles) promise advances that aren't merely a brief decade or two away and the assertions of what's coming next are just as spurious as the same kinds of crazy futurism predictions people made back in the 50s and 60s about what we would be enjoying in the year 2000...
Might want to fix your headline... Tom, not 'Ted' Gruber
Still the same by the time I got here. Proofreading is important.
also... artificial intelligence is a buzzword that still has yet to actually relate to actual intelligence. The best simulation of intelligence can't pass the most basic tests for intelligence. I wish there was a different term for this stuff so we could separate actual intelligence from cleverly made software tools that can't think.
The amount of information that gives rise to consciousness is so huge that silicon cannot handle this. An energy efficient biological machine is needed. All the rest is a cartoon of true consciousness.
Meanwhile we should not mix up intelligence and consciousness. There may be several types of intelligence that may come close to consciousness or not. What we need is "they" solve problems, not think.
Your first point was addressed above.
Your second point: How can intelligence exist without consciousness? How can intelligence exist without thought? Intelligence as pure data and no thought is not how i use the word. Your definition of intelligence does not match my own. Let me see what the dictionary says...
Number 5, which seems tacked on because of the pressure of the internet era, seems to defend your usage. At the very least, we are differing on the usage of "intelligence". I am using the definitions that involve thinking. I don't see how problems could be solved, or how systems can accommodate people without thought and thinking. I argue that the kind of intelligence people seem to be most fired up about involves thinking, which computers cannot do.
The proof that intelligence can exist without consciousness is all of those machines that can play Go, Chess, Jeopardy and alike. If you base your approach on the premise of "problem solving" you can immediately infer that such machines can be designed. You solve most of daily problems almost intuitively, it just happens in a snap. But complex problems cannot be solved that way. In order to solve an engineering problem you must call the math. Complex problems have some inherent math. In order to solve a complex problem you must mould it into some mathematical model. Once you perform such an abstraction, it becomes independent of the entity that will resolve it. You implement that abstraction on the silicon and what you get is called "AI" in pop-culture. I don't feel offended by the use of the term "intelligence" to define such an artificial entity. Actually it replicates some part of our "natural intelligence", the problem solving part. That problem solving part of ours is mostly algorithmic, and that algorithmic aspect can be replicated on the silicon.
Meanwhile, when you push the analysis deeper you come to a point where you observe that "mathematical understanding" cannot be replicated on the machine and is strictly tied to human existence. I refer you to Roger Penrose's "Shadows of the mind" on that. You can skip the math and just read the philosophy in that thick book. I find answers to many of my questions in his reasoning on "mathematical understanding" but I don't understand his and his colleague's "quantum consciousness" theory I find too simplistic.
Can computer code and silicon have reason and logic and consciousness? How does one program AI to know right from wrong? Where is the morality in all of this?
VERY interesting article and point of view on AI...
As a former (retired) IT pro I can say that I never once developed a system that replaced humans. Rather all of my systems made things better (mostly by improving information flow and making informed decision making possible). So, I can relate to what Gruber was saying...
I also like his examples of assisting those with dementia: I can easily picture various versions of AI assisting me in 20-25 years time with: -- managing my house (lights are on, doors are locked, etc) and cooking (food is cooked, stove is off, etc) -- managing my finances and bill paying -- managing medications -- self driving cars getting me around to where I need to be... -- and so on...
Currently, when an older person can no longer manage the daily tasks of living they are placed into a nursing home at enormous cost to them and to society. Far better for them and for society to help them live independently...
There is an interesting Black Mirror episode about folks who can remember everything they saw or did. Didn't turn out real well for them in the end because, well, Black Mirror.
Might want to fix your headline... Tom, not 'Ted' Gruber
Still the same by the time I got here. Proofreading is important.
also... artificial intelligence is a buzzword that still has yet to actually relate to actual intelligence. The best simulation of intelligence can't pass the most basic tests for intelligence. I wish there was a different term for this stuff so we could separate actual intelligence from cleverly made software tools that can't think.
The amount of information that gives rise to consciousness is so huge that silicon cannot handle this. An energy efficient biological machine is needed. All the rest is a cartoon of true consciousness.
Meanwhile we should not mix up intelligence and consciousness. There may be several types of intelligence that may come close to consciousness or not. What we need is "they" solve problems, not think.
Your first point was addressed above.
Your second point: How can intelligence exist without consciousness? How can intelligence exist without thought? Intelligence as pure data and no thought is not how i use the word. Your definition of intelligence does not match my own. Let me see what the dictionary says...
Number 5, which seems tacked on because of the pressure of the internet era, seems to defend your usage. At the very least, we are differing on the usage of "intelligence". I am using the definitions that involve thinking. I don't see how problems could be solved, or how systems can accommodate people without thought and thinking. I argue that the kind of intelligence people seem to be most fired up about involves thinking, which computers cannot do.
how about you take the first definition in the link you provided then? "The ability to learn or understand or to deal with new or trying situations." AI can do that TODAY. There's nothing wrong with the term "artificial intelligence" except (as is possible with every phrase) when people misinterpret it.
Might want to fix your headline... Tom, not 'Ted' Gruber
Still the same by the time I got here. Proofreading is important.
also... artificial intelligence is a buzzword that still has yet to actually relate to actual intelligence. The best simulation of intelligence can't pass the most basic tests for intelligence. I wish there was a different term for this stuff so we could separate actual intelligence from cleverly made software tools that can't think.
The amount of information that gives rise to consciousness is so huge that silicon cannot handle this. An energy efficient biological machine is needed. All the rest is a cartoon of true consciousness.
Meanwhile we should not mix up intelligence and consciousness. There may be several types of intelligence that may come close to consciousness or not. What we need is "they" solve problems, not think.
Your first point was addressed above.
Your second point: How can intelligence exist without consciousness? How can intelligence exist without thought? Intelligence as pure data and no thought is not how i use the word. Your definition of intelligence does not match my own. Let me see what the dictionary says...
Number 5, which seems tacked on because of the pressure of the internet era, seems to defend your usage. At the very least, we are differing on the usage of "intelligence". I am using the definitions that involve thinking. I don't see how problems could be solved, or how systems can accommodate people without thought and thinking. I argue that the kind of intelligence people seem to be most fired up about involves thinking, which computers cannot do.
how about you take the first definition in the link you provided then? "The ability to learn or understand or to deal with new or trying situations." AI can do that TODAY. There's nothing wrong with the term "artificial intelligence" except (as is possible with every phrase) when people misinterpret it.
Understanding is not a feature of AI. This is what differentiates "artificial" intelligence from "natural" intelligence.
Might want to fix your headline... Tom, not 'Ted' Gruber
Still the same by the time I got here. Proofreading is important.
also... artificial intelligence is a buzzword that still has yet to actually relate to actual intelligence. The best simulation of intelligence can't pass the most basic tests for intelligence. I wish there was a different term for this stuff so we could separate actual intelligence from cleverly made software tools that can't think.
The amount of information that gives rise to consciousness is so huge that silicon cannot handle this. An energy efficient biological machine is needed. All the rest is a cartoon of true consciousness.
Meanwhile we should not mix up intelligence and consciousness. There may be several types of intelligence that may come close to consciousness or not. What we need is "they" solve problems, not think.
Your first point was addressed above.
Your second point: How can intelligence exist without consciousness? How can intelligence exist without thought? Intelligence as pure data and no thought is not how i use the word. Your definition of intelligence does not match my own. Let me see what the dictionary says...
Number 5, which seems tacked on because of the pressure of the internet era, seems to defend your usage. At the very least, we are differing on the usage of "intelligence". I am using the definitions that involve thinking. I don't see how problems could be solved, or how systems can accommodate people without thought and thinking. I argue that the kind of intelligence people seem to be most fired up about involves thinking, which computers cannot do.
how about you take the first definition in the link you provided then? "The ability to learn or understand or to deal with new or trying situations." AI can do that TODAY. There's nothing wrong with the term "artificial intelligence" except (as is possible with every phrase) when people misinterpret it.
Understanding is not a feature of AI. This is what differentiates "artificial" intelligence from "natural" intelligence.
AI is apparently getting advanced enough to show the human trait of aggression when presented with circumstances it perceives (understands to be) stressful. Machines are also beginning to "understand" the value of certain learned skills, and how they might be re-used when encountering new challenges. I think think current machine "understanding" is a tad more advanced than you might realize, but still nowhere near as capable as even a baby.
Might want to fix your headline... Tom, not 'Ted' Gruber
Still the same by the time I got here. Proofreading is important.
also... artificial intelligence is a buzzword that still has yet to actually relate to actual intelligence. The best simulation of intelligence can't pass the most basic tests for intelligence. I wish there was a different term for this stuff so we could separate actual intelligence from cleverly made software tools that can't think.
The amount of information that gives rise to consciousness is so huge that silicon cannot handle this. An energy efficient biological machine is needed. All the rest is a cartoon of true consciousness.
Meanwhile we should not mix up intelligence and consciousness. There may be several types of intelligence that may come close to consciousness or not. What we need is "they" solve problems, not think.
Your first point was addressed above.
Your second point: How can intelligence exist without consciousness? How can intelligence exist without thought? Intelligence as pure data and no thought is not how i use the word. Your definition of intelligence does not match my own. Let me see what the dictionary says...
Number 5, which seems tacked on because of the pressure of the internet era, seems to defend your usage. At the very least, we are differing on the usage of "intelligence". I am using the definitions that involve thinking. I don't see how problems could be solved, or how systems can accommodate people without thought and thinking. I argue that the kind of intelligence people seem to be most fired up about involves thinking, which computers cannot do.
how about you take the first definition in the link you provided then? "The ability to learn or understand or to deal with new or trying situations." AI can do that TODAY. There's nothing wrong with the term "artificial intelligence" except (as is possible with every phrase) when people misinterpret it.
Understanding is not a feature of AI. This is what differentiates "artificial" intelligence from "natural" intelligence.
You must be using a different definition of "understanding." I would suggest that Watson "understood" the Jeopardy questions (answers) well enough to "deal with" the situation it was trained for. Do you disagree with applying the term artificial intelligence to what IBM is doing with Watson?
Might want to fix your headline... Tom, not 'Ted' Gruber
Still the same by the time I got here. Proofreading is important.
also... artificial intelligence is a buzzword that still has yet to actually relate to actual intelligence. The best simulation of intelligence can't pass the most basic tests for intelligence. I wish there was a different term for this stuff so we could separate actual intelligence from cleverly made software tools that can't think.
The amount of information that gives rise to consciousness is so huge that silicon cannot handle this. An energy efficient biological machine is needed. All the rest is a cartoon of true consciousness.
Meanwhile we should not mix up intelligence and consciousness. There may be several types of intelligence that may come close to consciousness or not. What we need is "they" solve problems, not think.
Your first point was addressed above.
Your second point: How can intelligence exist without consciousness? How can intelligence exist without thought? Intelligence as pure data and no thought is not how i use the word. Your definition of intelligence does not match my own. Let me see what the dictionary says...
Number 5, which seems tacked on because of the pressure of the internet era, seems to defend your usage. At the very least, we are differing on the usage of "intelligence". I am using the definitions that involve thinking. I don't see how problems could be solved, or how systems can accommodate people without thought and thinking. I argue that the kind of intelligence people seem to be most fired up about involves thinking, which computers cannot do.
how about you take the first definition in the link you provided then? "The ability to learn or understand or to deal with new or trying situations." AI can do that TODAY. There's nothing wrong with the term "artificial intelligence" except (as is possible with every phrase) when people misinterpret it.
Understanding is not a feature of AI. This is what differentiates "artificial" intelligence from "natural" intelligence.
You must be using a different definition of "understanding." I would suggest that Watson "understood" the Jeopardy questions (answers) well enough to "deal with" the situation it was trained for. Do you disagree with applying the term artificial intelligence to what IBM is doing with Watson?
Metaphorically it is "understanding" but internally it is not. Placing the question into the scope of some previously trained or learned information domains or finding the highest score match between a question and a list of possible answers is not understanding. None of these information domains or the sets of possible answers means absolutely nothing to Watson. There are just matches and Watson can never explain those matches as an eye cannot see itself. The superiority of Watson to humans is the speed at which it builds that list of possible answers. IBM was calling Watson as "machine learning" a couple of years ago, now they call it AI. I don't know, it is what they say it is.
Human understanding is not algorithmic. Algorithms may help and enhance understanding but they don't produce an output that can be defined as understanding. Human understanding is not an output, it seems like a state. Once we reach that mental state we say "understood".
Can computer code and silicon have reason and logic and consciousness? How does one program AI to know right from wrong? Where is the morality in all of this?
It isn't present because machines cannot think. Neither can many corporations, actually... ;-)
I feel like religion has hijacked the term "morality", so I prefer "ethics". I think there's somewhat a distinction between the two words, as well, as a result (morals being religious doctrine and dogma about what's sin or not, while ethics are societal rules that have the goal of protecting some kind of universal human rights regardless of the religious doctrines also found in the culture).
The ethics of computers is currently entirely in how people use them. People use computers for good, for bad, or neutrally.
It's kind of irrelevant, because software isn't an entity of its own. It doesn't do things without humans setting them in motion. You can write rules for human ethics just like you write rules for processing all kinds of other data. Until machines can actually think, there'll be no interpretation of those rules by the computer, only by the users (who violate ethics when it's personally profitable and we rarely see corporations held accountable for every breach of ethical conduct).
Once we have computers that can think, it would be nice to have something like Asimov's laws instilled in them, but we would have to teach computers what the meaning of "harm" is, and humans have a very inconsistent definition from society (and culture) to society (and culture). Aside from making hard-coded laws against harming humans (what about other animals??), we would have to teach these intelligences just like we teach our own kind. However, I suspect instinct would be absent from the software of intelligence. How much benefit or negative that would provide is unknown.
Might want to fix your headline... Tom, not 'Ted' Gruber
Still the same by the time I got here. Proofreading is important.
also... artificial intelligence is a buzzword that still has yet to actually relate to actual intelligence. The best simulation of intelligence can't pass the most basic tests for intelligence. I wish there was a different term for this stuff so we could separate actual intelligence from cleverly made software tools that can't think.
The amount of information that gives rise to consciousness is so huge that silicon cannot handle this. An energy efficient biological machine is needed. All the rest is a cartoon of true consciousness.
Meanwhile we should not mix up intelligence and consciousness. There may be several types of intelligence that may come close to consciousness or not. What we need is "they" solve problems, not think.
Your first point was addressed above.
Your second point: How can intelligence exist without consciousness? How can intelligence exist without thought? Intelligence as pure data and no thought is not how i use the word. Your definition of intelligence does not match my own. Let me see what the dictionary says...
Number 5, which seems tacked on because of the pressure of the internet era, seems to defend your usage. At the very least, we are differing on the usage of "intelligence". I am using the definitions that involve thinking. I don't see how problems could be solved, or how systems can accommodate people without thought and thinking. I argue that the kind of intelligence people seem to be most fired up about involves thinking, which computers cannot do.
how about you take the first definition in the link you provided then? "The ability to learn or understand or to deal with new or trying situations." AI can do that TODAY.
Examples of this, please?
What I see is software that can have data added or changed in an unintelligent manner. Example, autocorrect:
We don't "teach" autocorrect and it doesn't learn. Right now I'm struggling with how broken it is in Safari (completely inconsistent behavior elsewhere). It only gets more irritating over time because it cannot adapt to users (users get faster over time but the keyboard interpreter doesn't adapt to the "furthest reach" decreases seen in increased typing speed), it rejects complex vocabulary or language (because autocorrect wouldn't work at all if every possible word was anticipated, but maybe it should adapt the breadth of vocabulary for users who demonstrate a wider than average vocabulary), and because it adds unwanted words to its suggestion list.
Then there's Siri. It can barely respond to context at all. You can't "teach" it to pronounce your name correctly.
This is not "learning". Learning involves the ability to synthesize data into new and different usages and contexts. Software doesn't do this that I've seen.
I suspect you'll say these are poor examples. Please offer me some examples that you think demonstrate "The ability to learn or understand or to deal with new or trying situations."
Might want to fix your headline... Tom, not 'Ted' Gruber
Still the same by the time I got here. Proofreading is important.
also... artificial intelligence is a buzzword that still has yet to actually relate to actual intelligence. The best simulation of intelligence can't pass the most basic tests for intelligence. I wish there was a different term for this stuff so we could separate actual intelligence from cleverly made software tools that can't think.
The amount of information that gives rise to consciousness is so huge that silicon cannot handle this. An energy efficient biological machine is needed. All the rest is a cartoon of true consciousness.
Meanwhile we should not mix up intelligence and consciousness. There may be several types of intelligence that may come close to consciousness or not. What we need is "they" solve problems, not think.
Your first point was addressed above.
Your second point: How can intelligence exist without consciousness? How can intelligence exist without thought? Intelligence as pure data and no thought is not how i use the word. Your definition of intelligence does not match my own. Let me see what the dictionary says...
Number 5, which seems tacked on because of the pressure of the internet era, seems to defend your usage. At the very least, we are differing on the usage of "intelligence". I am using the definitions that involve thinking. I don't see how problems could be solved, or how systems can accommodate people without thought and thinking. I argue that the kind of intelligence people seem to be most fired up about involves thinking, which computers cannot do.
how about you take the first definition in the link you provided then? "The ability to learn or understand or to deal with new or trying situations." AI can do that TODAY. There's nothing wrong with the term "artificial intelligence" except (as is possible with every phrase) when people misinterpret it.
Understanding is not a feature of AI. This is what differentiates "artificial" intelligence from "natural" intelligence.
AI is apparently getting advanced enough to show the human trait of aggression when presented with circumstances it perceives (understands to be) stressful. Machines are also beginning to "understand" the value of certain learned skills, and how they might be re-used when encountering new challenges. I think think current machine "understanding" is a tad more advanced than you might realize, but still nowhere near as capable as even a baby.
Without specific examples, this doesn't tell me anything and I cannot respond with anything meaningful. It sounds to me as though you're anthropomorphizing software.
Can computer code and silicon have reason and logic and consciousness? How does one program AI to know right from wrong? Where is the morality in all of this?
It isn't present because machines cannot think. ....
You need to watch "The Imitation Game" where Alan Turing, the inventor of today's computers addressed that question very directly....
'Because, somebody or something thinks differently, does that mean that he/it is not thinking?' ... It's not clear whether he was speaking of computers or autistics -- but the question remains relevant.
Might want to fix your headline... Tom, not 'Ted' Gruber
Still the same by the time I got here. Proofreading is important.
also... artificial intelligence is a buzzword that still has yet to actually relate to actual intelligence. The best simulation of intelligence can't pass the most basic tests for intelligence. I wish there was a different term for this stuff so we could separate actual intelligence from cleverly made software tools that can't think.
The amount of information that gives rise to consciousness is so huge that silicon cannot handle this. An energy efficient biological machine is needed. All the rest is a cartoon of true consciousness.
Meanwhile we should not mix up intelligence and consciousness. There may be several types of intelligence that may come close to consciousness or not. What we need is "they" solve problems, not think.
Your first point was addressed above.
Your second point: How can intelligence exist without consciousness? How can intelligence exist without thought? Intelligence as pure data and no thought is not how i use the word. Your definition of intelligence does not match my own. Let me see what the dictionary says...
Number 5, which seems tacked on because of the pressure of the internet era, seems to defend your usage. At the very least, we are differing on the usage of "intelligence". I am using the definitions that involve thinking. I don't see how problems could be solved, or how systems can accommodate people without thought and thinking. I argue that the kind of intelligence people seem to be most fired up about involves thinking, which computers cannot do.
how about you take the first definition in the link you provided then? "The ability to learn or understand or to deal with new or trying situations." AI can do that TODAY. There's nothing wrong with the term "artificial intelligence" except (as is possible with every phrase) when people misinterpret it.
Understanding is not a feature of AI. This is what differentiates "artificial" intelligence from "natural" intelligence.
AI is apparently getting advanced enough to show the human trait of aggression when presented with circumstances it perceives (understands to be) stressful. Machines are also beginning to "understand" the value of certain learned skills, and how they might be re-used when encountering new challenges. I think think current machine "understanding" is a tad more advanced than you might realize, but still nowhere near as capable as even a baby.
Without specific examples, this doesn't tell me anything and I cannot respond with anything meaningful. It sounds to me as though you're anthropomorphizing software.
While I can certainly give you a link I think you'd learn more about it by looking for yourself. Briefly browsing a link will not give you a decent view of the bigger picture. There's been a quite a few recent articles at several blog and special interest sites detailing recent developments so they're not hard to find. Start with a simple search term like "AI learns to reuse skills" and go from there if you really have any interest in the topic.
Can computer code and silicon have reason and logic and consciousness? How does one program AI to know right from wrong? Where is the morality in all of this?
It isn't present because machines cannot think. ....
You need to watch "The Imitation Game" where Alan Turing, the inventor of today's computers addressed that question very directly....
'Because, somebody or something thinks differently, does that mean that he/it is not thinking?' ... It's not clear whether he was speaking of computers or autistics -- but the question remains relevant.
You don't have to be Alan Turing to think like that, men ask that question in their every interaction with women...
Joke aside, this is the wishful thinking of the writers.. In fact Alan Turing is the one who submitted one of the most powerful proofs against thinking machines: Halting Problem. Basically it is something like that: it is impossible to write a computer program that will determine whether it will get stuck or terminate succesfully with a given input.
Another proof comes from Gödel's Incompleteness Theorem. British scientist Roger Penrose has deeply handled both proofs in his famous book "Shadows of the Mind". This may not be available in pdf/epub but two papers from British philosopher J. R. Lucas are available online: "Mind, Machines and Gödel" and "The Godelian Argument".
Does failing to solve the Halting Problem mean that machines cannot ever be capable of reasoned response, "thinking" for all intents even if not actually sentient?
Does failing to solve the Halting Problem mean that machines cannot ever be capable of reasoned response, "thinking" for all intents even if not actually sentient?
Actually I'd like machines get that capability and that's all we need, nothing more. Only if that could be possible...
Suppose that machine learning will acquire some primary school level reasoning within a timespan of three years. That machine really talks with true reasoning and it is a very friendly and emotional companion to humans.
Then a startup emerges with a blazingly fast database solution and in three years it builds a companion able to provide university level companionship. That one doesn't use machine learning, it uses pre-programmed reasoning supported with continuously updated information vault, decorated with a speech-synthesis and natural language parsing module.
Which one would prevail? That is the problem of AI. As long as we commit resources and energy to an indeterminate AI journey, someone will always come up with more powerful conventional non-AI solutions to shadow our AI efforts.
Comments
Your first point was addressed above.
Your second point: How can intelligence exist without consciousness? How can intelligence exist without thought? Intelligence as pure data and no thought is not how i use the word. Your definition of intelligence does not match my own. Let me see what the dictionary says...
https://www.merriam-webster.com/dictionary/intelligence
Number 5, which seems tacked on because of the pressure of the internet era, seems to defend your usage. At the very least, we are differing on the usage of "intelligence". I am using the definitions that involve thinking. I don't see how problems could be solved, or how systems can accommodate people without thought and thinking. I argue that the kind of intelligence people seem to be most fired up about involves thinking, which computers cannot do.
Siri isn't speech to text, unless you're saying Apple is really confusing its naming of its computer products...
Metamaterials, huh? The more exotic the materials involved, the more extreme the conditions required for them to behave as desired, and the more difficult it is to actually MAKE those materials... the less likely we are to see the related technologies developed into any usable state. There seems to be a bell curve involved with the speed of advancements when it comes to electronics (and other areas of discovery) and we have reached the point where the rate of increase/improvement is seriously leveling off. All the "easily discovered" things have been discovered and physics shows us that there are hard limits to "more/faster".
Someone posted an article about quantum computing and artificial intelligence on facebook months ago. I argued back that the article was spurious, or the claims were, at least. The poster referenced Roger Penrose's quantum consciousness assertion, which i investigated and found to be just as problematic as the article. Here're a few excerpts from my responses:
The article they link to for the "quantum memory problem" is called "Scientists Have Built a Functional ‘Hybrid’ Logic Gate for Use in Quantum Computers". Take a look at the "computer" here. It's a huge pile of machinery to manipulate the atomic states of certain elements. This is not a computer. In terms of quantum computing, this is the equal to when the first vacuum tube was being designed, NOT equal to when thousands of vacuum tubes were actually being used in a real computer. Computers like Colossus and ENIAC were actually programmable and functional computers using the vacuum tube to do counting. While taking up a room, they were actual practical applications of vacuum tubes to do math. The quantum computer isn't even at vacuum tube status.
The difference being that manufacturing a vacuum tube is orders of magnitude simpler than manufacturing a Pentium CPU... which is orders of magnitude simpler than manufacturing a quantum logic gate.
Be careful what "science" articles you read. The headlines (and often the articles) promise advances that aren't merely a brief decade or two away and the assertions of what's coming next are just as spurious as the same kinds of crazy futurism predictions people made back in the 50s and 60s about what we would be enjoying in the year 2000...
The proof that intelligence can exist without consciousness is all of those machines that can play Go, Chess, Jeopardy and alike. If you base your approach on the premise of "problem solving" you can immediately infer that such machines can be designed. You solve most of daily problems almost intuitively, it just happens in a snap. But complex problems cannot be solved that way. In order to solve an engineering problem you must call the math. Complex problems have some inherent math. In order to solve a complex problem you must mould it into some mathematical model. Once you perform such an abstraction, it becomes independent of the entity that will resolve it. You implement that abstraction on the silicon and what you get is called "AI" in pop-culture. I don't feel offended by the use of the term "intelligence" to define such an artificial entity. Actually it replicates some part of our "natural intelligence", the problem solving part. That problem solving part of ours is mostly algorithmic, and that algorithmic aspect can be replicated on the silicon.
Meanwhile, when you push the analysis deeper you come to a point where you observe that "mathematical understanding" cannot be replicated on the machine and is strictly tied to human existence. I refer you to Roger Penrose's "Shadows of the mind" on that. You can skip the math and just read the philosophy in that thick book. I find answers to many of my questions in his reasoning on "mathematical understanding" but I don't understand his and his colleague's "quantum consciousness" theory I find too simplistic.
As a former (retired) IT pro I can say that I never once developed a system that replaced humans. Rather all of my systems made things better (mostly by improving information flow and making informed decision making possible). So, I can relate to what Gruber was saying...
I also like his examples of assisting those with dementia:
I can easily picture various versions of AI assisting me in 20-25 years time with:
-- managing my house (lights are on, doors are locked, etc) and cooking (food is cooked, stove is off, etc)
-- managing my finances and bill paying
-- managing medications
-- self driving cars getting me around to where I need to be...
-- and so on...
Currently, when an older person can no longer manage the daily tasks of living they are placed into a nursing home at enormous cost to them and to society. Far better for them and for society to help them live independently...
how about you take the first definition in the link you provided then? "The ability to learn or understand or to deal with new or trying situations." AI can do that TODAY. There's nothing wrong with the term "artificial intelligence" except (as is possible with every phrase) when people misinterpret it.
You must be using a different definition of "understanding." I would suggest that Watson "understood" the Jeopardy questions (answers) well enough to "deal with" the situation it was trained for. Do you disagree with applying the term artificial intelligence to what IBM is doing with Watson?
Human understanding is not algorithmic. Algorithms may help and enhance understanding but they don't produce an output that can be defined as understanding. Human understanding is not an output, it seems like a state. Once we reach that mental state we say "understood".
I feel like religion has hijacked the term "morality", so I prefer "ethics". I think there's somewhat a distinction between the two words, as well, as a result (morals being religious doctrine and dogma about what's sin or not, while ethics are societal rules that have the goal of protecting some kind of universal human rights regardless of the religious doctrines also found in the culture).
The ethics of computers is currently entirely in how people use them. People use computers for good, for bad, or neutrally.
It's kind of irrelevant, because software isn't an entity of its own. It doesn't do things without humans setting them in motion. You can write rules for human ethics just like you write rules for processing all kinds of other data. Until machines can actually think, there'll be no interpretation of those rules by the computer, only by the users (who violate ethics when it's personally profitable and we rarely see corporations held accountable for every breach of ethical conduct).
Once we have computers that can think, it would be nice to have something like Asimov's laws instilled in them, but we would have to teach computers what the meaning of "harm" is, and humans have a very inconsistent definition from society (and culture) to society (and culture). Aside from making hard-coded laws against harming humans (what about other animals??), we would have to teach these intelligences just like we teach our own kind. However, I suspect instinct would be absent from the software of intelligence. How much benefit or negative that would provide is unknown.
What I see is software that can have data added or changed in an unintelligent manner. Example, autocorrect:
We don't "teach" autocorrect and it doesn't learn. Right now I'm struggling with how broken it is in Safari (completely inconsistent behavior elsewhere). It only gets more irritating over time because it cannot adapt to users (users get faster over time but the keyboard interpreter doesn't adapt to the "furthest reach" decreases seen in increased typing speed), it rejects complex vocabulary or language (because autocorrect wouldn't work at all if every possible word was anticipated, but maybe it should adapt the breadth of vocabulary for users who demonstrate a wider than average vocabulary), and because it adds unwanted words to its suggestion list.
Then there's Siri. It can barely respond to context at all. You can't "teach" it to pronounce your name correctly.
This is not "learning". Learning involves the ability to synthesize data into new and different usages and contexts. Software doesn't do this that I've seen.
I suspect you'll say these are poor examples. Please offer me some examples that you think demonstrate "The ability to learn or understand or to deal with new or trying situations."
Without specific examples, this doesn't tell me anything and I cannot respond with anything meaningful. It sounds to me as though you're anthropomorphizing software.
You need to watch "The Imitation Game" where Alan Turing, the inventor of today's computers addressed that question very directly....
'Because, somebody or something thinks differently, does that mean that he/it is not thinking?'
... It's not clear whether he was speaking of computers or autistics -- but the question remains relevant.
Joke aside, this is the wishful thinking of the writers.. In fact Alan Turing is the one who submitted one of the most powerful proofs against thinking machines: Halting Problem. Basically it is something like that: it is impossible to write a computer program that will determine whether it will get stuck or terminate succesfully with a given input.
Another proof comes from Gödel's Incompleteness Theorem. British scientist Roger Penrose has deeply handled both proofs in his famous book "Shadows of the Mind". This may not be available in pdf/epub but two papers from British philosopher J. R. Lucas are available online: "Mind, Machines and Gödel" and "The Godelian Argument".
A video on halting problem:
Suppose that machine learning will acquire some primary school level reasoning within a timespan of three years. That machine really talks with true reasoning and it is a very friendly and emotional companion to humans.
Then a startup emerges with a blazingly fast database solution and in three years it builds a companion able to provide university level companionship. That one doesn't use machine learning, it uses pre-programmed reasoning supported with continuously updated information vault, decorated with a speech-synthesis and natural language parsing module.
Which one would prevail? That is the problem of AI. As long as we commit resources and energy to an indeterminate AI journey, someone will always come up with more powerful conventional non-AI solutions to shadow our AI efforts.