Dylan Smith retweeted
I don't think I'd use the word "grim," but I think I agree with the underlying point.
The study of evolutionary biology has shown that social learning (that is, education coming from other people) is quite literally what makes us human. We are so evolved for it that we cannot survive without it. Other animals also engage in social learning, but we are both the most helpless without it and the most powerful with it. The development of human teaching and learning is what made us human.
In that context, whether AI can more efficiently correct errors in the understanding of the quotient rule is sort of beside the point. Perhaps it can, and if it can, it should be used for that, and can be a great help.
*But that is not what education is.*
In the broader context of the role education actually serves in human life, there is no possibility in which an LLM gets so good that it replaces the human, because the human connections are entirely the point.
Replying to @DonaldClark @dylanwiliam
Grim reading, to be honest. It feels like we're ushering in a Big Tech takeover of one of the last bastions of meaningful human connection: education. Another opportunity to commodify something that should be a public good.
There's an underlying assumption here that AI will inevitably be smarter than humans, so why bother with human teachers at all? But surely education is about much more than delivering information efficiently?
Young people are completely right to be suspicious of Big Tech, its resource hungry data centres and the enormous power concentrated in private hands. And now education is next on the menu. Health, education, human interaction itself.. is there anything the billionaire class won't try to turn into a market?
The question I'd ask is who actually benefits from all this? And at what point do we stop asking whether AI can replace teachers and start asking what we lose when it does? 🤷🏻‍♂️
Dylan Smith retweeted
Very important for Ontario parents to know.
Accommodations are not the same as Modifications.
Please share this widely.
#OntEd #IEP #SpecialEducation #OntarioEducation #Ontario
Dylan Smith retweeted
Replying to @KateJones_teach
I consider myself to be a cognitive and educational psychologist as I've done cognitive and educational research and written >450 articles about the two. Learning scientist and learning science, as I've said before, is empty or could be anything and says absolutely nothing.
Dylan Smith retweeted
Do surgeons and pilots use checklists or scripts ? How are they different?
Would a surgeon follow a script during an operation? Why not?
I guarantee teachers are using mental and written checklists all the time.
Dylan Smith retweeted
Pedagogy is not like surgery or pharmacology. It is more like eldercare, psychiatry, or community medicine. Its effects cannot easily be isolated or generalized without reference to person or context. Research is a guide but the craft of practice and the knowledge of the people and place, in the moment have to be taken into consideration. It is that judgment that defines the professionalism of teaching.
Dylan Smith retweeted
Replying to @ShakinthatChalk
The principles are so clear that whole swathes are still doing di wrong, and edutwitter experts need to come to the rescue to correct lethal mutations.
Dylan Smith retweeted
The two camps are actually “Do As You’re Told” and “Let’s Figure This Out Together”
Dylan Smith retweeted
This seems pretty important ---------Project Follow Through: A study so old it makes me look young mathematizing4all.com/2026/1…
Dylan Smith retweeted
Replying to @cbokhove
The other side of the coin: there's nothing wrong with apples and pears.
If countries A, B, and C delivered public ed in identical fashion, there would only be more of the same for them to observe from one another. It’s a rule of thumb in life. Diversity is a positive because it signals there is something to be learned. That’s why Darwin waxed on about divergence and modification, why researchers pay attention to outliers, and why many educational systems flew observers into Finland.
A word re discussion here on X... There are many apparent universals in education that educators the world over could productively discuss, e.g., terminology, subject matters, student behaviour, child development, etc. There are also many apparent differences in ed systems around the world that could provide fodder for discussion and learning. And that’s not all because no sooner do conversations begin and we start to realize that we could spend much time discussing how the "universals" actually vary and how the "differences" are actually similarities in disguise.
Perhaps educational research could help us? Well, no, not really, a century of equivocal research findings do little to clarify the status quo of public ed, let alone steer us into the future.
And then there’s ideology, the tragically confounding layer that’s overlaid on top of all of the above.
For all those reasons and more, discussion here on X is greatly challenged. Even with fresh topics in new conversations, “progress” soon lock ups and recedes from reach because even though all discussants have a voice, few seem to understand or care that greater discipline and skill and sensitivity is needed to participate in a cross-border conversation.
Dylan Smith retweeted
Archer’s words also hilite the need for *responsible* academics to acknowledge the diversity and equivocality of evidence.
Scholars such as yourselves have a choice: either help others overcome cherry-picking and confirmation bias to find value in discussion, or add to the BS.
Dylan Smith retweeted
Where’s the buzz? Great teaching needs evidence, consistency and strong routines. But it also needs personality, judgement and a little bit of magic. Enjoying the writing process again, my latest blog on why we shouldn’t take the “teach” out of teaching.
smithsmm.wordpress.com/2026/…
Dylan Smith retweeted
"Considering all of these factors, anyone claiming that all teachers should teach in a particular way, or for that matter, all students should learn in a singular way is a charlatan." - Larry Cuban
Dylan Smith retweeted
Trust might be the most important condition to cultivate in our schools.
In this #LeadershipHandbook post, I looked at the wider literature on relational trust and proposed a unified model.
mrnickhart.substack.com/p/le…
Dylan Smith retweeted
"Novices learn best from explicit instruction" is a robust (though far from universal) finding.
But what is a novice? In American educational discourse, students are often considered perpetual novices, because each math class, for instance, introduces a new fact or procedure.
That student=novice appears so obvious it is rarely questioned. But this is less a fact than a perspective, a choice.
Japanese and Singaporean math classes do not treat children at perpetual novices. Neither do Socratic seminars. Neither do basketball practices for 10 year olds.
In fact, rarely is a child a true novice at whatever you're doing with them. My daughter is 9 years old and was asking me this weekend about exchange rates, and we got into a conversation about international monetary policy. Even there, she had background knowledge to build on!
And that's before you start talking about something like literary analysis, a field in which she already has maybe 100 hours of experience (what is a discussion of a Disney movie from the perspective of the villain, if not a form of literary analysis?) which itself builds on maybe 1000 hours of reading. Again, not a novice.
Knowledge matters. But students rarely have no knowledge of a subject, just as teachers rarely have perfect knowledge of it. That fact provides opportunities for unscripted discussion and mutual learning that are the foundation of meaningful human relationships.
Dylan Smith retweeted
“How about a guide who is knowledgeable enough to periodically stop and deliver some sage advice, and who is wise enough to know when to lead and when to get out of the way? It won’t sell many stories, but it will get the job done.”
CURMUDGUCATION: Here Comes Direct Instruction, Again curmudgucation.blogspot.com/…
Dylan Smith retweeted
Replying to @alexanderrusso @ehanford
I’ve been around teachers and schools for a long time and I have never heard anyone voice that fear
Dylan Smith retweeted
I tried to explain why I didn't like the initial run of "Sold a Story" very much. pershmail.substack.com/p/emi…
Dylan Smith retweeted
Replying to @mathillustrated
The view that DI is “passive” rests on 3 matters of fact:
❎ Ss remain in seats, aren't “active learners” in it's long-accepted sense
❎ requiring all to think/act identically isn't an opportunity to participate
❎ raising one’s hand, chanting are compliance demands—not agential
Dylan Smith retweeted
Replying to @C_Hendrick @mattyglesias
Phonics reckoning? What phonics reckoning could @mattyglesias be referring to? He couldn’t mean the decoding push in the wake of Sold a Story because that didn’t really work out, did it?
Perhaps he means the scientifically based “Five Pillars” approach in USA following the “Reading First” legislation in 2002. No, national testing showed that basic skills programming didn’t lead to gains either despite 6 years of a billion-dollars-a-year support.
Well, maybe he means the push after “The Simple View” of Reading in 1986, or the one after Jean Chall’s review in 1967. Or the one after Flesch published “Why Johnny Can’t Read—And What You Can Do About It” in 1955. No, those didn't lead to phonics nirvana either and just turned around too.
I wonder what phonics reckoning he was talking about?




















