Just to be clear: it's not for every student (at least not yet!). We are in a research phase sharing it with a limited subset of users. More details about our approach to the responsible development of AI: https://blog.khanacademy.org/aiguidelines/
It better not be for every student. I am very familiar with Khan Academy as I am currently guiding a student through several AP courses on khanacademy. In my opinion, khan academy's time would be better spent fixing the UI for teachers, and improving the organization of the physics curriculum.
I would prefer that khan academy not be dragged into some PR fluff piece for some AI shill.
This argument has a pretty standard counterargument: organizations can and should do more than 1 thing at a time. It's a non-profit so you're probably better off contributing your ideas to them.
Do you know if/how someone can get involved with this at the research phase? I have a daughter who will be entering 1st grade next year and I'd be interested in having her try this out if there was a way of signing up.
I feel young again!! Logged into my account and an attempt to join the waitlist returns this:
"If you want to test Khanmigo, your parent or guardian needs to sign up.
At this time, only people ages 18 and older are eligible to sign up. Please ask your parent or guardian sign up and add you to their account."
I was a high school English teacher for about ten years before transitioning into tech last July. Has KA explored assistive assessment tools for in-person instruction?
I frequently find myself in the middle of many things, one of them is that although I oppose the grotesque plunder of the public and citizens of countries by neo-aristocrats claiming to do one thing or another amidst blatant lies of social welfare, I also object to slavery and therefore cannot condone that everything should be free since people like being compensated for their work, and not by being indentured and forced to do things.
Things have to cost, even if they wouldn’t have to cost as much if we could stop the psychopaths at the top constantly plundering and raiding the resources of the productive people to maintain their parasitic and decadent lives.
Are the tolerances in education really that tight?
How many terrible teachers are allowed to continue to teach after decades of disastrous results?
How many ideologues with no interest in teaching, but every interest in indoctrinating young minds, are tolerated because the alternative is "no teacher" and the class wouldn't run?
How many teachers who would fail state exams, teach, despite relying on answer sheets to be "competent"?
I agree with you that on the top end of education, this is no replacement and at best a supplementary tool. For the poor kid in a bad neighbourhood whose teacher is more interested in "de-colonizing" mathematics than teaching mathematics, this is a Godsend.
I feel the question is, should we be making this worse with unreliable AI.
Currently AI seem to have 2 weaknesses, it’s quite bad at reasoning and it hallucinates.
Yes, humans screw up reasoning too but at least they try. AI like ChatGPT seem to skip critical thinking altogether - there are examples of ChatGPT contradicting itself multiple times in a single conversation.
Then there are hallucinations. When people don’t know something, most just say they don’t know. AI just can’t seem to help itself but make up complete BS that it confidently try to pass off as facts.
P.S. Frankly I think the 2 might be related. AI can’t “sanity check” it’s own output.
> How many terrible teachers are allowed to continue to teach after decades of disastrous results?
Counterpoint: Why "hire" a known-bad teacher? Is a teacher that is known to give fake information better or worse than none at all?
I don't know the answer to this, but I hope that Khan Academy thought about this and decided that the good of GPT is better than it's bad. Or, they can't know so thats why they're doing a limited study before rolling this out.
My thoughts exactly. What is the baseline here? Teachers already give biased and/or inaccurate information at a rate much higher than I have seen in ChatGPT.
Is there a name for this fallacy (is it a fallacy?). I feel like we see the same issue with self driving cars crash rates. People are hesitant to adopt because they aren't perfect, but ignore that it's better than the status quo.
I'll don't think it's a fallacy. People are more predictable and can be accountable for their actions, and even admit ignorance, eg answering "I don't know ask someone else".
It's worth discussing if we want AIs replacing them - maybe the scalability is better, but these are usually worse.
> People are more predictable and can be accountable for their actions...
I would disagree. I think that at this stage, ChatGPT is very predictable, and we have excellent statistics on how likely it is to respond correctly to a particular type of question. I would have much less predictability about any particular human teacher, who might have an especially good or bad day.
Also, ChatGPT can be much more accountable, in the sense that when issues pop up, a filter can be implemented to prevent these particular issues for arising again. I don't see how humans are better in this regard - even if they are "accountable" in some vague sense, it can take many months to get rid of an underperforming teacher, and can take decades to change something across the system.
Heh, when Texas dictates what gets into schoolbooks.
"What happened between 1850 and 1960 regarding civil rights"
A: "Oh nothing, it was great for everyone"
History is a mess here. That said it would be nice if it were incorporated with written texts, lessons, and the AI so you can bounce questions off of it.
There is being misinformed and making a mistake when reasoning. Then there is making s** up.
Most humans have the decency to not do the latter while AI (at the moment) freely makes up complete lies when it doesn't have the answer instead of just admitting it doesn't know.
At least misleading instruction from teachers tends to be of a sort that others also experience, so can later be addressed en masse (e.g. mythologized history, poor-but-common explanations for things like how airplane wings work). Or spreading the same political propaganda that can be found all around, so is pretty easy to spot.
It's not usually that you ask them a question they can't answer, and they just confidently make up something entirely wrong, and possibly also an incorrect explanation that few or no other people have seen.
It's a different kind of wrong. One can write a book addressing common misconceptions picked up in school, from human teachers, and cover much of it. Does that work for lies AIs tell? They could be about almost anything, they could be broadly wrong, they could be subtly wrong, et c. "Show me some things my teachers may have been wrong about" seems to me an easier request to adequately address than "show me some things my AI tutors lied about", because the former tend to cluster around a kind of common space and form an experience with a great deal of overlap between students, while I'm not sure the latter do.
I suppose that this is true. But at least with an AI under your control (as a Khan Academy admin), you can run a query to analyze any unexpected decrease in student performance, find correlations with changes in the keywords that these students received from the AI tutor, isolate the particular misleading AI outputs by running some (possibly automated) experiments, and then once identified, fix the problem systemically by prompt engineering, fine-tuning or output filtering.
What's the analog of that with human teachers, and how long would it take to implement a solution there? Even if you had all the budget in the world, how would you go about making all teachers stop misleading students about how airplane lift works?
Firstly, the unreliability may not be permanent, and we may improve the AI accuracy in the future.
Secondly, we need to figure out what works and what not, and we're in the middle of such phase.
Finally, there is no need to be alarmed, even with hallucinations, the amount of advanced (and reliable) knowledge provided by the AI vastly counterbalances the small mistakes. The elitist mindset needs to die, let's give access to advanced knowledge to everybody and stop putting it behind some unreachable walls.
> Firstly, the unreliability may not be permanent, and we may improve the AI accuracy in the future.
How about we fix the unreliability first and not put the cart before the horse.
> The elitist mindset needs to die, let's give access to advanced knowledge to everybody and stop putting it behind some unreachable walls.
What are you talking about? We have high quality university courses on the internet for free. We have things like the Khan Academy and similar sites. How is advanced knowledge behind unreachable walls?
Without providing context or examples, GPT-4 is already better at answering questions than the average teacher, with the unlimited patience that only a computer can provide.
With context, which Khan academy has in abundance due to their lesson plans and transcripts, accuracy will be higher than even the best teachers and tutors.
Once you give context and known-true facts to best-in-class LLMs like GPT-4, the output is shockingly good.
If you don't know how they are using GPT-4 it's not fair to say it will hallucinate.
As far as I understand the preferred way to use LLMs nowadays for domain specific information retrieval is through embeddings that insert the related context in the prompt. GPT-4 is specially good for this since they increased the prompt size almost by an order of magnitude.
This means that the model can be given a very specific task: to extract information from the context or avoid providing an answer at all.
The answer doesn't rely on the neural memory of the model, since it doesn't need to store information, just understand the task, and they are really good at that.
There are two indicators that this is PR bullshit produced by whoever is trying to capitalize their investment, 1) is the use of the marketing term 'AI' or 'AI-powered' and 2) use of the 'think of the children' trope.
'AI' is a marketing term? I don't see how that's more than saying that 'cloud' is a marketing term - while being useful in marketing, it's just an industry term.
I mean that is but one thing to worry about, we've gone about nowhere with the alignment problem, and we're screaming ahead at full speed with making these things more powerful.
I wonder if this kind of intelligent tutoring could be the answer to Bloom's Two Sigma Problem. The limiting factor with that problem was that not everyone can afford a personal tutor. Having an AI tutor that can breeze through the SAT seems like it should give every student a major boost.
Circa late 2015, early 2016, my cofounder and I spent ~6 month attempting to build exactly this.
Our hypothesis was that students self studying for the SAT used books (eg Kaplan) but the solutions in the back were poor in explaining how they arrived to the solution because they usually just cut to the chase rather than explaining the approach & helping you arrive to the solution.
So we took each practice SAT math question and wrote out a solution spread over four swipeable pages. Our hope was that if you saw the first or second page, you'd arrive at the solution without having to see the answer, thus increasing your chance to solve it on your own next time.
We spent over 100 hours tutoring kids and had them use our MVP to mild success. The obvious next step was to automate the solutions rather than hand write it ourselves - like GPT-4 has done here.
As someone who has logged over 2000 hours teaching, there's a lot of soft nuances and tactics to tutoring. Also, motivation is a big thing a good tutor brings. I don't see this solution as quite there yet but getting close.
The weird thing is how rarely approaches like the one you describe here are applied. It's not exactly a secret that simply providing the answer to math questions is great if someone gets them all correct, but nearly useless when someone gets them wrong. It's a problem that's been solved before - in a classroom, the traditional answer is working through the problem on the board and hopefully students are following along. Accompanying books with videos that work through the problems has been done at least since we had VHS.
My mind went to the same place. Knowing that "the average tutored student was above 98% of the students in the control class", imagine how transformative it will be to lift the average student to the 98th percentile!
This may be the most transformative change to education since the invention of public schooling.
Every now and then I think about Marc Andreessen’s “software is eating the world” article. He was extremely excited about investing in EdTech and HealthTech software companies at the time—for obvious reasons!
But EdTech hasn’t really delivered much value. The modern LMS feels to me like a reinvention of Facebook groups in many ways…
LLMs feel like something that _could_ very well end up eating education, and that’s a really exciting prospect. My personal experience with schooling was less than stellar, and figuring out alternative and more scalable ways to get knowledge into kids’ heads has a lot of possible upside for society imo
> The modern LMS feels to me like a reinvention of Facebook groups in many ways
I agree with your sentiment that there's a lot more potential value to be derived from EdTech than we've seen so far, but even if LMS's are "just" like Facebook groups, that's already the equivalent of hundreds of billions of dollars in economic value.
When I was a kid, if I was sick for a few days, I had to manually copy my friend's notes after I returned to school, and would still feel confused. Now, with an LMS, I have access to all the learning content from home, for my own class, and for resources from similar classes taught by amazing teachers from across the world, at any modality I prefer. This on its own is incredible value. (although with the caveat that being sick is no longer the excuse to rest and relax as it used to be)
It's not the tutoring alone. It's also the social effect. A machine simply does not exert that kind of influence, nor do online courses require the same amount of dedication and ambition. Compare MOOCs: they have incredibly high drop-out rates.
Chances are if you have a tutor you are paying good money for it, and have made appointments at regular intervals, that probably helps with motivation too. I think its possible that chatchpt and other learning ai will end up being like how office hours are used today in college. Like office hours, these tools are available for everyone, but for one reason or another not everyone will bother with them or use them to their fullest.
On the other hand, I think it could also trigger an arms race, and might very well incentivize learning how to write a good prompt more than learning the concept well enough to succeed in a class. This reminds me in my era of the divide in success in my era, between students who were able to navigate the internet to the benefit of their academic success, and those that weren't able to do things like quickly paraphrase a wikipedia article on the subject and use the wikipedia citations as their own, find a relevant flashcard set someone else made and shared on quizlet, or even find pdfs of the teacher's solutions accompanying the textbook. Cheat codes are cheat codes. They will help in terms of your grade but usually hurt in terms of understanding.
> On the other hand, I think it could also trigger an arms race, and might very well incentivize learning how to write a good prompt more than learning the concept well enough to succeed in a class.
I think general public having access to these AI tools is going to create a shift from scholastic aptitude testing to more performance testing. So no more homework, you use your time out of class to train so you can use the in-class time to perform based on the training you've been doing with your personal AI tutor.
Every student might get an absolute boost, but the relative gap may even widen. Consider students who can afford to use cutting edge, expensive to train/run models, vs students who can only use off the shelf, older models.
More likely AI will be cheap and widespread like electricity. Data likes to be free.
Assuming the divide will be real, you can just record a million problem solutions from ExpensiveLM and use it to tune your CheapLM, it will be a specialised cheap model that works ok on your task of interest.
Since using T5 to generate quiz questions in 2019 I've thought this. Transformers now can understand misconceptions, quote source materials... it will equalise education.
A tutor doesn't just imply parents who can afford one but also who care enough to do so in the first place and see the value of an education.
I think you'd find that even by the sat-prep age the actual limiting factor for many (most?) of the tutor-less is a fate-accompli attitude towards their path in life.
Might as well turn on the chat bots at the cradle. Ah, and make sure everybody gets the same quality AI to parent them for it to be egalitarian.
Right... but the whole point Bloom's experiment is that if you give every kid a tutor, they perform (on average) at the 98th percentile. Not just the rich kids who can afford tutors.
I do think there are valid concerns that this technology (like old-school human tutoring) could be disproportionately available to already-privileged children. This could potentially widen the achievement gap. But it seems way more likely that we can get poor kids a laptop to tutor them vs. getting them a dedicated human tutor.
Blooms experiment would probably more relevant here if he gave the kids the option of using a tutor, at their own liberty and pacing (not sure if he did in fact do so).
Everyone has chatgpt here but not everyone will use it. Just like how not everyone shows up to office hours or asks questions of their teacher.
GPT-4 definitely seems to be doing better on a lot of benchmarks and that's impressive. But it still hallucinates facts and I don't think anyone really has a good understanding of when and how that happens. Given that, is it really a good idea to be positioning this model as some kind of factual authority figure?
Human teachers make mistakes too - perhaps even at higher rates than GPT in some cases.
Instead of isolating a single factor of hallucinations I think you have to also consider the cost and experience quality and make a more holistic judgement on whether these models are good or bad.
Yes, but human teachers can eventually come to understand why they made a mistake. And they probably are more able to communicate how confident they are in their answers. I think that's because human teachers have a mental model that they're working from. On the other hand, we have no idea if transformer models are working from anything like a "mental model."
Just because humans make mistakes (or even as many or more mistakes than machines) doesn't mean the nature or consequences of the mistakes are the same.
>Just because humans make mistakes (or even as many or more mistakes than machines) doesn't mean the nature or consequences of the mistakes are the same.
True, but this shouldn't be hard to measure. It's not like we will go from these AIs will go from teaching no one to replacing all teachers overnight.
Give a random sample of classes access to a GPT-n powered tutor, and see if they do better or worse over the next few years.
I remember CS Lewis in "That Hideous Strength" talking about schools; he observed that no parent would let an organization experiment on their children, but that they can easily be convinced to send their kids to an experimental school, which in effect is the same.
That said, you have to compare results from different strategies, or else you'll never improve or react to changing circumstances. I think it's like medical trials; you experiment, but be willing to stop the moment you get good evidence that one option is worse than another.
I've tutored people at times. A good tutor needs to understand the subject very well, so they can not only understand the right answer but also figure out why the student is coming to the wrong answer.
I personally think that assigning GPT as a "tutor" is devaluing the real skill involved in tutoring and I doubt it will work out.
I would like to divert your attention to this recent post: https://openai.com/research/gpt-4 specifically on the topic of "Steerability: Socratic tutor" (Sample 1 of 3)
I was trying to get it to teach me Ordinary Differential Equations and it just kept getting everything wrong. It has no concept of checking its work etc, that example makes it appear like it can but I guess it's just getting lucky for linear equations.
Very interesting, thanks for referencing it as I had missed it.
There have been occurrences of people asking chat gpt to ignore instructions on the initial prompt: have these been solved? It's not clear to me if the first prompt is given any extra weight. Or maybe during training of version 4 they have heavily penalized these sorts of attacks to make it more resilient.
In my brief asking ChatGPT question and to rephrase/clarify, it seems at least as good as the average tutor/teacher/TA which for me has been marginally better than warm body rephrasing the theories and working through textbook problems in front of you.
And in time, much better. To the point of raising the standard of education for all of humanity with access to the internet. I'd count on it "working out".
I'm 42 and still don't always like asking questions for fear of looking stupid if I ask one the wrong way. chatgpt has none of that because it's just an ai and won't judge me or make my imposter syndrome worse.
My daughter does children's martial arts. Often in the dojo there's an atmosphere where the yellow belts are helping the white belts, the red belts are periodically chiming in, and the black belt adult instructor is free to float around the room supervising everything, giving individual attention where it's most effective, but leaving the lower ranks to take care of the lessons they're capable of giving.
In the same way, AI will be able to make an expert human tutor 100x more effective by escalating the tricky cases but learning how to handle the ones that come up over and over.
This is something I’m working on. I think having a good understanding of the material for good explanations is doable in the near future, but the main problem is actually building a good model of what the learner knows to guide the LLM
Highly recommend the Neal Stephenson book The Diamond Age: Or, A Young Lady's Illustrated Primer, for an interesting exploration of a custom AI tutor for each student.
In that book, students have a "magic book" that teaches them lessons in a story form while encouraging certain life paths. It's pretty fascinating to consider the implications and whether that's something we'll want, bc it may soon be possible.
Khan Academy's integration of GPT-4 already steers students towards particular modes of thought. According to the article, they want "to get students thinking deeply about the content that they’re learning" and in the examples, Khanmigo prompts students to recall information from the lesson rather than directly explaining how to solve the problem.
I think what Khan Academy has done here is desirable, but just as Stephenson's Primer enforces its creator's values through AI tutoring, Khanmigo is enforcing Khan Academy's values. It's easy to imagine how someone with more authoritarian values could use this same technology (e.g. to teach students to follow instructions without doubt) to indoctrinate students with scale.
AI Teacher: "And that was how the civil war ended... Now remember kids, the most important thing you can do is be nice to each other and contribute more data and processor time to growing the AI. Thank you, have a good day"
If openAI can do it, so can other companies in time. At best they would cause competitive systems faster and likely with fewer safeguards by hard blocking languages or users.
More authoritarian values? Isn't this what's happening now from the perspective of foreign countries? The USA enforcing values foreign to these countries.
Unfortunately the guard rails of the current systems prevent, in my opinion, interesting antagonists in fiction or games. I’ve experimented with some writing and when you can convince ChatGPT to give you some examples of negative behavior it will attempt to bring things back to a wholesome version. GPT-4 having more effective guard rails makes this problem worse.
For younger readers Monica Hughes' classic "Devil on my back" is more explicitly about a society where those who can interface with the web via computers installed into them become higher caste and those who cannot become lower caste, resulting in a revolution. Definitely aimed at a younger audience than Stephenson.
The eventuality I see is everybody talking to each other through LLM's, and the world ending up in some kind of tower of babel situation. You wear an AR/MR headset, and all conversation is dual translated, with some kind of bytecode intermediary. Each persons language evolves to be incompatible with anyone else's, and only your LLM can understand you. The power goes out, and nobody can talk to anybody.
>It's pretty fascinating to consider the implications and whether that's something we'll want, bc it may soon be possible.
I wouldn't be surprised if it told people to take some path of least resistance to $10+ million so you don't have to work to survive, then you can mostly do whatever you want.
Having heard many recommendations, I tried reading that book more than once, but have real issues with Stephenson's writing style there.
Could you please recommend an overview of the qualities of the AI tutor that I could go over instead?
I’m sitting on the couch with my new grandson. He’s six months old. What is school going to look like for him over the course of his education? Should be interesting.
My daughter is in 2nd grade and math is her favorite subject. I liked math too. I've looked forward to helping her learn more advance math through the years, but a doubt has entered my mind; will I have any reason to help her?
AI can efficiently teach your daughter how to solve math problems, but as a human, you can provide her with the context and understanding of why and what she's learning. AI is great at handling the technical aspects, freeing us to focus on cultivating our uniquely human qualities of empathy, creativity, and critical thinking. Letting the AI teach the how and the human teach the why and what can lead to a more holistic and enriching education.
I'd like to agree, but it does seem like AI is getting better rapidly at providing the context and the "why" too (as well as creativity and critical thinking)
the language models can tell you how to do a fourier transform, and then you can have a conversation about what it all means, why it's important, the discovery, etc
[me] why are fourier transforms relevant to learn? what is the history behind them
[bing] Fourier transforms are relevant to learn because they are used in many fields such as signal processing, physics, and engineering. They are used to convert a function from the time domain to the frequency domain, which can help to identify the frequencies present in a signal and in what proportions. The Fourier transform was first introduced by Joseph Fourier in 1822 as a way to solve the heat equation. Would you like more information on the history of Fourier transforms or its applications?
This sounds like an argument from when calculators were invented. We will keep getting more powerful technologies, and you will need to understand them. What we should be learning will continue to gradually shift.
Absolutely. With mobile internet/search engines, knowing how and where to find information became a more important skill than having to memorize that information
Honestly it depends on the growth rate of our models and if they converge with human intelligence without requiring have the computing power on the planet.
If we get slow incremental progress over the years then it will act like an interactive personalized encyclopedia that can teach you based on your strengths and help bolster you where you are weak, as the technology improves that is.
Now, it's still scary as hell in some ways even in this slow growth mode. The authoritarians will love to have a system like this that just forgets particular parts of history and make reports back to the mothership if you ask too many 'wrong' questions. So actual implementation will be interesting to see.
Now, if we have explosive growth to AGI and beyond, then pretty much every bet about the future is cancelled and we'll have more luck in taking random guesses, gluing them to the wall, and then throwing darts at them blindfolded.
Xaeon-12 wakes up at 8am, has a bowl of cereal and put's on her VR goggles. She is greeted by her tutor Tom. Tom will guide her through advanced "AI prompt" class today. Xaeon-12 learns to create her own short story based on her dreams that were recorded the night before. 12am, time for lunch. After lunch, the kids all connect to play Roblox VR for 30 mins. Suddenly, the doorbell rings, but no one else is home. Fortunately, Xaeon-12 doesn't have to answer the door, as everything is automated. The mailbot dropped off some packages, before flying back to the zeppelin mothership to retrieve the next package. Xaeon-12 is just finishing her last class, "circuit boards", when her dads come home from the golf range – with an abundance of free, renewable energy, working for a basic income has become irrelevant.
It's funny that you mention "The Fun They Had", because if you visit its Wikipedia page now, the summary looks written by a bot. It looks like "basic" English, only with a weird quality to it -- either written by someone whose grasp of language is very poor, or by ChatGPT. I'd say worse than ChatGPT actually: the sentences are short and almost mechanical. Go look at it.
The edit changing a way more human-sounding plot summary to this version was made on September 2022, in case anyone is wondering.
PS: if GPT derives much of its data from Wikipedia, and people start using GPT to write Wikipedia articles, I wonder what kind of strange feedback loop we're getting into.
The cynic in me wonders if "AI" has realized that the most efficient way to take over the world is to teach the next generation to be dependent on it for learning.
> “Once men turned their thinking over to machines in the hope that this would set them free. But that only permitted other men with machines to enslave them.”
To me, this quite really captures the most likely immediate negative of a huge AGI breakthrough. Not that it'll rise up like the robots in The Matrix movies... but rather simply having it owned and controlled by an American mega-corporation or a state actor... quietly influencing everything and everyone behind the scenes.
There is no way I can imagine to define true neutrality. You need to make value judgements. Everyone has their own set of values they judge the world with.
There is no way I'm aware of to completely avoid bias. Ultimately, it has to be up to someone to define what bias and neutrality both mean.
There is no algorithm for "truth". As soon as you stray outside of extremely well-defined systems like math and physics, ambiguity enters the equation and value judgements need to be made. Who will make those value judgements?
> most likely immediate negative of a huge AGI breakthrough.
I mean if the machines are generally intelligent in the human sense and unaligned with human interests I'm assuming that the mega-corporation would be its own slave to the machine too. The machine would give the mega-corporation more power, simply because that's the number one way to increase the power of the machine. If super intelligence is possible, it seems that a greedy power hungry corporation would gladly create it to gain more power, and then lose control of the situation.
this was my '1984'. i was 15, had it in my head that i was destined to become a revolutionary.
i gave it to a friend and they came away feeling like they needed to become a teacher; i came away wanting to be a novelist, or a father. in any case, it definitely changed my world view. not sure how i'd feel about it nearly a decade later, but i'll take this a sign to re-read. cheers!
A plot point in an episode of Person of Interest, where AGI "Samaritan" funds a charity (free tablets for students, Samaritan access pre-installed) to take over education and recruit more mercenaries.
There will probably be a janky but usable open source version of the model and tool eventually. The "Linux" option. That way the 1% that cares about this stuff can do the self host route and retain control by them controlling what the model tells them(hopefully).
They literally announced that they're testing this on a small sample of kids whose parents opt-in. I don't see any better approach, seeing how the sentiments of "untested" and "using kids as guinea pigs" are contradictory.
A lot of what GPT-4 can return is ideas based on assumptions and beliefs. Who chooses what of these to train the system? Those subtle ideas will then lead students who are mostly children.
I don't think this is such a problem if it is teaching calc-3 and stays on topic. If it's teaching humanities I can see there would be more of an issue.
Who is "we all"? For instance, the US is 5% of the worlds population. What we call the west is 20% of the global population? Not even to a quarter of it. Most of the population is outside the west. And, within the west there are a number of topics that people don't agree on which are more base.
I wonder if the students can direct the AI to give them questions similar to what would show up in the tests. Do the teachers use the same AI as the students?
I don't think tests will work at all like that much longer; they'll be more like oral exams, where you have a back-and-forth conversation with an AI that probes the breadth and depth of your knowledge with sufficient variability as to detect overfitting (i.e., memorizing the answers to a few previous exams).
> "It's important that you learn how to do this yourself!"
I find this response by the AI tutor to be unbearably ironic - if there's an AI that could do it better than me, and well enough to teach me, then it seems quite unimportant that I learn to do it myself. While of course I see a benefit to some people learning mathematics from scratch, for advanced research and continuity purposes, at this point I think that saying that it's important for "everyone" to learn math (at least beyond the very basics) is almost equivalent to saying that it's important that everyone learn how to grow grain, weave fabric and mix cement.
I had a similar reply in a separate thread, but human-to-human interaction is extremely important for young children. Teachers teach much more than school subjects in the classroom - they teach how to behave, how to ask for help, how to interact with others, etc. I often feel like teachers of young children teach them how to be "human".
I think as people age they can probably rely more on Khan Academy, etc.
Maybe this will change in the future as we find better ways to teach young children with technology, but I would be apprehensive with dropping a 1st grader in front of a screen and having zero human interaction.
> teach how to behave, how to ask for help, how to interact with others, etc. I often feel like teachers of young children teach them how to be "human".
Isn't that the job description of parents and/or day care?
Small modern families don’t have a ton of people around and the relationships between those people tend to settle into equilibrium. It’s not really adequate for practicing the adaptive social skills we usually (used to?) expect from adults.
And day care can provide more of that, but is naturally supplanted for 25-30 hours per week once kids start attending school.
Socialization is absolutely a function of schools. You can expect big societal repercussions if it were substantially removed.
Yes, but at that age it's important to be taught those things and be socialized through all aspects of life. You can kind of think of it like learning a language - you learn it faster when you're fulling enveloped in it.
It's easy for parents to miss things that a public school teacher with an education degree would not.
Also, children can be taught all of that by their parents and through day care, but still fail to do what they've been taught in a school setting. It's important for teachers to teach those things or reinforce those teachings.
I highly recommend asking teachers of young children about this, or looking into what education materials teachers are trained on. Many schools have professional development days, so what a teacher knows and works with changes/updates often.
One important function of schools is not only to offer education, but to force people to become educated. (Nearly) everyone in a society sharing some common skillsets (like writing, basic math, critical thinking etc) is hugely beneficial, and allows individuals and society at large to get long-term benefits where on their own they might have chosen short-term benefits. Though arguably that effect is strongest in the early grades.
b) stability/refuge for kids/adolescents from unstable families
c) interpersonal connections with experienced humans (opposed to isolated discovery)
d) pedagogy
that was its main purpose for a while now; most people learn complex concepts through youtube these days. an engineers/hackers ability to dive into topics on their own doesn't map to the requirements for the healthy development of a child.
Its not there yet. Im testing out a teaching assistant and it doesn't catch wrong answers (eg question was whats 2 + 4, answer given was 4). I think its good for explaining concepts but not yet structuring the lesson or basic reasoning (this is 1st grade)
Just when many education systems are struggling, this comes up. It's gonna be interesting to see if it helps self organizing and self learning (removing the negative side effects of group learning in school settings) or if it will just add noise and chaos.
In this case, as I understand it, the idea is that the existing Khan Academy system generates a random question, the AI helps the student solve it, and then the student provides the answer to the original system, which then validates it.
I would also assume that Khan Academy could inject the true answer into the GPT prompt and instruct the AI to hide it from the user, but rely on it in its explanation, in order to reduce the risk of misleading outputs.
This is why it doesn't matter if AI lets students cheat on essays: Essays will soon be obsolete — not just for students, but for everyone.
The two main purposes of an essay are to teach others about something in your head, and to develop and refine such ideas in the first place.
Very soon, it'll be much faster, easier, and more prolific to spread an idea by conversing with an AI about it and letting it go off and talk to others about it. If some of those people have interesting feedback, questions, or find a flaw, the AI can distill it all down to a time-efficient debriefing it can come back to you with. And an advanced AI makes for a great partner to bounce new ideas off and chew them over with.
I mean honestly, if I took the corpus of our companies support data and used it as a training set for our first line support, I believe that GPT could answer the questions better than they do first shot.
No idea what the cost difference would be, but for many businesses I could see the AI response be a costs savings.
I'm currently homeschooling two teenagers and one of the most interesting aspects for me is learning all the material that I 'supposedly learned' way back when I was in high school.
I graduated, went to a great university, and yet virtually all of the stuff that we are learning in the current HS curriculum is new to me despite my having taken the same classes. I honestly don't know how I graduated high school, except to consider that I was very good at parroting back recently received information.
Most of what students are learning will be obsolete by the time they enter the job market, but on the plus side they will have forgotten it anyway. The most important thing from my perspective is ensuring that those topics don't use up too much of their time as I think most intelligent people are able to acquire the knowledge and skills they need as they come to need it.
I assume the one from the research page would work:
"You are a tutor that always responds in the Socratic style. You never give the student the answer, but always try to ask just the right question to help them learn to think for themselves. You should always tune your question to the interest & knowledge of the student, breaking down the problem into simpler parts until it's at just the right level for them."
Given that learning will probably soon not be required anymore to live a successful life (AIs will do everything better than you), I guess such learning-as-a-hobby apps will be the remaining niche to entertain the humans on basic income.
I wonder how many people will actually make use of it, though, and how many are instead content with being wireheaded by the various AI-based content apps that will appear, full with virtual partners (see Replika etc.).
Learning is absolutely going to be required, we are just going to need to spend more time learning patterns and motifs rather than lots of facts. Learning facts and trivia was already devalued by everyone having a hyper-encyclopedia in their pockets, this will just extend that. Knowledge of patterns and motifs in domains of information will let you work together with the machines in a way that an untrained person will not be able to match.
I admire your optimism if you think that a human in the loop will add any value. Other than for humoring/entertaining the human, of course, which is what I meant with learning-as-a-hobby.
Also, I'm not educator or very knowledgeable about the theory of learning, but I can imagine that "learning facts and trivia" is to some extent a necessity for exercising our brains when we are young. Of course, learning patterns and theories is the end goal, but some rote learning may be necessary on the way.
Humans are very efficient computers for many types of tasks. Humans also have a large body of multi-modal data with detailed annotations inferred by observation of other people's responses passed through our mirror circuitry that is unavailable to the AI, so they can sometimes provide relevant guidance even if they aren't better at the whole task. Finally, human neural networks are chaotic, so they can potentially find better global maximums in some cases than a model, which might end up trapped in a local minimum due to its inability to entertain sufficiently unlikely hypotheticals.
My job is fighting IT crime. Right now this means reverse-engineering malware, tracking criminal groups in various graph databases, and writing Python scripts. Maybe in 20 years I will reverse-engineer AI models (using AI tools), discuss criminal groups with ChatGPT, and prompt-engineer utilities. But I don't see my (broadly understood) job becoming obsolete as long as people are capable of crime and software exists.
"Learning facts and trivia was already devalued by everyone having a hyper-encyclopedia in their pockets, this will just extend that."
I hope you don't end up having surgery done by a doctor who has to check a "hyper- encyclopedia" in his pocket because learning facts is supposedly obsolete...
I don't know why comments like this take for granted "basic income" as like this inevitability. We currently let people all around the world die on the streets for not participating well enough in the economy. Why would this moment in history be any different? You think just because you get paid a lot right now, the guys in charge will make sure you are ok in the future? That is not really how this whole capitalism thing works unfortunately.. No amount of technology will save us from waged labor in itself, its simply the character of our labor which will change.
What you want is communism, and buddy, I'm with you!
Oh yeah, the basic income thing I just threw in there because I think even in the best case where the disempowered and soon-to-be unemployed workers will be cared for, humanity will be reduced to AI-governed content consumers without much agency left (cf "The Machine Stops").
That said, I agree that it's more likely that the almost all of the returns of this AI-based rationalization will flow towards the wealthy, as usual.
I am not even that optimistic about it. What would be the incentive to keep humans entertained in a world run by AI? Why would they keep billions of us around? I won't make any sense economically.
This seems like a no-brainer next step for OpenAI, to derive an AI instructor from GPT-4 without a partner like Khan Academy. I suppose that may require a "truer" multimodality of ChatGPT being able to generate images or visual patterns. Imagine being able to onboard new employees via such an agent or learning about complex subjects.
When I was a kid, I remember one of my teachers saying the same thing about the internet ( google, computers, etc ). Us kids were so lucky to have such powerful tools and everyone in the future will be smarter for it. Turns out, people are dumb as ever and these tools made us lazy and dependent. Why memorize the multiplication table when you could ask siri?
I think it's fair but there's a reason why it almost seems natural to raise the question in this scenario. It's OpenAI's hypocrisy and doublespeak. It's their non-stop messaging about equality, alignment, unbiasedness, etc. They can't have it both ways and they will be judged for it.
I don't know that I would classify Khan Academy as a tech giant. For the sake of argument let's say they are. Yes it is fair for them to get access earlier. Why would it not be?
I'd assume that's because the Gates Foundation is one of the biggest financial supporters of Khan Academy, and they can reach the right ears at Microsoft and by extension OpenAI. As always, the right connections are everything
OpenAI is an opinionated startup, they don't owe everyone the same things. Probably they thought KhanAcademy would be a very ethical way to use GPT-4 for good and they gave it early access.
