AI, tech, and education
a year ago
https://picarto.tv/SimonAquarius for my livestreams, everyone's invited
https://www.patreon.com/simonaquarius to support my comic on Patreon
https://www.patreon.com/simonaquarius to support my comic on Patreon
Years back I remember teachers telling us we couldn't use calculators for some classes because we weren't going to always have access to one in future. The real reason was that we were learning the basics and they wanted us to memorize how to do it manually, because in later years we were given math to do but were also expected to use a calculator because it was difficult. I don't trust people to either spell words right or get math right if they don't use tools, because the internet's full of examples in their comments. Spell-check is built into everything and they still get it wrong.
Anyways, wikipedia eventually came out, and the immediate reaction was a distrust from the school system. Good reasons were had, as before, but when you're entering highschool and have seen two things shot down you start to question if the school system is up to date and trying to advance society or not.
Then AI comes out and the same familiar pushback is seen, and I think my viewpoint is solidified.
The school system wants people to be taught in a completely unbiased and factual manner, because teaching the students to question what they are being taught isn't part of their education until much later in life. If students lean on educational crutches too early then they aren't going to learn the basics, is what seems to be the issue.
The problem is that there's schools who are paid by religious organizations and corporations who both push for the inclusion of books that are inherently biased and teach viewpoints instead of things that meaningfully improves an individuals quality of life. If you teach someone math then they have the capability of becoming a merchant, if you teach them to read then they can learn from more than just teachers, but if you teach them to love oil companies then they only learn to accept other people making money while their own prospects are reduced in number, and if taught religion then they learn to be hostile to outsiders who might have made good allies or gained new technological insights other cultures provide.
With this, I think it's reasonable (in that we can use reason to understand) that the educational system dislikes things that will alter the foundation of society. But if it continues to push back against reality, then the result will be a loss of faith in the education system. This is an issue, because schools lean heavily on faith the age of the facilities creates to justify the cost of attending school there. If people lose faith then they might argue in general that education is good, but they might not be willing to pay for it if a better and cheaper alternative exists.
For a while schools also used the idea that an education would lead to jobs, but job hunting is mostly useless thanks to automation and nepotism making it impossible if you don't pay a service to apply for jobs for you. Companies ask for years of experience in fields that only existed recently, specifically because looking like they have openings means they can hire someone instantly the moment a position opens instead of creating the opening and then waiting for others to notice it (and if you the worker are wasting your time, who cares, the business didn't have to pay you for all the work you did applying everywhere, it's at your expense and not theirs).
AI's progress isn't what some expected. We're seeing things that at one time was predicted would exist a full century from now, but for everyone who only heard of modern AI in the last nine years this pace seems slow as hell, the quality is bad, etc. There's also the fear of cheap labor being favored by corporations when it comes to art of a certain quality. I've always hated advertisers, some are good, artistic, but then there's ads that blow out your eardrum when they come on, there's ads that take advantage of people who might not realize it's a scam, there's advertisers who put pressure on creators to not swear or show too much skin or any number of things. Of all forms of art, advertising has the greatest potential to do the most harm while the most people accept such harm is normal and expected. Which is why I'm not entirely against the education system's biases, because they (to my knowledge) ban advertising in schools.
Anyways, it seems the school system is banning smartphones again, some at least are, and I'm curious if this will increase people's faith in the education system, or worsen attendance as people question whether the current system is modern enough. Personally, I think this is more about how the parents are voters while children are not, and so there's going to be an increase of parents who believe schools are relevant and important for future worker production, and an increase of parents and children who believe schools are a waste of time that could be better spent learning from the internet and starting a small business at home. Who's more likely to be hired at a company that needs five years of work experience after all?
Anyways, wikipedia eventually came out, and the immediate reaction was a distrust from the school system. Good reasons were had, as before, but when you're entering highschool and have seen two things shot down you start to question if the school system is up to date and trying to advance society or not.
Then AI comes out and the same familiar pushback is seen, and I think my viewpoint is solidified.
The school system wants people to be taught in a completely unbiased and factual manner, because teaching the students to question what they are being taught isn't part of their education until much later in life. If students lean on educational crutches too early then they aren't going to learn the basics, is what seems to be the issue.
The problem is that there's schools who are paid by religious organizations and corporations who both push for the inclusion of books that are inherently biased and teach viewpoints instead of things that meaningfully improves an individuals quality of life. If you teach someone math then they have the capability of becoming a merchant, if you teach them to read then they can learn from more than just teachers, but if you teach them to love oil companies then they only learn to accept other people making money while their own prospects are reduced in number, and if taught religion then they learn to be hostile to outsiders who might have made good allies or gained new technological insights other cultures provide.
With this, I think it's reasonable (in that we can use reason to understand) that the educational system dislikes things that will alter the foundation of society. But if it continues to push back against reality, then the result will be a loss of faith in the education system. This is an issue, because schools lean heavily on faith the age of the facilities creates to justify the cost of attending school there. If people lose faith then they might argue in general that education is good, but they might not be willing to pay for it if a better and cheaper alternative exists.
For a while schools also used the idea that an education would lead to jobs, but job hunting is mostly useless thanks to automation and nepotism making it impossible if you don't pay a service to apply for jobs for you. Companies ask for years of experience in fields that only existed recently, specifically because looking like they have openings means they can hire someone instantly the moment a position opens instead of creating the opening and then waiting for others to notice it (and if you the worker are wasting your time, who cares, the business didn't have to pay you for all the work you did applying everywhere, it's at your expense and not theirs).
AI's progress isn't what some expected. We're seeing things that at one time was predicted would exist a full century from now, but for everyone who only heard of modern AI in the last nine years this pace seems slow as hell, the quality is bad, etc. There's also the fear of cheap labor being favored by corporations when it comes to art of a certain quality. I've always hated advertisers, some are good, artistic, but then there's ads that blow out your eardrum when they come on, there's ads that take advantage of people who might not realize it's a scam, there's advertisers who put pressure on creators to not swear or show too much skin or any number of things. Of all forms of art, advertising has the greatest potential to do the most harm while the most people accept such harm is normal and expected. Which is why I'm not entirely against the education system's biases, because they (to my knowledge) ban advertising in schools.
Anyways, it seems the school system is banning smartphones again, some at least are, and I'm curious if this will increase people's faith in the education system, or worsen attendance as people question whether the current system is modern enough. Personally, I think this is more about how the parents are voters while children are not, and so there's going to be an increase of parents who believe schools are relevant and important for future worker production, and an increase of parents and children who believe schools are a waste of time that could be better spent learning from the internet and starting a small business at home. Who's more likely to be hired at a company that needs five years of work experience after all?
FA+


bobingabout
WhiteChimera
Samhat1
MrSandwichesTheSecond
In the face of something like that, tools like calculators or Wikipedia are extremely tempting, as they promise to finish your task in record time. However it means you miss out on the actual lessons meant to be taught. Being able to work through problems in a logical manner, or to source facts from multiple places rather than taking the first thing for granted.
So I suppose the bigger problem is how to teach those things on top of the actual skills like math or facts of life like science, and gauging a student's absorption of such, in an accurate and engaging way without it devolving into "Do this thing 50 times and if you get it right most of the time you pass", which leads to the incentive for tool use as a cheat code which skips the actual learning in favour of getting the boring task done ASAP.
Smartphones I think are confiscated more for their potential for distraction than as cheat tools. Sometimes for tests and exams I recall them being called out, but in most circumstances it's because, due to things like videos and games being designed more explicitly for engagement than school lessons are, they often arrest a student's already tenuous attention. Which itself is a whole other problem the education system faces, in that even without cell phones it's difficult to serve every student's needs so they can learn in whatever way suits them, rather than the one-size-fits-all that often leaves folks behind simply due to its execution, rather than content.
The truly hardest thing to learn is the thing for which you don't know the name nor its existence. If schools can teach of the existence of an idea then all we need to do is learn how to find the things we can't name specifically but know are true.
Like, half the video so far has focused on how consumers might use AI, but I haven't seen anything about alphafold yet, which is a job I'd consider an actual thing only AI could do, provably. Sure, humans had been doing the same work for years, but the process is too slow, and by comparison the AI is insanely fast at solving these kinds of problems. And it's still improving.
Since 2015, how much have cars improved? How much has construction improved? Farming? The longer something is around, the slower the changes. Comparing AI from one year to the next shows it is getting better, but if you snapshot AI at present and compare it to reality, yeah it's going to suck. That's not why people are paying so much into it.
Businesses don't give a shit about how fancy tech is, my mom only recently stopped faxing papers to work because the business had no reason to switch off of that until covid screwed things up. The only things businesses care about are their stockholders and risk. Know why so many companies are trying to figure out if AI is useful? It's because if the competition discovers a use, then they will quickly out-compete the rest of the market who then has to start their research at the bottom. Everyone knows what happened to Kodak, and I think by now everyone knows Kodak invented the first digital camera and then decided not to sell something that could affect their source of revenue. That's why there's so many companies pushing AI products that suck, they are spending a fortune knowing that it won't see profits for decades maybe.
Picture for a moment a world where AI works as well as we hope it will, basically as smart as or smarter than a human in all fields, able to operate on its own to solve problems. Know what I'd use it for? Mining 16 Psyche, have it make more of itself and send those out to other similar asteroids. We're never going to accomplish this with human labor as quickly or efficiently as with automation. So how do we reach that point without first creating an AI that can look at sensor data and figure out things on its own? And even if we magically created an AI in one try that could do it all, how safe would it be, would we only find out after turning it on? I'd rather people, all people, be aware how AI can screw up or misinterpret things before we give it an order like "mine everything valuable in the solar system".
I do think governments are too hands off when it comes to regulation, but again, no government wants to be dealing with another nation that has AI piloted drones first.
If I was more tech savvy and looking for work I'd hire a bot service to flood the company inbox with resumes that lead nowhere but are otherwise perfect, then physically hand in a resume once it was clear they were no longer reading them all.
I think your last point is definitely a strong one, presently the way AI works is like a democracy, it can only choose what seems to be correct and not what is objectively correct. The issue is that it's taught based on what humans say and do on average, and on average people are wrong about something pretty frequently. Artists struggle with hands, so does AI trained on artists work, who'da thunk right?
But still, it's better now than it was a couple years ago, and within a decade we're going to have a harder time telling what's the work of AI. And that's only if there isn't a breakthrough that replaced the existing framework with something that's more reliable. Tech is like a house of cards, new discoveries topples the old one over and the funding is then funneled into the new stuff. So, at worst we'll have AI that's close to human level, at best there's a couple major shakeups before 2030.
I disagree with banning smartphones because learning to juggle work and play is important, and the students who can do both will excel. Students in every class ever are distracted by things, for me it was pokemon yellow, for past ones it was comics. There is no reward when a student instantly grasps a subject, they need to repeat things just as much as the other students. I used to eat lunches early because of this, and one day the teacher asked me to stop doing that. I asked if finishing my work early meant I could leave class early and read in the library, they said no and if I finished my classwork then I could do my homework. I told them that's what I already did, I could hand that in early if that meant I could leave class, they said no, so I said I'd eat my lunch then since I'd make more noise being bored for the remaining 45 minutes each day. And I ate lunch like that for the rest of my years there. If I'd had a smartphone I'd have used that instead of eating, but the easier solution would be rewarding students for handing work in early by letting them leave early.
I picture a school system where you can get that weeks homework immediately, finish it all, hand it in at the exit, and as long as it's not a test day be free to go home. The people who don't want to be there could accomplish the same thing by becoming overachievers as opposed to just not going to class. And if homework comes up that people don't know how to do, then they know it's an important class day that will teach them just enough to avoid staying in school for another week if they can master it immediately. And if they can't, then they can attend the classes and get more attention from a teacher who has less to deal with.
I thought that only in my country smartphones are banned in public schools in the classroom, not to mention calculators.
90% of my country's citizens attend public school from the age of 7 to 18, and then there are already universities and colleges or compulsory military service (for men), so I will not talk about chat rooms, because they are expensive and the children of rich deputies study in them.
I believe that school attendance is mandatory, but up to a certain point. To give a basic idea and basics to children under the age of 12 (this is from grades 1 to 5) This is a must. From the ability to read and count to the basic rules of language and mathematics. And then the child must choose what he will study until the age of 18, choose a narrow education profile, which our government cannot allow, therefore, until the age of 17 they try to teach the child everything, try to cram into him all the information that he needs and does not need, eventually killing interest even in what he liked and what he had a talent for.... for example, I loved literature and was able to write good and fascinating stories, but this love and talent killed me when I was forced to study boring novels for people over 50 years old.
As for new technologies.
It's great that things like neural networks have appeared and are developing rapidly, but they also need to be able to use them, and I think their training will be included in the computer science section, and schools will have only the basics, and universities will have more in-depth study.
But it all comes down to the reform of school education.
582 * 296 is 150,000 ish. Anyone should be able to do that estimate so they can sanity check their tool.
This kind of blind trust extends to more than just arithmetic, spelling, writing, etc. It extends to life as a whole. The media is full of outrageous lies and most people don't even question it because they simply trust the media in the same way as they trust their tools to do the work for them.
AI presents a similar problem. It can give hilariously wrong answers (e.g. Google Gemini recommending to add glue to your pizza crust for extra tackiness) and people just blindly trust it because they're not knowledgeable enough to know it's wrong.
When IQ tests were first created it took ages for people to realize stuff that's common at home isn't common everywhere in the world, there's lots of things that are basic concepts today that people back then didn't need to grapple with on a daily basis.
AI has existed in a functional form for all of ten years and people hate that it's still bad, and by bad they mean getting facts wrong, or looking at a photo and knowing what bird species is in it. For several decades prior there were contests to achieve even a little accuracy but it was one of those impossible tasks. It started as a student project or something for summer break, the professors thought it'd only take a short while to solve. We're in that era again, thinking certain problems will be easy to solve, getting it wrong, badly.
But now, totally can look at photos and tell you some of the contents in it. We've built it into AI systems to make prompting easier, and it's not limited to birds any more. Impossible into possible into useful is amazing progress.
Another example is people in cars driving on bike paths because their GPS said to take the path, not realizing they set it to walking/biking instead of driving. You could fault the device for not being clear, or you could fault the driver for not having enough training to use the machine they own.
I just feel that if we work out the kinks, AI is eternal and we are not. If we can create AI that drives better than a normal person with a map in their head, then we don't need to worry about bad drivers, just glitches that get rarer and rarer. Nobody wants to share a road with a student driver when it's an AI that may never reach human level capabilities before it's replaced, but that's probably not going to stay true forever. We just need to make it a little safer.