What are we doing to our kids?
-
dandy72 wrote:
It's very little more than the sort of parlor trick BS artists manage to pull off.
um not really, at all - could not be further from the truth. ChatGPT passes MBA exam given by a Wharton professor - University Business[^] https://healthitanalytics.com/news/chatgpt-passes-us-medical-licensing-exam-without-clinician-input[^] ChatGPT passes exams from law and business schools | CNN Business[^] ChatGPT is very real and will get better, faster, and more accurate every day going forward. The real danger of AI bots like ChatGPT is its eventual use as a people/political power control weapon.
Slacker007 wrote:
um not really, at all - could not be further from the truth.
Not true. For example from the first link. 1. It was one test 2. It was one class 3. It did not score perfect. Then the second link 1. The 'passing' score was just barely and that was 60%. 2. Hardly the only thing that goes in to becoming certified. 3. Text suggests this is not something new. They have run this test before and this is just the first time it got a score that high. 4. Why would it matter? There are studies that suggest medical errors are in the top 10 causes of death in the US. And it could be as high as the top three. So are you worried that the software might make the wrong choice?
-
Slacker007 wrote:
um not really, at all - could not be further from the truth.
Not true. For example from the first link. 1. It was one test 2. It was one class 3. It did not score perfect. Then the second link 1. The 'passing' score was just barely and that was 60%. 2. Hardly the only thing that goes in to becoming certified. 3. Text suggests this is not something new. They have run this test before and this is just the first time it got a score that high. 4. Why would it matter? There are studies that suggest medical errors are in the top 10 causes of death in the US. And it could be as high as the top three. So are you worried that the software might make the wrong choice?
Don't you see that ChatGPT is only going to get "smarter" with time? Don't you see that? it's passing all the tests, barely, but passing. It won't be long at all when it passes all the tests with 100% scores. Humans make silly mistakes, like forgetting to remove all the gauze from a site before sewing up. AI bots will not forget. I will be laughing at all of this, especially at you haters and doubters, everyday till I die.
-
Now, when ChatGPT can write essays better than school kids, and has answers to lots of questions, it seems to me that it could also answer the exam questions kid are getting in school or college. Of course, the AI proponents are going to praise this as proof of how "intelligent" ChatGPT is - it's so good, it could pass a college exam! But is it? Isn't it rather a poor comment of what nonsense we are doing in schools? Is schooling really meant to be repeating random facts, regurgitate what you have been told so you can spit it out again on an exam paper? Is this "learning"? If you think that's learning, THEN of course ChatGPT is "intelligent". Even Einstein apparently said "most of my work came from imagination, not logical thinking. And if you have problem with mathematics, I assure you mine are still greater." A school should prepare kids for life, give them some competence they can use, some knowledge they can apply, make them curious to create and use their imagination. Cramming data down their throat is, in my opinion, NOT what a school should do. It's just another example of how "automation" takes something away from humans. But is it really taking something away, or is it not rather pointing out that this was, after, not really human to do this stuff? Was it human to die as a slave while carrying stones to the pyramids in Egypt, or rowing the Roman boats? Certainly it wasn't - and now it's replaced by machines. It certainly created some unemployment, I guess - the real stupid people were then unemployed. But what business does anyone have to be stupid? That's where schools come in. But they, now, just make kids into parrots, easily replaced by chatbots. Maybe ChatGPT just points out that the "robotic" repetition really does not have a place in our schools. Something needs to change here, doesn't it?
I dunno. Take out loans. Go to grad school. Have debts forgiven. Live long and prosper.
"Before entering on an understanding, I have meditated for a long time, and have foreseen what might happen. It is not genius which reveals to me suddenly, secretly, what I have to say or to do in a circumstance unexpected by other people; it is reflection, it is meditation." - Napoleon I
-
I think you only have to look at how ChatGPT actually works - it tries to figure out the "best" next word in the sentence. That is NOT how to reason or think. Do you think that way? Certainly not - I would guess you have a CONCEPT first before you open the mouth. ChatGTP has no concept. It is just word babble. Thinking is not talking, no matter how many "scientists" may tell you that the way we think is through words. Einstein did not. And what about musicians? They don't think "now I need to put a F# semiquaver here in this position" (and if they do, their music is balderdash)
That is how it works now. You have to look past the now, and at the future. New versions, new updates, new branches. It eventually become what we all fear and know to be true.
-
nepdev wrote:
Is schooling really meant to be repeating random facts, regurgitate what you have been told so you can spit it out again on an exam paper? Is this "learning"?
For the most part the answer is yes. Despite what many might claim I see little evidence that formal education in its entirety teaches people to 'think'. Might get lucky in some cases but for the most part it is just a matter of learning the exact date of '1492', where the comma goes in the sentence and how to add and subtract.
nepdev wrote:
Cramming data down their throat is, in my opinion, NOT what a school should do.
Many others have that opinion as well. And there have been many attempts to provide an alternative way. But none of those work (they seem to work in very, very small tests but in larger rollouts they change nothing or even lead to more problems.) So certainly if you know a way that can provide revolutionary change to education then you should step up and start proving it and then popularizing it.
nepdev wrote:
Even Einstein apparently said "most of my work came from imagination, not logical thinking. And if you have problem with mathematics, I assure you mine are still greater."
The educational system targets the average not the above average. That is necessary because, by definition, most of the users (students) of the system are and always will be average.
nepdev wrote:
Something needs to change here, doesn't it?
Discard the calculator? Discard the keyboard? Discard the slide rule? Discard pencil and paper? Should students be taught solely by making new copies of religious texts? Or recognize those are changes that do in fact help students. If AI agents were capable of, in general, answering all questions correctly then why not use such a useful tool. Doing so does not directly lead to no one thinking - it is just another tool. However nothing suggests that AI agents are even close to being capable of providing correct answers to even most things. They provide answers to many things but the validity is often in doubt and is in fact often wrong. But in the same way if you rely solely on the results of a calculator to build a bridge you should expect that it will fall down (probably even before it starts being used.)
jschell wrote:
Many others have that opinion as well. And there have been many attempts to provide an alternative way. But none of those work (they seem to work in very, very small tests but in larger rollouts they change nothing or even lead to more problems.)
Hear! hear! If I may add a metaphor. A juggler who can keep many balls up in the air needs two things: the dexterity and the balls. Without the balls, only the dexterity will not help. During school years our brains acquire the ability to connect disparate facts in new and surprising ways but they need those facts. Knowing that they exist on Google or Wikipedia does not help. They need to be inside the brain so they can be rotated, rearranged, looked at from a different vantage point for in the end, with a bit of luck, the brain to come up with that magical thing, a novel idea. The theory of education pendulum has moved from rotten memorization to creative thinking but there is a golden midpoint where you teach young brains both facts and how to operate with those facts.
Mircea
-
dandy72 wrote:
It's very little more than the sort of parlor trick BS artists manage to pull off.
um not really, at all - could not be further from the truth. ChatGPT passes MBA exam given by a Wharton professor - University Business[^] https://healthitanalytics.com/news/chatgpt-passes-us-medical-licensing-exam-without-clinician-input[^] ChatGPT passes exams from law and business schools | CNN Business[^] ChatGPT is very real and will get better, faster, and more accurate every day going forward. The real danger of AI bots like ChatGPT is its eventual use as a people/political power control weapon.
If it's not linked, it's "unknowable". There will be a new movement to not record anything; we'll just exchange information with winks and nods so it can't be used.
"Before entering on an understanding, I have meditated for a long time, and have foreseen what might happen. It is not genius which reveals to me suddenly, secretly, what I have to say or to do in a circumstance unexpected by other people; it is reflection, it is meditation." - Napoleon I
-
Don't you see that ChatGPT is only going to get "smarter" with time? Don't you see that? it's passing all the tests, barely, but passing. It won't be long at all when it passes all the tests with 100% scores. Humans make silly mistakes, like forgetting to remove all the gauze from a site before sewing up. AI bots will not forget. I will be laughing at all of this, especially at you haters and doubters, everyday till I die.
Slacker007 wrote:
I will be laughing at all of this, especially at you haters and doubters, everyday till I die.
Agreed. I'm amazed by the number of developers and computer scientists (supposedly smart people) that are burying their heads on this one. Automation and robotics will be eliminating physical / manual jobs soon enough. AI will be eliminating MANY white collar jobs in roughly the same timespan. The world needs to figure out what to do with 8.5 billion idle humans.
-
Slacker007 wrote:
I will be laughing at all of this, especially at you haters and doubters, everyday till I die.
Agreed. I'm amazed by the number of developers and computer scientists (supposedly smart people) that are burying their heads on this one. Automation and robotics will be eliminating physical / manual jobs soon enough. AI will be eliminating MANY white collar jobs in roughly the same timespan. The world needs to figure out what to do with 8.5 billion idle humans.
AI bot haters and doubters remind me of this historical event: Get A Horse! America’s Skepticism Toward the First Automobiles | The Saturday Evening Post[^]
-
Now, when ChatGPT can write essays better than school kids, and has answers to lots of questions, it seems to me that it could also answer the exam questions kid are getting in school or college. Of course, the AI proponents are going to praise this as proof of how "intelligent" ChatGPT is - it's so good, it could pass a college exam! But is it? Isn't it rather a poor comment of what nonsense we are doing in schools? Is schooling really meant to be repeating random facts, regurgitate what you have been told so you can spit it out again on an exam paper? Is this "learning"? If you think that's learning, THEN of course ChatGPT is "intelligent". Even Einstein apparently said "most of my work came from imagination, not logical thinking. And if you have problem with mathematics, I assure you mine are still greater." A school should prepare kids for life, give them some competence they can use, some knowledge they can apply, make them curious to create and use their imagination. Cramming data down their throat is, in my opinion, NOT what a school should do. It's just another example of how "automation" takes something away from humans. But is it really taking something away, or is it not rather pointing out that this was, after, not really human to do this stuff? Was it human to die as a slave while carrying stones to the pyramids in Egypt, or rowing the Roman boats? Certainly it wasn't - and now it's replaced by machines. It certainly created some unemployment, I guess - the real stupid people were then unemployed. But what business does anyone have to be stupid? That's where schools come in. But they, now, just make kids into parrots, easily replaced by chatbots. Maybe ChatGPT just points out that the "robotic" repetition really does not have a place in our schools. Something needs to change here, doesn't it?
Me? I'm cashing out my home equity to invest in HVAC and septic. Everyone wants to be warm, cool and their toilets work.
Charlie Gilley “They who can give up essential liberty to obtain a little temporary safety deserve neither liberty nor safety.” BF, 1759 Has never been more appropriate.
-
jschell wrote:
Many others have that opinion as well. And there have been many attempts to provide an alternative way. But none of those work (they seem to work in very, very small tests but in larger rollouts they change nothing or even lead to more problems.)
Hear! hear! If I may add a metaphor. A juggler who can keep many balls up in the air needs two things: the dexterity and the balls. Without the balls, only the dexterity will not help. During school years our brains acquire the ability to connect disparate facts in new and surprising ways but they need those facts. Knowing that they exist on Google or Wikipedia does not help. They need to be inside the brain so they can be rotated, rearranged, looked at from a different vantage point for in the end, with a bit of luck, the brain to come up with that magical thing, a novel idea. The theory of education pendulum has moved from rotten memorization to creative thinking but there is a golden midpoint where you teach young brains both facts and how to operate with those facts.
Mircea
After having taught my children basic math, algebra, geometry and trig - I absolutely despise walking away from "rote memorization". The add/sub/mult/div tables dealt with the fundamentals of basic facts that got children over the details that actually allowed them to think. They get to Algebra - which is a fascinating time - and rather than wrestle with basic math, they can focus on the abstract concepts. Creative thinking as it were... that golden mid point. But Bureaucracy gets paid for elephanting basic concepts to prove they need a job.
Charlie Gilley “They who can give up essential liberty to obtain a little temporary safety deserve neither liberty nor safety.” BF, 1759 Has never been more appropriate.
-
I dunno. Take out loans. Go to grad school. Have debts forgiven. Live long and prosper.
"Before entering on an understanding, I have meditated for a long time, and have foreseen what might happen. It is not genius which reveals to me suddenly, secretly, what I have to say or to do in a circumstance unexpected by other people; it is reflection, it is meditation." - Napoleon I
-
Now, when ChatGPT can write essays better than school kids, and has answers to lots of questions, it seems to me that it could also answer the exam questions kid are getting in school or college. Of course, the AI proponents are going to praise this as proof of how "intelligent" ChatGPT is - it's so good, it could pass a college exam! But is it? Isn't it rather a poor comment of what nonsense we are doing in schools? Is schooling really meant to be repeating random facts, regurgitate what you have been told so you can spit it out again on an exam paper? Is this "learning"? If you think that's learning, THEN of course ChatGPT is "intelligent". Even Einstein apparently said "most of my work came from imagination, not logical thinking. And if you have problem with mathematics, I assure you mine are still greater." A school should prepare kids for life, give them some competence they can use, some knowledge they can apply, make them curious to create and use their imagination. Cramming data down their throat is, in my opinion, NOT what a school should do. It's just another example of how "automation" takes something away from humans. But is it really taking something away, or is it not rather pointing out that this was, after, not really human to do this stuff? Was it human to die as a slave while carrying stones to the pyramids in Egypt, or rowing the Roman boats? Certainly it wasn't - and now it's replaced by machines. It certainly created some unemployment, I guess - the real stupid people were then unemployed. But what business does anyone have to be stupid? That's where schools come in. But they, now, just make kids into parrots, easily replaced by chatbots. Maybe ChatGPT just points out that the "robotic" repetition really does not have a place in our schools. Something needs to change here, doesn't it?
-
Now, when ChatGPT can write essays better than school kids, and has answers to lots of questions, it seems to me that it could also answer the exam questions kid are getting in school or college. Of course, the AI proponents are going to praise this as proof of how "intelligent" ChatGPT is - it's so good, it could pass a college exam! But is it? Isn't it rather a poor comment of what nonsense we are doing in schools? Is schooling really meant to be repeating random facts, regurgitate what you have been told so you can spit it out again on an exam paper? Is this "learning"? If you think that's learning, THEN of course ChatGPT is "intelligent". Even Einstein apparently said "most of my work came from imagination, not logical thinking. And if you have problem with mathematics, I assure you mine are still greater." A school should prepare kids for life, give them some competence they can use, some knowledge they can apply, make them curious to create and use their imagination. Cramming data down their throat is, in my opinion, NOT what a school should do. It's just another example of how "automation" takes something away from humans. But is it really taking something away, or is it not rather pointing out that this was, after, not really human to do this stuff? Was it human to die as a slave while carrying stones to the pyramids in Egypt, or rowing the Roman boats? Certainly it wasn't - and now it's replaced by machines. It certainly created some unemployment, I guess - the real stupid people were then unemployed. But what business does anyone have to be stupid? That's where schools come in. But they, now, just make kids into parrots, easily replaced by chatbots. Maybe ChatGPT just points out that the "robotic" repetition really does not have a place in our schools. Something needs to change here, doesn't it?
In my teaching days (community college level), I created the final exams. Each problem was structured according to the Bloom taxonomy of learning. The students could bring any printed material to the exam - the course textbook or anything else. The 'a)' question asks for 'simple facts' that can usually be copied directly from the course textbook, 'knowledge'. 'b)' is for the student to show that (s)he understands the meaning of facts, 'comprehension'. 'c)' asks the student to demonstrate how the understanding of facts is used to solve a specific task, 'application'. 'd)' asks for an explanation of a suitable breakdown of a complex situation/system, 'analysis'. 'e)' asks the student to combine various elements / principles / ... into a larger, more complex whole to create a solution, 'synthesis'. 'f)' asks for a critical, 'professional' evaluation of some technique / solution / ..., 'assessment'. In recent years, many people put 'synthesis' (aka. 'creating') at the very top of the learning pyramid, 'assessment' (aka. 'evaluation') at the second level from the top. I tend to agree more with the original ordering: I have met lots of people - both in programming and in other fields - sprouting creative ideas like a fountain, producing lots of results, yet completely unable to do any sort of critical evaluation / assessment of either their own creations or the works of others. The other way around: In order to make a true assessment, you cannot be a stranger to the process of synthesizing a whole from constituents; you must master it quite well. Assessment is going a step further in mastering your field. It did take some practice & experience to create exams suitable to tell how far up the pyramid a student could climb, but after a handful of exams, as I got the grip on how to phrase the a) - f), I could quite easily see where the candidate was starting to fall off, not mastering that kind of questioning. Or, the candidate was a true master in the area presented in 1a) to 1f), but didn't handle problem 2) much higher than to 2d), clearly weaker in synthesis and assessment in that area. I am quite sure that ChatGPT would handle most problems up to the c) level quite well - although I guess I could fool it by 'trick questions' that would be handled by a human. For d) to f), it is not that difficult to phrase the problem statement so that a mechanical, robotic search would easily be revealed as a fake. If I were creating exams nowadays, I would
-
Slacker007 wrote:
I will be laughing at all of this, especially at you haters and doubters, everyday till I die.
Agreed. I'm amazed by the number of developers and computer scientists (supposedly smart people) that are burying their heads on this one. Automation and robotics will be eliminating physical / manual jobs soon enough. AI will be eliminating MANY white collar jobs in roughly the same timespan. The world needs to figure out what to do with 8.5 billion idle humans.
fgs1963 wrote:
The world needs to figure out what to do with 8.5 billion idle humans.
M.D.V. ;) If something has a solution... Why do we have to worry about?. If it has no solution... For what reason do we have to worry about? Help me to understand what I'm saying, and I'll explain it better to you Rating helpful answers is nice, but saying thanks can be even nicer.
-
I'll need to give that more thought. Personally, I'm unclear on what constitutes instinct anyway, so I may be a bit lost. As to choice, I'd still be unsure where to draw the line. For instance: When a pack of predators attacks the weakest members of a herd of prey, is that instinct or choice? Wouldn't instinct demand they attack the largest/meatiest? Is attacking the weakest members a learned strategy? This reminds me of "A Beautiful Mind". I think humans have probably lost much of the instinct our ancestors must have had and replaced it with learned knowledge. Maybe that's what makes the difference today, but there still must have been chooser-zero who had the ability and acted on it. Probably some bratty kid refusing to eat his mammoth.
PIEBALDconsult wrote:
Is attacking the weakest members a learned strategy?
The most amount of energy for the least amount of work, plus some morale/long-term thinking? Human babies could grow up and eventually help us so we let them be, some animals think very differently. (This sort of went dark awfully fast). At any rate: AI could definitely do that since its basically just math: Calculus of variations - Wikipedia[^]
-
Now, when ChatGPT can write essays better than school kids, and has answers to lots of questions, it seems to me that it could also answer the exam questions kid are getting in school or college. Of course, the AI proponents are going to praise this as proof of how "intelligent" ChatGPT is - it's so good, it could pass a college exam! But is it? Isn't it rather a poor comment of what nonsense we are doing in schools? Is schooling really meant to be repeating random facts, regurgitate what you have been told so you can spit it out again on an exam paper? Is this "learning"? If you think that's learning, THEN of course ChatGPT is "intelligent". Even Einstein apparently said "most of my work came from imagination, not logical thinking. And if you have problem with mathematics, I assure you mine are still greater." A school should prepare kids for life, give them some competence they can use, some knowledge they can apply, make them curious to create and use their imagination. Cramming data down their throat is, in my opinion, NOT what a school should do. It's just another example of how "automation" takes something away from humans. But is it really taking something away, or is it not rather pointing out that this was, after, not really human to do this stuff? Was it human to die as a slave while carrying stones to the pyramids in Egypt, or rowing the Roman boats? Certainly it wasn't - and now it's replaced by machines. It certainly created some unemployment, I guess - the real stupid people were then unemployed. But what business does anyone have to be stupid? That's where schools come in. But they, now, just make kids into parrots, easily replaced by chatbots. Maybe ChatGPT just points out that the "robotic" repetition really does not have a place in our schools. Something needs to change here, doesn't it?
"we" are doing the same thing to "our kids" that you are doing in the Lounge: increasing methane production.
«The mind is not a vessel to be filled but a fire to be kindled» Plutarch
-
I think you only have to look at how ChatGPT actually works - it tries to figure out the "best" next word in the sentence. That is NOT how to reason or think. Do you think that way? Certainly not - I would guess you have a CONCEPT first before you open the mouth. ChatGTP has no concept. It is just word babble. Thinking is not talking, no matter how many "scientists" may tell you that the way we think is through words. Einstein did not. And what about musicians? They don't think "now I need to put a F# semiquaver here in this position" (and if they do, their music is balderdash)
-
PIEBALDconsult wrote:
But that means that "something else" must separate humans (and probably our extinct proto-human ancestors) from "the lower animals" -- but not "intelligence".
The simplest concept of what that "something else" is, is the ability to choose. Animals generally respond instinctively (they can be trained to not respond instinctively, but that's still not choice.) We humans are unique in that we can choose not to respond by instinct. Ooh, that cake looks delicious, I'm going to eat it. Or, nice cake, but I'm watching my calories. Or I'm lactose intolerant so eating that would not be a good idea. Conversely, my cat loves to chew on certain plant leaves regardless of how many times he barfs them up later. Therefore, I would say that intelligence is making good choices based on knowledge and skill, and also making poor choices for reasons we are conscious of but choose to ignore.
Latest Article:
SVG Grids: Squares, Triangles, Hexagons with scrolling, sprites and simple animation examplesAs your cat does with the leaves.
-
fgs1963 wrote:
The world needs to figure out what to do with 8.5 billion idle humans.
M.D.V. ;) If something has a solution... Why do we have to worry about?. If it has no solution... For what reason do we have to worry about? Help me to understand what I'm saying, and I'll explain it better to you Rating helpful answers is nice, but saying thanks can be even nicer.
Humans feel most comfortable at a temperature of ~20-22 degrees Centigrade (293-295 Kelvin), and have a body temperature of 37 degrees Centigrade (310 Kelvin) Given 8.5 billion people, each of whom produces ~100W of heat, we have for the total usable energy: 8.5 * 109 * 100 * (310 - 295) / 295 = 43 GW. This is the total energy production of ten large power stations. From this, you need to subtract the energy required for growing & distribution of food and waste elimination for all those bodies. 'The Matrix' is not very efficient at power production. :sigh:
Freedom is the freedom to say that two plus two make four. If that is granted, all else follows. -- 6079 Smith W.
-
If it's not linked, it's "unknowable". There will be a new movement to not record anything; we'll just exchange information with winks and nods so it can't be used.
"Before entering on an understanding, I have meditated for a long time, and have foreseen what might happen. It is not genius which reveals to me suddenly, secretly, what I have to say or to do in a circumstance unexpected by other people; it is reflection, it is meditation." - Napoleon I
So the ubiquitous cameras will be used by the AI to learn the "winks and nods" language. Resistance is futile!
Freedom is the freedom to say that two plus two make four. If that is granted, all else follows. -- 6079 Smith W.