Why Sal Khan's AI Tutor Failed: He Learned by Making, but Taught by Telling
Why Sal Khan't: On Learning by Making but Teaching by Telling
Sal Khan launched Khanmigo, an AI chatbot tutor, with world-changing ambitions, but even he admits it was 'a non-event' for most students. This piece argues the failure wasn't a lack of effort but a fundamental design flaw: Khan's own learning was active, purposeful, and driven by Dewey's four impulses—inquire, construct, express, and communicate—yet he offered students only the passive consumption of his explanations. The author traces this to a lack of pedagogical knowledge, suggesting that edtech's focus on delivering content misses the real driver of learning: purpose and active construction.
He gave them the destination without the journey. And because the journey is where motivation lives, where purpose lives, students quite reasonably declined the offer.
- skulk
Overall, I do agree with the thesis but:
> Now… how does Sal Khan want my kid to learn?
> Watch a video.
I find this a little uncharitable. I learned a lot of math at a young age from Sal Khan's early videos (back when he used to erase by switching to a black pen and scribbling across the entire screen) and, to put it in the author's terms, it served as an easy-to-digest scaffolding on top of which to build a deeper understanding. I don't think Khanacademy ever sold itself as a magic learning resource that removes the need to work out problems yourself.
- dctoedt
From what I know of the Khan approach, it's akin to "flipping the classroom" [0], which has gained pretty wide acceptance. It was pioneered by Harvard physics professor Eric Mazur.
(I use the flipped classroom approach extensively. But the legal profession being the eager embracers of novelty that we are </sarcasm>, I refer to the approach as "enhanced Socratic method," given that Socratic method was something we all endured in law school.)
For math and engineering courses, the basic flipped-classroom approach was used at West Point for over 150 years: In the so-called Thayer method, students study textbooks in advance, then in class, they work on problems on chalkboards. The instructor moves from chalkboard to chalkboard, coaching the students in real time. [1]
(I do much the same thing in my contract-drafting classes: In small groups, students do drafting exercises using online whiteboards. I lurk on my laptop, quietly coaching in person as needed.)
[0] https://bokcenter.harvard.edu/flipped-classrooms
[1] https://flippedlearning.org/syndicated/avoiding-flipped-lear...
- ChaitanyaSai
A comment I had made on Dan Meyer's article, also relevant here (Also, as others are pointing out here, there is a crucial difference between the videos and their chatbot; you can inspire with videos because your human interests are on display):
This resonates so much with me. I am the co-founder of a microschool and also a neuroscientist. And these two have intersected so much in fascinating ways. Four years in, the big lesson for us is that for learning to make sense it should have the following progression: meaning, motivation, mechanics, measurement.
Conventional schooling, and everything designed to help with conventional schooling like Khanmigo give in to this systemic gravitational pull of measurement (we must know what they are learning). And from that ensues your standardized curriculum, measuring the visible and easy, because that's easier than something like "mathematical thinking", and the focus on mechanics (multiplication tables, fractions, chemical names, whatever). And in doing this we leach out all meaning and focus on the drills.
Chatbots, the way they are set up now, are doomed to fail because the entry point is mechanics. "How do I solve this particular quadratic equation?" It's a meaningless question to a child because it simply does not connect to their world. We cannot start with that and work Socratically towards meaning. We must instead offer or build a world where meaning and sense-making are possible. (Wrote about this here https://blog.comini.in/p/ […]
- PostOnce
Khan Academy has a problem lately:
Go visit a page like https://www.khanacademy.org/math/linear-algebra and count the number of:
* cookie banners
* donation panhandling modals
* account signup widgets
* gamification modals to tell you about leveling up
* etc
It does not communicate that their primary goal is to educate, it feels more like the primary goal is fuelling the bureaucracy behind the charity.
- narmiouh
This articles core argument appears to be that a live instruction is better than a video because there can be feedback at the moment of confusion while a video can't. But that depends on student's impulse being triggered and I'm not sure if its any better or in some cases can be worse if the teachers content is not as thorough as Khan's or anyone else's video - as these benefit from a worldwide audiences feedback on correcting any confusing or unclear parts.
In the classrooms I see through my kids, the teacher is doing roughly what a Khan video does: walking through a worked procedure at the front of the room. If a video fails to provoke the "wait, why does that work?" moment, then a teacher following the same flow likely fails in exactly the same way as the author claims and missing the opportunity for hands'on learning. A video actually allows the learner to go back and review content if anything was unclear, and presumably any interactive AI agent can help with the feedback part even though the teacher in the video isn't.
I also wish the piece had stayed with Khanmigo, where nobody has a good theory yet on where the ai tutoring loop actually breaks and why people fail to engage it and that putting it in their face may not work, why exactly personalization at that scale hasn't produced the gains it promised.
Also irked by the title built on a pun on Khan's name, sets up an awkward framing that kind of lets me know it may not be unbiased.
- tbonza
As someone who has earned over 3 million points on Khan Academy over a 3-4 year span about ten years ago, I am very appreciative of my experience. What I appreciated most is when Sal would derive formulas such as the quotient rule in Calculus from the power rule by using negative exponents. Deriving formulas rather than memorizing them opened up a new perspective for me.
This article appears to focus on the chat bot which I have not used and was not part of my experience. At the time, I had a subscription to Wolfram Alpha that was helpful if I got stuck on an intermediary step.
The high level of goal of the post which is directed at asking good questions raises a point that seems valid to me.
- wafflemaker
Is the article about something else than how math on Khan Academy was thaught?
I remember minimal videos (3-13 min, usually 5-8), and then exercises drilling the techniques until you understand it and don't make mistakes. If you got a perfect score (no mistakes whatsoever), you went a little faster.
That's learning by doing, or am I wrong?
Or are there no exercises for things other than math?
- vanderZwan
Whenever this topic comes up I feel the urge to point out that not enough people have read Audrey Watter's "Teaching Machines: The History of Personalized Learning", or her blog. We might stop repeating the same mistakes over and over again if we did.
[0] https://direct.mit.edu/books/book/5138/Teaching-MachinesThe-...