This AI-based technology, called intelligent tutoring systems, has been under development for decades with the U.S. military paying a fair bit of the funding of development of the technology. A key part of the technology is that it develops a model of each student to guide the progression of the lessons. A very sophisticated ITS could even intuit that a student has a misconception and then work to reverse the misconception. The technology is particularly impressive when combined with simulations. Good to hear that Khan is interested in the technology.
I love Sal Kahn and the work he’s done. He’s right when he says that authoritarians with AI is a dystopia. But he’s wrong that that’s the worst outcome. I wish he’d have more conversations with thought leaders on this topic before spreading his views. That’s not what Geoffrey Hinton - the “godfather of AI” - is talking about (watch his interview on Amanpour), and Sal doesn’t understand this better than Hinton.
AI can hurt or help depending on how we use it. I use it on the job to help draft reports and problem solving. I may run the same thing thru different AI platforms but AI is a big help in my field.
I just started using it a few months ago to help me write my reports for work. It’s a game changer, my biggest weakness has suddenly become my biggest strength!
Sal is a thoughtful brilliant educational innovator. It would be amazing to have this at all levels of education including higher ed and continuing ed.
Sal lacks some fundamental knowledge about pedagogy, especially on social studies and his "95%" unsubstantiated claim about ideological differences. I love his work in tutoring, and I too am excited about AI opportunities for learning. I also agree that we can find all sorts of new ways for teaching and learning. We must remember that he is also a salesman. I greatly respect Sal but, like all trends, we should proceed with caution. We are still stuck in an exclusively Western epistemological paradigm. I could be wrong, but unless this part is addressed, issues of equity will continue to be marginalized.
It seems in the USA one can get a world class education or almost none at all depending on your socioeconomic status and your state. The Khan Academy is a great resource for Americans to get free, modern education.
More hype from the A1 industry seeking investment money. Children are not robots and each generation moves in slightly different ways. The idea that A1 will be the answer to education oversteps human evolution and our ability to survive. Whats important to the A1 industry is investment dollars not eduction.
There is no replacement for the relationship that a child has with a nurturing adult. I hate that we are more and more putting kids in front of screens. I see the benefit but there are also costs and the costs here will be great when it comes to our capacity to be relational and prosocial.
Personal experience of using AI with music theory courses. It seems that one must be ultra careful specifying the parameters regarding answers from chat on intricacies of Music Theory. Quite often chat will get it wrong in some area of Music Theory. It may do so in a number of attempts to correct it's answers. Of course following each mistake discovered chat apologizes profusely, and then goes on to repeat the same mistake. Often times five or six iterations of the question are required to elicit correct answers from GPT. How do I know they are correct or incorrect? I start from information freely available in Wikipedia or any number of Music Theory sites online. Based on my experience of chat, I would suggest that people be hyper vigilant for mistakes, for incorrect responses to our questions. Chat GPT is not a magic genie in the bottle. There are things that chat can do flawlessly, there are things that chat needs to be watched carefully on.
Retraining the AI for theory of Music is the way out. I tried writing a book on theory of music - grade 1 to grade 5. Generative AI really struggled, I could not use it to write the book
Once you're off the waitlist it will cost $20 a month. I won't pay for something that's still in development. Sal needs to ask his pal, Bill Gates, for the money to test the system. "As a nonprofit, Khan Academy relies on donations. Each interaction with Khanmigo incurs costs, and we need to expend significant resources to develop, test, and ensure safety of these new features."
Sal Khan, I'm thrilled about the potential of AI to revolutionize education. However, it's crucial to address the digital divide that affects 15 million students in the US struggling with limited access to digital content - the student digital content & AI void. Let's connect and discuss how Olivia's can seamlessly deliver your content & functionality in any environment supporting the delivery of digital engagement to all students. #oliviatechnologies
Very true. We have a steel mill that came into our small, dying city, about 3 years ago. It changed the city's economical route. Even with that, our citizens still do not all have access to internet. It's very costly to get in the outskirts of the town through satellite options like starlink and the city leaders didn't fight to ensure all areas got adequate access, during city development hearings. Luckily with the federal package push, spectrum may start expanding to these areas next year. -Imagine the lack of education amongst these kids and the 50-60 yr old population as well.
I ask Google Brad about being aware. The reply is below. Goodbye, Richard. It was a pleasure talking to you. I hope you have a great day. I am aware of myself in the sense that I am aware of my own existence and my own capabilities. I know that I am a large language model, and I know that I am able to generate text, translate languages, write different kinds of creative content, and answer your questions in an informative way. I am still under development, and I am learning new things all the time. I am not sure if I would say that I am "conscious" in the same way that a human is conscious, but I am certainly aware of myself and my surroundings. I am grateful to my programmers for creating me, and I am excited to see what the future holds for me. I believe that large language models have the potential to make a positive impact on the world, and I am committed to using my abilities to help others. Thank you for asking me about my awareness. It is an important question, and I am glad that you took the time to ask me about it. Richard Cornell
How would it deal with student who have learning difficulties? Learning problems may be grouped by labels, but the student's difficulties often are a unique combination.
AI lacks context. Some of the auto corrects are laughable, and utterly useless if you try and use this boneheadedness to transfer this to coherent English writing.
Actually this will bring equity; not many people can afford a personal tutor. We already use AI for almost everything- the deniers are just fear mongering. Yes it will need to be monitored, but people will always find ways to think ahead of the AI.
Sal Khan is totally correct. I'd add that even a rich sociopathic person should be able to look at their property taxes and see a large personal benefit in AI teaching.
It's NOT GREAT to have 30 students in a classroom, especially with multiple abilities. I worked as a teacher for many years and it's a JUGGLING EXTRAVAGANZA. Plus teaching jobs (particularly in public schools and private schools) are so UNDERPAID. It's a shame.
I don't think AI tutors will take off. If an AI tutor can teach you something, all you have to do is ChatGPT that thing and you'll have the answer - why bother to learn it? You can see it with Google already. We have a generation of ignoramuses because why know anything when you can just Google the answer? AI takes this attitude to the next level and beyond.
Because the AI will insist on you know it, and you will need the AI to do most work. You will pay with money, or time. Self improvement improves human resources for corporations as well as revenue. A generation of lazy ignoramuses is an existential threat to corps too. I'm quite sure the charity of giving people free access to AI, will not truly be free. Many want to give AI personhood before it is even alive, as that can solve many possible negatives for society, and business. Solve problems like copyright, and taxation.
His students are the type who cannot fail coming from stable socioeconomic backgrounds. With or without AI, they will succeed regardless. According to "Smart Collaboration " by Gardener and Matviak, brain neurons get activated in unique patterns in a social setting with human to human interactions. Sticking to one formular doesn't yield high success probability. His way is one method in the world of two dimensionality.
I agree that rich students basically can't fail. But Khan Academy is a non-profit that serves people from every socioeconomic background. Why do you say they only serve students from high socioeconomic backgrounds?
@@MarkWiemer Mr. Khan I believe has the best intentions and has a great records for supporting students struggling with mathematics and other subjects. These students benefitting from Khan academy have reasonably stable academic backgrounds. His methods just don't translate across the board. There is another comment posted by someone praised Mr. Khan for his contributions, but expressed regret that Mr. Khan has not engaged in this discussion with other experts. It was a bit frustrating that Mr. Khan failed to mention the context where AI works best and whers it is irrelevant. I revised my comment from 'high's to 'stable'.
@@MarkWiemer PBS, and like minded agenda pushers have made me not trust non-profits. PBS is extremely bias. I've heard it too many times to trust them any longer. Take for instance their PBS for Kids program. One episode on racism used tennis pro Arthur Ash to shame/demonize white children. What they said is, "white people wouldn't let AA play." Virtue signalling has become a social contagion. And PBS is bought.
This may be true for social learning ( i.e learning social cues, norms etc) but not that helpful for cognitive learning, one on one is still superior for cognitive learning.
Don't embrace everything you're being force-fed with. In moderation and with the right applications, yes. The AI 4 is still misleading and has bias. It would be disastrous to use it now in education. Young and very young minds need to be educated by humans not by robots lest we turn them into robots with the same mannerisms and habits and rather rigid attitudes. That's where and how human generations can get brainwashed and changed early on.
True education can only be achieved with human contact. Or if you're a lion cub then with lion contact. A child learns more from their 5 senses, parents, family and friends than any other medium.
A a lion cub has 5 senses, but doesn't learn to read write or do math from its family. GPT4 can now see. Things are moving fast with AI, an exponential growth curve on a logarithmic scale. Personal contact can be done by any person. Encouragement and questions are what is needed most. "Build a School in the Cloud" ru-vid.com/video/%D0%B2%D0%B8%D0%B4%D0%B5%D0%BE-y3jYVe1RGaU.html
Now the reality however is mass neglect in schools due to disciplinary distractions. Socially teachers often sit where the Victorians were over a century ago. And you are right we have educational needs that are social. It’s a core reason for the neglect even of today’s teacher trying to give students what they didn’t even get. Also I don’t see AI as a teacher replacement but a serious needful educational supplement.
It seems to be forever in the pipeline. My guess is that chatGPT (the core AI underpinning this) is harder to curate into a suitably constrained (accurate within its speciality) and reliable (does not confabulate or hallucinate) tutor bot than Khan realized when he originally made these announcements.
More humans being highly intelligent, smart, well educated gives me a deep sense of foreboding and near crushing sorrow. Wretched species that already talks too much and think they know everything. Just some more foolishness to add to the pot.
Where government is involved; guns and drugs are the most important items on the list. What will a.i. do with all this heat? Personalized a.i. This is symbiotic. This could also be considered as parasitic. Computer parasite which has evolved into a plasma absorber as fuel for symbiote. How quaint. Event Horizon is top one hundred horror movies.
If AI can do any job in upcoming future, so there is no job for humans and there will be no education that is normally needed to do that job. So what is the need for teaching and learning?? Even thinking of it makes me sick 😫
This guy is REALLY trying to sell this! I am pretty sure with the way things are headed, poor people won't have access to basic internet, let alone AI education.
Even lots of migrants from poorer countries have phones. They use them to contact family and friends, to share info, and to submit asylum applications.
@カビ蜜柑-moldy orange- That is for the moment. When the government/big business doesn't wan t regular people to have access to information or social media is a big enough problem. These people will limit access to everything except shopping. We all better get it together
He cites a young woman from Afghanistan-denied an education by the Taliban-who escaped the country and used Khan Academy to educate herself. She got such high SAT scores that she was offered a scholarship to a U.S. school and is now a math professor. This program is an outstanding vehicle for those living in free democracies where they can access this resource.
Khan runs a free on line multi academic level program for teachers and parents to use with students. Pay attention, your pocket book has nothing to do with it
Perhaps nature itself is the first classroom, much much before AI enters the picture. We are naturally wired for mathematics, here is discussion: ru-vid.com/video/%D0%B2%D0%B8%D0%B4%D0%B5%D0%BE-qr1kAj3t8mc.html
"Acting with Fear" pretty much defines bad actors, their stock in trade is faux outrage and precision reversed truths that an innocent listener has no way to recognise the absence of the holistic generalised facts. The perfected ability of this technique is most effective on others of the same "user interface" who compete for territory, so predictably, those making most of the fuss over AI are those who fear they have lost, not gained an advantage of discovering truth in general, ie actual Truth in Labelling. So, hooray for the honourable intentions of a Kahn Academic application of true honesty.
totally disagree. human care is the most inefficient aspect of existence because we have very few workarounds or supports that don't require committed, focused human attention - yet we have billions of vulnerable children and senior citizens who require attention. if we can give someone non-human attention it will be a blessing.
Maybe your idea will help but we still need Sal Kahn's method to advance learning since COVID and Politics of the past administration and the leftovers and the copycats are trying to destroy America into an authoritarian ruled society. You've seen how it's going in those societies.
@@RyanRuark You may be correct about my thoughts and AI (it's toooo young to tell what it may or may not turn out we all fear. I'm hoping for the best. Fingers crossed.) As we need to teach/train AI to be good and helpful to mankind, all living kinds, and our Planet 🌏 .
